Choose timezone
Your profile timezone:
Powered by Indico
v3.3.4
Since 18 of December 2019 conferences.iaea.org uses Nucleus credentials. Visit our help pages for information on how to Register and Sign-in using Nucleus.
KEY DEADLINES
30 June 2022 Deadline for submission of abstracts through IAEA-INDICO for regular contributions
30 June 2022 Deadline for submission of Participation Form (Form A), and Grant Application Form (Form C) (if applicable) through the official channels
29 July 2022 Notification of acceptance of abstracts and of assigned awards
It is expected that first generation fusion power plants will use a mixture of two heavy isotopes of hydrogen — deuterium and tritium (DT) — as fuel, which then fuse to produce helium and neutrons. Inside the reactor, additional tritium will be ‘bred’ or created from the reaction of the neutrons with a lithium blanket covering the inner side of the reactor vessel. However, several uncertainties remain regarding the physics and technology of the DT fuel cycle in ITER and demonstration fusion power plants
Objectives
The event aims to review and discuss – in an integrated manner – plasma physics and technology aspects of the tritium fuel cycle in magnetic fusion reactors, from ITER to demonstration fusion power plants.
Target Audience
The event aims to bring together technical representatives from a wide range of fusion specialties including core plasma physics, plasma material interactions, edge plasma (scrape-off layer) physics, materials science, plasma component design, vacuum systems, tritium behaviour, and tritium processing.
Efficient plasma fuelling, impurity exhaust and power load control are essential for the successful operation of fusion reactors and have a direct impact on the fuel cycle and the achievable tritium burn-up fraction.
The physics of plasma fuelling in fusion devices and the processes determining the balance of particles in the reactor chamber are described, with emphasis on the core and edge plasma conditions required to provide the impurity exhaust and power load control necessary to sustain high Q operation. These processes include the ionization and transport of neutrals into the edge plasma (produced by recycling of ions lost from the plasma and from gas injection) as well as of pellets injected to fuel the confined plasma. The impact of fusion reactor edge plasma conditions on the effectiveness of these fuelling schemes will be described, together with the open R&D issues in this area.
Once fuel neutrals (either from recycling/gas injection or pellet injection) are ionized in the confined plasma, they have to reach the core plasma region where fusion reactions take place. This is dominated by turbulent transport. The impact of the expected reactor scale plasma density and temperature profiles and the impurity level and mix composition on fuel transport will be discussed, together with open R&D issues.
Based on the understanding of the fuelling and plasma transport physics processes highlighted here, approaches to optimize tritium recirculation and burn-up will be discussed.
The regulation of the amount of fusion power produced by future reactors will require precise control over the plasma density and temperature. Therefore, the control of the core-plasma kinetic state, usually referred to as burn control, arises as one of the most fundamental problems in nuclear fusion and will be critical to the success of burning-plasma devices like ITER. Due to the nonlinear coupled dynamics of the plasma, feedback control of the burn condition will be necessary to avoid undesirable transient performance and to respond to changes in plasma confinement, impurity content, or operation conditions, which could significantly alter the plasma burn. A well-design burn controller should be able not only to achieve tight regulation around a desired operating point by rejecting perturbations in temperature and density but also to drive the plasma from one operating point to another during the burning plasma mode (e.g., different Q or fusion power). Moreover, the controller should be capable of accessing to and exiting from the burning plasma mode. The isotopic fuel mix in the plasma is a critical reactor parameter as it has a major influence on the fusion power produced. Differences in deuterium (D) and tritium (T) transport and fueling efficiency, as well as perturbation introduced by other sources of particles not under the burn controller such as NBIs, may lead to a non-optimal fuel mix in the core even with an optimal 50:50 DT injection. Additionally, depending on the operating scenario, it may be desirable or even necessary to operate at a lower tritium fraction or vary the tritium fraction during operation. The regulation of the tritium ratio is possible thanks to a method of fueling referred to as isotopic tailoring, in which the relative mix of deuterium and tritium injected by the fueling system is varied in real-time. The pellet injection system for ITER will include two separate injectors—one with pellets made of primarily deuterium and the other with pellets made primarily of tritium. A gas injection system will be used to supply deuterium at the edge of the plasma. Together, these systems will allow for fuel mix regulation. The role of isotopic tailoring as an actuator for burn control will be discussed within an overall control scheme that may also include auxiliary power, fueling, impurity injection, and magnetic coils. The impact of recycling on the ability of changing the fuel mix via isotopic tailoring will also be discussed. Moreover, the accessibility of steady-state, Q=10, operation points in ITER will be analyzed for different particle recycling and confinement conditions as a function of the tritium concentration in the fueling lines, and methods will be proposed to robustify the control of the burn condition against unmeasurable variations over time (biases and drifts) in the tritium concentration of the fueling lines.
