Pellet Injection Research Articles

Fast plasma dilution in ITER with pure deuterium shattered pellet injection

JOREK 3D non-linear magnetohydrodynamic (MHD) simulations of pure deuterium shattered pellet injection in ITER are presented. Considering a 15 MA L-mode plasma with a thermal energy content of 36 MJ from the non-activated phase of ITER operation, it is shown that such a scheme could allow diluting the plasma by more than a factor 10 without immediately triggering large MHD activity, provided the background impurity density is low enough. This appears as a promising strategy to reduce the risk of hot tail runaway electron (RE) generation and possibly to avoid RE beams altogether in ITER, motivating further studies in this direction.

Structural Design for ITER Gas Injection System Gas Fueling Gas Valve Box

The ITER gas injection system delivers gases from the tritium plant to the vacuum vessel, fueling pellet injection system, and neutral beam for plasma operation and fusion power shutdown. In this system, the gas fueling (GF) gas valve box (GVB) is an indispensable part that mainly provides functions of gas throughput control and measurement of gas pressure, flow rate, and temperature. The preliminary structure design is largely driven by the requirements of magnetic field compatibility and limited integration space. A strong magnetic field of over 0.2 T exists around the GVB locations, so a magnetic shielding design is required to ensure the normal function of susceptible components. Instead of the previous overall shielding, a local magnetic shielding has been developed by a validated analysis method. As a result, the total weight of the shield has been reduced from over 7000 kg to about 200 kg. Furthermore, considering the limited space reservation, a highly compact flat layout for the GF GVB has been developed to ensure enough maintenance space in front of it. In addition, other requirements such as structure integrity under various load combinations, leak detectability, in situ maintainability, etc., have all been taken into account.

Preliminary Experimental Results of Shattered Pellet Injection on the HL-2A Tokamak

A shattered pellet injector based on in situ technology was installed on the HL-2A tokamak, and preliminary experiments were performed recently. In this paper, a fast current shutdown experiment introduces shattered pellet injection (SPI). In comparison with spontaneous disruptions and massive gas injection, SPI has advantages for disruption mitigation. The experimental results show the hard-X-ray radiation intensity (40 to 60 keV) rapidly falling from 20 to 0 when SPI is used. From this, we can infer that runaway electrons are suppressed. This observation indicates that SPI should be a good candidate for current fast shutdown in the future.

Active Radiative Liquid Lithium Divertor for Handling Transient High Heat Flux Events

The extreme heat flux anticipated in fusion reactor divertor plasma facing components (PFCs) is perhaps the most challenging technology issue for fusion energy development. Most divertor PFCs are designed based on the maximum steady-state operational limits. Application of lithium (Li) in NSTX resulted in improved H-mode confinement, H-mode power threshold reduction, and reduction in the divertor peak heat flux while maintaining essentially Li-free core plasma operation even during H-modes. These promising Li results in NSTX and related modeling calculations motivated the passive and active radiative liquid lithium divertor concepts (RLLD and ARLLD) (Ono et al. in Nucl Fusion 53:113030, 2013; Fusion Eng Des 89:2838, 2014). This radiative process has the desired effect of spreading the localized divertor heat load to the rest of the divertor chamber wall surfaces, facilitating divertor heat removal with relatively small amount of lithium ~ few moles/s to handle the expected steady-state high divertor heat load. However, in addition to the high steady-state heat flux, the fusion reactor divertor PFCs could also experience significant transient heat flux such as ELMs and/or other magnetic reconnection events which can deposit large transient heat flux onto the divertor PFCs. If unprotected, it could damage the divertor PFC surfaces which could lead to a highly undesirable unplanned shutdown for PFC repair and/or replacement. In this paper, we explore feasibility of LLD and ARLLD concepts in handling such transient heat flux to protect the divertor PFCs from the extreme transient heat flux while maintaining the normal plasma operations. We also suggest a possible implementation technique using inductive pellet injector for the reactor PFC protection from transient heat flux which can be tested on NSTX-U.