Benign power exhaust in a tokamak relies on the injection of radiating impurities (plasma enhancement gases), to be controlled for the achievement of the desired power crossing the separatrix and sufficient divertor radiation. At least partial divertor detachment is required, reducing the heat flux impinging the divertor target at the separatrix below about 5 MW/m$^2$. Candidate gases for divertor radiation are N, Ne and Ar, while potential core radiators are Ar, Kr and Xe. A high throughput of D+T (gas puffing, pellet injection) is generally required to obtain a high compression of impurities in the divertor, avoiding too strong core fuel dilution. This applies as well to the He atoms generated by the fusion reaction.
The talk will explain the general relationships between fueling, radiative losses, divertor compression and pumping, using integrated scenarios in the ASDEX Upgrade tokamak as examples which combine core and divertor radiation and no-ELM conditions. The measured impurity and fuel distributions are described by a simple particle balance model, which provides the size dependent extrapolation to reactor conditions and defines the open physics parameters. Here, the impurity divertor enrichment is the most important quantity, describing the ratio of impurity and fuel divertor compression. Much longer time scales are expected for reactor conditions compared to those observed ASDEX Upgrade, due to much larger relevant volumes but a similar pumping speed. The fuel particle throughput, which is connected to the separatrix density, appears as important optimization quantity: while a cleaner core plasma and easier divertor detachment are obtained at high gas throughput, a reduction of core energy confinement has generally been observed.
Deuterium (D)-Tritium (T) plasmas are considered the most promising hydrogen isotope combination for the generation of fusion energy in tokamaks. However, in contrast to electron particle transport, ion transport in mixed plasmas, notably in the presence of T, is less understood.
Differences in the electron density behavior clearly indicated in the first DT campaigns in TFTR and JET that electron particle transport was significantly different in DD vs DT plasmas. However, the understanding of the ion particle transport is more elusive as the direct measurement of ion density profiles is challenging. From the theoretical point of view, several works with gyrokinetic codes pointed out an asymmetry between T and D particle transport in particular plasma conditions, which would lead to lower T particle fluxes than D for the same ion density gradients [1,2]. Integrated modelling in preparation of DT plasmas at JET showed that higher T density peaking than D could be expected in such conditions, resulting in a higher electron density in DT compared to DD [3]. Furthermore, recent analyses have shown that in multi ion plasmas, ion particle transport can behave quite differently to electron transport leading to an insensitivity of core ion densities to particle sources in different isotopes [4,5].
Recent campaigns at JET have given the opportunity to further explore this issue. In the presence of T, higher electron density is obtained, due to mainly an increase on the pedestal density. Possible increase of core density peaking in DT compared to DD is under analysis. The DT ratio has been well controlled by external gas puffing or pellet fueling. No large discrepancies between the DT ratio at the divertor and in the core have been detected. However, some discrepancies between the neutron rate calculated in TRANSP and the experimental one might indicate some differences on the D and T density profiles. Such discrepancies are particularly strong, up to 40%, at low NBI power and low neutron generation. Finally, T plasmas heated with high NBI power in D do not seem to be significantly polluted in the plasma core, confirming the resilience of multi ion plasmas to core particle source.