Shatter Thresholds and Fragment Size Distributions of Deuterium–Neon Mixture Cryogenic Pellets for Tokamak Thermal Mitigation

Reliable mitigation is necessary to eliminate the detrimental effects of a disruption event in large high-current tokamaks such as ITER. To avoid serious damage to plasma-facing components during the thermal quench phase of a disruption, material is injected to radiate the plasma energy over the inner surface of the machine. The most promising method of material injection is a process known as shattered pellet injection (SPI). SPI utilizes cryogenic cooling to desublimate gas into the barrel of a pipe gun to form a solid pellet. High-pressure gas or a mechanical punch is used to dislodge the pellet and accelerate it into a bent tube to intentionally fracture it. Pellets made of a mixture of deuterium and neon are likely candidates for thermal mitigation. The survivability of these pellets throughout their flight path, before striking the shatter tube, is essential for reliable SPI operation. Experiments were conducted to determine intact speed limits for various mixtures. This paper outlines the details of brittle fracture theory and compares a theory-based model to experimental results from various mixtures of deuterium and neon pellets.

TOKES simulations of mitigated disruption thermal quenches in ITER

• TOKES numerical simulations of mitigated disruption thermal quenches, caused by SPI in ITER have been performed. • Bremsstrahlung and recombination radiation contributes significantly to the core cooling upon injection of mixed pellet • The minimum Ne amount in two mixed pellets, which radiates >90 % of the core plasma energy has been found. TOKES simulations of mixed Ne/D 2 pellet injection through 1 equatorial launcher into an ITER discharge of 280 MJ thermal plasma energy was carried out. Two kinds of pellets were simulated: ‘large’, containing 1.1∙10 24 D 2 molecules and ‘small’ with 2.6∙10 23 D 2 molecules, both with various amount of Ne and correspondingly lower quantities of D to preserve the pellet size. It was found that the physics of the core energy radiation by the mixed pellet is different from that of the pure Ne pellet. In case of the mixed pellet injection, bremsstrahlung and recombination radiation make a significant contribution to the core cooling along with the Ne line radiation, the latter being the only mechanism for pure Ne pellet injection. The minimum Ne amount in the mixed pellet, which radiates more than 90 % of the core plasma energy (the threshold) has been found to be 1∙10 20 Ne atoms for the large pellet and 2∙10 21 atoms for the small one.

Turbulence Mechanisms of Enhanced Performance Stellarator Plasmas

We theoretically assess two mechanisms thought to be responsible for the enhanced performance observed in plasma discharges of the Wendelstein 7-X stellarator experiment fueled by pellet injection. The effects of the ambipolar radial electric field and the electron density peaking on the turbulent ion heat transport are separately evaluated using large-scale gyrokinetic simulations. The essential role of the stellarator magnetic geometry is demonstrated, by comparison with a tokamak.

Measurement of Radial Correlation Lengths of Electron Density Fluctuations in Heliotron J Using O-Mode Reflectometry

The radial correlation length Lr of electron density fluctuations has been measured using a two-channel Omode reflectometer system in the helical-axis fusion plasma experimental device Heliotron J. The experimental results show that Lr is around 1.4 n 0.7mm in three different magnetic field bumpiness configurations (i.e. low, medium and high) for the low-density electron cyclotron heating (ECH) discharges. In high-density neutral beam injection (NBI) discharges with HIGP and pellet injection, Lr is found to be around 1 n 0.2 mm.

Advanced ASDEX Upgrade pellet guiding system design.

Cryogenic pellet injection will be the prime candidate to fuel future fusion power plants. In order to harvest optimum fueling performance, it is essential to inject pellets from the magnetic high field side of the tokamak. The pellet launching system of the tokamak ASDEX Upgrade injects cryogenic hydrogen pellets with a speed of up to 1000 m/s from the magnetic high field side via curved guiding tubes. Pellets passing the guiding tube are sliding on a gas cushion, generated by the Leidenfrost effect. The actual track has a rectangular cross section and is composed of a series of ellipses in order to generate the required 270° looping type turn; the path length is 17 m. The last part of this track is marked by strong geometrical constraints from the vacuum vessel port. The previous design was composed of a sequence of three sections of ellipses too, tangentially constant but discontinuous with regard to the curvature. It had been in operation for almost 20 years. Its steps in the curvature are supposed to limit the system performance. A novel and advanced geometry concept, adopting a method well-known from civil engineering (e.g., for the railroad track design), has been applied to develop an improved design. It relies on clothoid shape sections keeping the track curvatures continuous and, thus, provides a smooth transition between all the elements. The new design presented improves the pellet launching system performance on ASDEX Upgrade and provides knowledge for an advanced design of pellet guiding tubes in future fusion devices.

Non-Newtonian and non-isothermal numerical modelling of a twin screw hydrogen extruder with slip imposed boundary condition on the screw surface

Abstract Fueling a plasma reactor by pellet injection is the most efficient method. Production of a solid hydrogen filament using a Twin screw extruder (TSE) is reported to be most stable among various extrusion techniques. The non-Newtonian behavior of solid hydrogen and non-linearities arising due to temperature dependent thermophysical properties in addition to rotation of two intermeshing screws make the flow modelling challenging. In the present study, a non-isothermal and non-Newtonian model has been developed taking into account the temperature and shear rate dependent Herschel Bulkley viscosity model. The ‘slip’ boundary condition has been imposed on the screw surface to account for the localized melting at the screw surface. The highly nonlinear viscous heating term in the energy equation is handled by using evolution scheme. It is important to accurately estimate the viscous dissipation rate which determines the cooling load required and input power. The viscous dissipation rate and developed torque predicted using the CFD model is in close agreement with the experimental available in the literature. The temperature rise due to viscous heating at the clearance gaps and the intermeshing region is found to be 0.6 K higher than that developed in the non-intermeshing region at 4 rpm screw speed for the given extruder dimensions.

Investigation of the dithering-like cycles induced by deuterium pellet injection during ELMy H-mode plasmas on EAST

By injecting cryogenic deuterium pellets into H-mode plasmas with edge localized modes (ELMs) from the low-field side (LFS) on EAST tokamak, the impact of these pellets on the edge plasma can be investigated. In some cases of pellet injection (PI), the intermediate phase (I-phase), i.e, the dithering-like cycles, have been observed. Experiments show that during one dithering-like cycle, three phases can be distinguished, which are closely related to the variations in the pedestal density profiles and density gradient after PI. Phase 1 is characterized by rapid particle loss and a density pedestal crash, phase 2 has a relatively low density gradient and enhanced edge density fluctuation and phase 3 corresponds to the relatively fast recovery of the density pedestal and reduced edge density fluctuation. Through an experimental analysis, it is shown that the plasma will enter the I-phase when the edge line-averaged density increases beyond a certain threshold value after PI. H-I-H transitions are more likely to be observed when the edge line-averaged density is greater than the core line-averaged density.

A reduced model for edge localized mode control by supersonic molecular beam injection and pellet injection

We develop a diffusive, bistable, tri-unstable cellular automata (CA) model to study the dynamics of H-mode pedestal with edge localized modes (ELMs) and their control by supersonic molecular beam injection (SMBI) and pellet injection (PI). It is shown that the new CA model can reproduce the key features of H-mode pedestals with various types of ELM, including Type-I ELM. SMBI and PI are modeled as additional grain injections into pedestal with varying degrees of injected materials and profiles. It is found that H-mode pedestal responds to SMBI differently depending on the baseline fueling. If the baseline fueling is large enough to allow Type-I ELM, SMBI enhances large transport avalanches caused by ballooning instabilities. These avalanches prevent the total pedestal current from reaching the boundary for peeling instability. On the other hand, if the baseline fueling is low to avoid Type-I ELM, SMBI enhances small scale avalanches, which prevent the pedestal from growing to profiles globally vulnerable to ballooning instabilities. These imply that SMBI can mitigate different types of ELM by converting them to more benign types. From CA modeling of pellet injection, it is shown that Type-I ELM can be triggered by pellet injection with sufficient strength and depth. Scanning the frequency of pellet injection, it is found that a maximum efficiency of pellet pacing is achieved when the injection frequency is approximately ten times the natural frequency of Type-I ELM.

A mechanism of ion temperature peaking by impurity pellet injection in a heliotron plasma

Experiments on the Large Helical Device with the injection of carbon pellets into discharges of low density have demonstrated a significant reduction of the ion heat conduction in the plasma core and an increase in the central ion temperature by a factor of up to 2. These results are interpreted in the framework of a transport model elaborated on the basis of those applied previously to explain the improvement in confinement by impurity seeding into the tokamak devices TEXTOR and JET. The calculations performed reproduce well the strong peaking of the ion temperature profile with increasing carbon density nZ and the consequent drop in the confinement as nZ exceeds a certain critical level. The importance of different elements in the model, such as braking of the main ion rotation by friction with impurity ones and the shape of the density profiles, are investigated. A qualitative assessment of the applicability under fusion reactor conditions, e.g. of much higher plasma density and heating power, is performed.