[1] C Estrada-Mila et al., Phys. Plasmas 12, 022305 (2005) [2] J. Garcia et al., Physics of Plasmas 25, 055902 (2018) [3] F.J. Casson et al 2020 Nucl. Fusion 60 066029 [4] M. Maslov et al 2018 Nucl. Fusion 58 076022 [5] C. Bourdelle et al 2018 Nucl. Fusion 58 076028
In a fusion reactor plasma, impurities are present due to multiple sources. In the center, He ash is produced by fusion reactions, at the edge plasma facing components can release impurity atoms and impurities are actively seeded to reach a tolerable heat exhaust. The consequent impurity density profiles will be the result of a combination of the strength of the sources and of the transport, determining the concentration levels and the profile shapes. As a consequence of the strong increase with charge of collisional transport, the transport of highly charged impurities is generally determined by a combination of collisional and turbulent transport, where the collisional (neoclassical) transport is rarely negligible. Light impurities, like He, are usually more strongly governed by turbulent transport. The relative roles of collisional and turbulent transport also depend on the plasma region. Collisional transport can prevail close to the center, where turbulent transport is weak, and in regions of reduced turbulence, like in the transport barrier of the H-mode pedestal. While turbulent transport does not produce strong mechanisms of impurity accumulation, collisional transport can produce strong impurity convection towards the plasma center, critically depending on the relative strength of the main ion density and temperature gradients and the collisionality regime. Plasma toroidal rotation also strongly increases neoclassical transport and can affect the direction of the convection. Turbulent transport can counteract the unfavorable neoclassical convection by increasing the diffusion. Decades of tokamak operation allowed the development of methods to avoid impurity accumulation by controlling both the strength of the sources and the transport in the edge region, as well as preventing strong neoclassical accumulation to occur in the center, where the profiles of the radiated power density connect edge and core effects. Progress in theory is also allowing these operational methods to be increasingly understood and their effects to be predicted.
Plasma-material interaction (PMI) imposes a number of challenges on the operation of a next step fusion device or reactor associated with the lifetime of components, the sustainability of the tritium cycle, and ultimately with safety aspects. The underlying critical processes under steady-state plasma operation can be splitted into two categories: (i) erosion, transport, deposition, and dust formation described in general in the term material migration and (ii) co-deposition, implantation, diffusion, and permeation labelled in general as fuel retention. The role and strength of the individual process depend primarily on the choice and energy of plasma or projectile species (D ,T and seeding species Ar, Ne, etc.) and plasma-facing or target materials (low-Z species like C, Be, Li etc.- and high-Z species like W. Mo, steel, etc. ). DT fusion neutrons can induce additional damage to the materials in the main chamber above a damage threshold resulting in enhancement of e.g. fuel retention above roughly half a dpa, which is in the order of the end of lifetime damage in the only full burning device under construction - ITER. In a reactor of DEMO size with much higher neutron dose, neutron induced effects will play the dominant role in the retention process.
Here, we present an overview of the different processes, how they interact with each other, and define potential limits. Predominantly experimental results from the JET and ASDEX Upgrade tokamak will be use to describe the differences between plasma operation in low-Z and high-Z first wall in deuterium. A massive reduction of the material migration and fuel retention by a factor 10-20 have been identified in both device when transferring from carbon-based materials towards a metallic material device, but adaptation of the operational space was required with tungsten. Comparison between different fuel isotopes will be addressed in particular with the aid of JET, which operated recently in H, D, T and DT, and allows to establish and extrapolate previous non-T operation towards reactor like conditions. However, none of the present-day tokamak operation can directly contribute to the assessment of the role of neutrons on the PMI processes. Here, we rely on accompanied research in dedicated laboratories, which can mimic the impact of neutrons by heavy ion, proton or fission neutron impact. Modelling is used to transfer the physics towards fission neutrons in the lack of a facility to mimic material damage by 14MeV neutrons.
The overall physics understanding regarding PMI will be brought into perspective of a DEMO-like device. The proposed modelling tools to address the critical issues first wall erosion, dust production, and tritium retention will be briefly introduced and linked to recent revised assessments done for the PFPO and FPO phases in ITER. The preliminary status of corresponding plasma edge (SOLPS-ITER) and plasma-material interaction modelling (ERO2.0) for the European DEMO - without consideration of neutron impact - will be presented.
Tritium (T) retention in plasma facing components (PFCs) subjected to burning plasma-material interactions (BPMI), defined here as simultaneous plasma exposure and 14 MeV neutron irradiation at reactor-relevant temperatures, will materially impact the in-vessel T inventory, achievable tritium breeding ratio (TBR), and performance limits of PFCs in fusion pilot plants (FPPs). Validated model predictions of these issues in FPP PFCs are needed to support FPP designs but require a deep understanding of PFC material evolution under BPMI conditions.
Experiments on linear plasma devices are contributing to the development of this understanding in several ways. First, these experiments show that PFCs exposed to mixed D-He plasmas undergo profound surface and near-surface morphology changes, including formation of He-nanobubble layers at the surface, development of arrays of micron-scaled cone-shaped features and, at high surface temperature, spontaneous growth of a dense interwoven nano tendrils that emerge out of the underlying substrate. Experiments then show that these changes have a profound impact on fuel retention within the PFCs, and that these changes lead to macroscopic changes in transport of eroded surface material within the surrounding scrape-off layer plasma; thus, any model of redeposition and material migration must account for these microscopic morphology changes. Second, ion-beam based studies show that the displacement damage from energetic particles leads to defects that can greatly enhance fuel retention. When exposed at reactor-relevant temperatures, interstitial-vacancy recombination in W-based materials occurs quickly enough to largely eliminate this elevated retention. However, recent sequential ion-beam/linear plasma device PMI experiments clearly indicate that trapping of plasma-implanted hydrogen isotopes within these defects can stabilize vacancy-interstitial recombination, and thus may largely stop this beneficial annealing. Finally, recent experiments show a clear degradation of material thermomechanical properties from plasma exposure and from displacement damage, and thus PFC operational limits in an FPP will likely be profoundly impacted by BPMI effects.
After presenting these existing results, we close with thoughts on future linear device and confinement device BPMI experiments that can help develop the validated predictive multi-scale PFC models needed to guide the design of any magnetic confinement based FPP device.
Acknowledgement: Work supported by the U.S. Department of Energy under grant DE-FG02-07ER54912 and cooperative agreement DE-SC0022528 .
A critical challenge for the long-term operation of ITER and the future U.S. fusion pilot plants will be the development of plasma-facing components (PFCs) that demonstrate erosion resistance to intense heat and neutral/ion particle fluxes under the extreme fusion nuclear environment while minimizing in-vessel inventories and ex-vessel permeation of tritium.
INL leverages a series of U.S.-Japan collaborations (TITAN, PHENIX, FRONTIER) to irradiate tungsten and tungsten alloy material in High Flux Isotope Reactor at ORNL and investigate irradiation response on fuel behavior in irradiated tungsten using INL’s linear plasma device, Tritium Plasma Experiment.
This talk describes the challenge in modeling tritium behavior in neutron-irradiated PFCs, the U.S.-Japan plans for neutron-irradiation and post-irradiation examination, and the recent findings on fuel retention in neutron-irradiated and also ion irradiated tungsten.