Numerical Investigations on Enhanced-Performance Superconducting Linear Acceleration System for Pellet Injection

Improvement of the achievable speed of the pellet container in the superconducting linear acceleration (SLA) system has been investigated numerically. To this end, a numerical code has been developed for analyzing the shielding current density in a high-temperature superconducting film. Moreover, another coil is placed on the outside of the conventional one. The results of computations show that the acceleration time of the SLA system is 2.1 s, and this time becomes shorter 5.7 s as compared to the only use of the conventional coil. In addition, it is found that the distance of electromagnetic rails is reduced to 1/3 or less compared with the numerical results of the previous study. However, we found that even this value is too long. We conclude that it is necessary to make the shape of the rail not straight but circular. Furthermore, our study also needs to be discussed in the circular shape of electromagnetic rails.

Neutronics related integration studies of EU-DEMO pellet injection system

• The problem of the integration of the pellet injection line system from neutronic and thermal point of view is presented. • The evolution of the design of the pellet injection system from inboard side of the machine is presented. • A neutronic preliminary assessment of an alternative solution for pellet injection system from the upper port of the machine has been studied. • All analyses have been done using MCNP-5 Monte Carlo code and ADVANTG hybrid has been used to implement specific variance reduction techniques. In the frame of the EU DEMO development program within the EUROfusion Consortium, the integration of the in-vessel components is crucial even at an early stage of the design process. The auxiliary, heating and fueling systems have to be integrated into the Breeding Blanket and, thus, they will undergo to a harsh nuclear environment during operation, and have a significant impact on the Tritium Breeding capability, loads and shielding performances. This work presents the neutronics and thermal analyses performed to support the integration studies of the pellet injection systems. This study is mainly devoted to optimize the design of a fueling line option consisting in a protrusion of the Vacuum Vessel (VV) in the inboard side supporting a guiding tube to drive the pellets along the ideal trajectory as close as possible to the First Wall. An alternative concept with a free-flight injector through the upper port, has also been studied evaluating the feasibility and the shielding needs. The three-dimensional neutronics simulations were carried-out with the MCNP5 Monte-Carlo code using a DEMO model with the integrated systems to calculate neutron fluxes, damage and nuclear heating distribution. Thermal analyses were carried-out using ABAQUS. Results of neutronics and thermal analyses on the fueling systems are presented and discussed.

Analytical formula for pellet fuel source density in toroidal plasma configurations based on an areal deposition model

An analytical formula for the fuel source density averaged over a magnetic flux surface following pellet injection is derived. During pellet motion the ablated and ionized fuel will be initially deposited nearby the pellet, prior to subsequent homogenization within the magnetic flux surfaces. In this paper, it is assumed that the deposition area projected normal to the magnetic field is on the order of the cross-sectional area of the ionized ablation column whose axis is aligned with the magnetic field. Previous pellet deposition codes (Houlberg et al 1979) employed a point source deposition approximation, which overpredicts the surface-averaged density, particularly when the obliquity between the pellet trajectory and the flux surface normal approaches 90 degrees. A key finding of this study is that as the obliquity approaches 90 degrees, grazing trajectory, the areal model predicts a well behaved , and removes the unphysical divergence stemming from the point model. To highlight the difference exhibited between areal and point models for near grazing trajectories, we considered vertical pellet injection into a CFETR-like tokamak with a negative triangularity plasma cross section using a pellet ablation/deposition module.

Interpretive MHD modeling of dispersive shell pellet injection for rapid shutdown in tokamaks

Dispersive shell pellet (DSP) injection is modeled with the extended-MHD code NIMROD for interpretive insight into the results of recent DIII-D DSP experiments and to explore the dynamics of an inside-out thermal quench for disruption mitigation in tokamaks. Simulations of the pre-thermal quench (TQ) phase indicate that the upper bound for the quantity of ablated carbon shell material that will not perturb the flux surfaces is in the ballpark of, but somewhat below the experimental quantity. Even below this quantity, sufficient electrons are added to the plasma by the shell material to produce significant dilution cooling before the TQ is triggered. Simulations carried through the end of the TQ have very large amplitude MHD fluctuations (δB/B > 10−2) at the time of the plasma current spike associated with current profile redistribution. After the plasma current spike, which is of comparable amplitude to that measured in DIII-D experiments, none of the runaway electron test-particles whose orbits are tracked throughout the simulation remain confined.