A self-sufficient fuel cycle is a significant contributor to enable commercial fusion energy. It also puts additional requirements on existing fuel cycle concepts. Driven by the need to reduce the tritium inventory in the systems to an absolute minimum, the work package TFV (Tritium – Matter Injection – Vacuum) of the European Fusion Programme has developed a three-loop fuel cycle architecture. This requires the continual recirculation of gases in loops without storage, avoiding hold-ups of tritium in each process stage by giving preference to continuous over batch technologies, and immediate use of tritium released from tritium breeding blankets. In order to achieve this goal, a number of novel concepts and technologies had to be found and their principle feasibility to be shown.
Ths talk will start from a functional analysis of the fuel cycle and introduce the results of a technology survey and ranking. The main interfaces to the plasma (breeding blanket, matter injection and particle exhaust) will be described. Based on this, the fuel cycle architecture will be delineated and required operational windows of the sub-systems defined, based on suitable figures of merit. To validate this, various R&D lines were established, the main results of which are reported, together with the remaining key technology developments to be addressed in the next years.
A future fusion reactor is anticipated to mainly utilize pellet injection for particle fuelling. Pellets are mm-sized bodies formed from solid hydrogen fuel. Undoubtedly, delivering an adequate fuel amount with an isotope mixture adjusted to establish the optimum deuterium:tritium composition expected in the vicinity of D:T = 1:1 to the plasma core has to lay the foundation for any pellet actuator concept. In a tokamak, this calls for high-speed injection of pellets from the torus inboard side via guide tubes. Hence, any reactor is in need of a highly capable pellet injection system located at the interface of plasma physics and the fuel cycle technology. Beyond this primary fuelling purpose, hydrogenic pellets have proven their potential to serve for additional actuation tasks.
In ITER, gas fuelling efficiency will be strongly limited by the poor penetration of the neutrals into the core plasma. To access and maintain a core density level adequate for a high fusion gain, controllable steady-state pellet delivery is needed. Inboard launch of pellets containing deuterium and tritium with high reliability for up to 1 hour plasma duration is foreseen. In addition, the ITER pellet system is expected to serve for ELM pacing – the controlled triggering and mitigation of these potentially plasma-facing component damaging edge instabilities that expel significant amounts of plasma energy. This yet unprecedented challenge for a pellet system is expected to be covered by a launcher under development at ORNL. The launcher, based on well proven technology for high mass throughput pellet production and acceleration in a gas gun, will be set up for delivery of pellets with adjustable size and composition via different injection guide tube routes optionally to be chosen as needed for the fuelling and ELM pacing tasks.
For the follow-up project DEMO, requirements for the pellet system are even more challenging. In order to harvest an ample amount of fusion power, here steady operation at core densities significantly above the Greenwald density is envisaged. Operation in this high-density regime is possible virtually only by pellets. Since burn control in DEMO becomes very sensitive to variations in the core density, the demands for reliability and controllability performance are significant. To foster the development of a pellet actuator capable to cover the multifaceted control requirements of a burning plasma, investigations are under way at ASDEX Upgrade (AUG). Efforts reported span a wide range of pellet technology control techniques utilized in plasma physics experiments. For example, fuelling experiments were performed mimicking D/T by H/D pellets demonstrating safe and reversible isotopic control with high core density operation. However, a significant disagreement with scaling predictions of a continuing increase of energy confinement with density was found. Instead, above a distinct level, the energy content can be kept at best constant. Thus far, experimental investigations focussed on the ELMy H-Mode scenario. Recently, encouraging results achieved with naturally ELM-free regimes rendered the possibility for a more benign DEMO scenario. Consequently, a critical review of pellet actuation for its compatibility with these DEMO relevant scenarios is still required.
The divertor system of a fusion device is always a compromise which has to meet power exhaust, particle exhaust and neutron shielding requirements at the same time. The design space of the tokamak particle exhaust function results from a number of requirements, such as geometrical parameters (for instance the divertor cassette configuration, and the position of the pumping port relative to the divertor), the effective pumping speed that can be provided, the intra- and inter-cassette gaps which define the recycle flow pattern of the divertor cassettes as well as from the presence of the dome which is mainly defined by neutron shielding. Hence, a feasible divertor design has to properly consider particle transport physics. Furthermore, since the neutral density in the private flux region (PFR) is expected to be high enough to justify viscous flow conditions, the corresponding gas collisionality increases and therefore a nonlinear (i.e. collisional) neutral particle transport treatment is required.
In that context, a numerical tool called DIVGAS (Divertor Gas Simulator) has been developed at Karlsruhe Institute of Technology (KIT). The DIVGAS code is based on the Direct Simulation Monte Carlo (DSMC) method. The aim of this code is to investigate and reliably describe the flow conditions in the particle exhaust of a fusion device. That said, DIVGAS takes into account all the physics and engineering aspects of plasma‐wall interactions in the divertor, which influence the generation of neutral particles at the targets and consequently the overall flow behavior of the particle exhaust, including the attached vacuum system. For validation purposes, the DIVGAS code has been implemented to model the neutral gas flow in the JET sub-divertor. Moreover, DIVGAS has been applied for simulating the particle exhaust of ITER, JT60SA, AUG, EU-DEMO, DTT and recently of the stellarator W7-X.
This contribution will exemplify the main workflow which uses DIVGAS process to provide a self-consistent coupling between the sub-divertor volume and the vacuum systems. DIVGAS requires the total neutral flux (i.e fuel gas and impurities) in the PFR as input boundary condition. This information is usually provided by an edge plasma code. Additionally, the actual 2D/3D divertor geometry is introduced. The outcome of the simulation is given by the total pumped throughput, the recycle flow from the divertor to the core plasma, and the distribution of the neutral pressures in the whole sub-divertor area, which directly points to the required total pumping speed, distributed among a certain number of pumps located in the divertor pumping ports.
Based on the aforementioned workflow, this talk aims to highlight the impact of the main design drivers, illustrated by corresponding results, for various machines and divertor configurations (i.e. Single-Null, X, Super-X and liquid metal divertors) on the particle exhaust. In all cases, important design directions for achieving high pumping efficiency will be presented.
Present and planned fusion machines rely heavily on the use of neutral beam injectors, to provide plasma heating and current drive. In the case of large experiment, like ITER and beyond, the atoms injected in the plasma require a high energy (>500 keV) to penetrate the dense and large plasma and deliver the power at plasma center. This calls for the use of negative ions as precursors of the atomic beam. Negative ions are typically created in a cold plasma ion source, by a complex interplay of physicochemical processes requiring a careful control of the main parameters, including the source pressure and gas purity.
This contribution summarizes the basic concept of the neutral beam injection systems, focusing on the needs for high purity of the gas that in turns sets a demanding requirement to the fueling system. The reasons behind the definition of the specific values in the case of ITER NBIs are discussed and the consequences of a deviation from prescribed values are addressed.
The views and opinions expressed herein do not necessarily reflect those of the ITER Organization
The deuterium/tritium fuel cycle for fusion reactors is linked to how the reactor is designed and operated. Sometimes these links are obvious such as the choice of seeding gases and burn fraction, though others are subtle or not obvious. This paper presents an introduction to the key considerations stemming from physics decisions that impact the design and operation of the fuel cycle.
Tritium permeation from Breeding Blanket (BB) towards Primary Heat Transfer System (PHTS) constitutes a relevant issue for operation of DEMO machine. As a matter of fact, once permeated into PHTS, tritium can migrate to working areas and environment via permeation and leaks. In order to control radioactive release, tritium concentration within primary coolant must be kept below fixed limits. To do this, two strategies are under assessment: the use of anti-permeation barriers and/or the adoption of the Coolant Purification System (CPS).