Numerical simulation of Li-pellet injection experiments for ELM-pacing in EAST

The injection of light impurity granules into fusion research discharges has shown a great potential for ELM pacing in future nuclear devices. The present work employed a model considering both pellet ablation and transport to systematically study for the first time the evolution of both impurity Li species and target plasma in the edge plasma during single and repeated Li pellet injections, with all initial information extracted from the EAST experiments. The evolutionary distributions of atomic and ionized Li species were presented and show good consistency with experimental CCD camera images. Simulation studies were carried out for single and repeated pellet injections at the outer mid-plane, and statistical uncertainty of ELM pacing rates triggering ELMs is discussed with regard to related injection conditions. It is found that the pressure locally caused by an ELM pacing pellet (with a diameter of 0.6 ∼ 0.7 mm) was 1.8–2.4 times its unperturbed magnitude in the EAST shots of interest. Accumulation from repeated pellet injections results in the rise of pedestal height at a high injection frequency and this effect becomes stronger with increasing pellet size. Pressure changes are also evident in a scenario of multiple small pellets per stroke, which could lead to a nonnegligible, probabilistic ELM triggering rate observed experimentally. The effects of other possible factors are also discussed. With the assumption that the ELM trigger rate of a Li pellet of 0.7 mm diameter is 100%, the predicted trigger rates are in good agreement with those observed in experiments. This work can help understand the statistically probabilistic behavior of ELM triggering efficiency observed in previous studies.

Evaluation of fuelling requirements for core density and divertor heat load control in non-stationary phases of the ITER DT 15 MA baseline scenario

To ensure optimal plasma performance at high Qfus for the baseline scenario foreseen for ITER, the fuelling requirements, in particular for non-stationary phases, need to be assessed by means of integrated modelling to address the special additional challenges facing plasma fuelling on ITER. The fuelling scheme needs to be adjusted to ensure robust divertor heat load control, avoiding complete detachment while still maintaining low divertor temperatures and heat fluxes to minimise W sputtering and contamination of the plasma by impurities. At the same time, the core density needs to be controlled to fulfil requirements for: the application of neutral beam heating with acceptable shine-through losses; a robust transition from L-mode to stationary fusion burn; the maximisation of the fusion yield; and a fast reduction in core particle content in the termination phase. Coupled core-edge-SOL transport calculations have been performed, simulating for the first time the entire ITER plasma evolution from just after the X-point formation until the late termination phase. These calculations are being exploited to find the most effective ways of fuelling and heating DT plasmas without exceeding ITER operational limits (e.g. divertor power density). The most efficient ways to fuel ITER with gas and/or pellet injection have been investigated self-consistently with the integrated core + edge + SOL transport suite of codes, JINTRAC, developed at JET (Romanelli et al 2014 Plasma Fusion Res. 9 3403023). Our modelling is exploited to study schemes for gas and pellet fuelling for main ion SOL and core density control, respectively, and for impurity seeding by Ne for the control of SOL radiation, that allow ITER to approach Qfus10, with plasma evolution successfully controlled to respect major operational limits through all transients from the early ramp-up until the late ramp down phase.

High-performance plasmas after pellet injections in Wendelstein 7-X

A significant improvement of plasma parameters in the optimized stellarator W7-X is found after injections of frozen hydrogen pellets. The ion temperature in the post-pellet phase exceeds 3 keV with 5 MW of electron heating and the global energy confinement time surpasses the empirical ISS04-scaling. The plasma parameters realized in such experiments are significantly above those in comparable gas-fuelled discharges. In this paper, we present details of these pellet experiments and discuss the main plasma properties during the enhanced confinement phases. Local power balance is applied to show that the heat transport in post-pellet phases is close to the neoclassical level for the ion channel and is about a factor of two above that level for the combined losses. In comparable gas-fuelled discharges, the heat transport is by about ten times larger than the neoclassical level, and thus is largely anomalous. It is further observed that the improvement in the transport is related to the peaked density profiles that lead to a stabilization of the ion-scale turbulence.