CPS aims at recovering tritium permeated into PHTS, keeping its concentration below design limits. Preliminary sizing has been recently carried out for both helium and water CPS technologies, belonging to ancillary systems of Helium-Cooled Pebble Bed (HCPB) and Water-Cooled Lithium Lead (WCLL) BBs, respectively. CPSs dimensioning was performed to fulfill certain fixed conditions, mainly in terms of thermodynamic and tritium concentration, in primary coolant. However, during DEMO operation, CPS is called to work under an evolving tritium inventory within PHTS. In fact, tritium concentration in primary coolant will have an evolution with time according to the tritium permeation rate, which is linked to the plasma operation, and to the CPS functionality beyond its design point.
The present work deals with the assessment of tritium inventory and CPS operation over the alternation between plasma operation and scheduled maintenance periods, focusing on DEMO first operation phase. Starting from the most recent outcomes of tritium permeation analysis, the evolution of tritium concentration within PHTS is predicted for both helium and water cases, considering CPS efficiency under dynamic conditions. For the analysis, the size of main technologies of both helium and water CPS is kept fixed, according to the preliminary dimensioning. The objective is to predict a reasonable trend of the increase of tritium inventory into coolant over the whole DEMO first operation phase, according to the allowable performances of the CPS under different conditions.
DTT (Divertor Tokamak Test Facility) is a new facility, currently under build, in which various scaled experiments for testing different magnetic configurations and alternative solutions for the power exhaust system of DEMO will be performed. Although the divertor system is not finalized yet, the machine and port geometry set limitations on the divertor pumping system operational space. In the present work, an in-depth numerical study of neutral gas dynamics in the divertor region is performed based on the Direct Simulation Monte Carlo (DSMC) method by applying the Divertor Gas Simulator (DIVGAS) code, which over the last years has been proved as an efficient numerical tool for the simulation of the particle exhaust of fusion reactors. The information about the neutral particles imposed as boundary conditions on the DIVGAS simulations has been extracted by corresponding plasma simulations (B2-EIRENE). The scope of the present study is twofold. Firstly, to derive information on the particle exhaust fluxes and neutral pressures in the vicinity of the actual pumping opening based on the outcome of the plasma modelling of edge plasma of the DTT at full power in detached conditions, and secondly, to present a self-consistent approach that allows for a coupling between DIVGAS and the plasma code based on a physics basis satisfying the overall particle flux balance. Moreover, the albedo values, which up to now were treated as a free handshake parameter in the plasma modelling, are provided based on real physics strictly linked to the neutral gas dynamics of the sub-divertor area. The presented methodology is general and can be applied for the self-consistent evaluation of the divertor pumping efficiency in any fusion device. Overall, this study demonstrates how to successfully deal with one of the major interfaces between the plasma and the fuel cycle.
Future deuterium-tritium fueled fusion power plants must breed tritium and sustain a burning plasma using a semi-closed loop fuel cycle. The DT fusion fuel cycle is an important aspect of any fusion energy configuration whose purpose is to provide fuel to the plasma, pump and separate plasma exhaust products, and recover fuel from breeding and plasma exhaust products. The current method of separation requires a large building for palladium membranes, cryogenic distillation columns, and other separation equipment. The hydrogen isotope separation method utilized by ITER is time consuming which results in a large tritium inventory in the tritium plant. A 2 GW fusion power plant utilizing the same technology will result in a tritium inventory of ~4 kg, four times the inventory required for ITER. Scaling a fusion power reactor tritium plant using ITER technology represents an enormous hurdle in tritium inventory required, and solutions to reduce the size and cost of such a system are essential. The concept of directly recirculating the exhaust gas to make fuel pellets was proposed in the 1990s and later termed Direct Internal Recycling (DIR). In this approach the residual fusion fuel in the plasma exhaust stream is separated locally and diverted directly to the fueling systems, bypassing the isotope separation and other processing equipment, and therefore significantly reducing the required size of the tritium plant and resulting plant tritium inventory.
A concept for DIR uses a series of cryogenic pumps to separate the impurities from the machine exhaust gas by utilizing the different triple point temperatures of exhaust constituents. In this concept, the plasma exhaust is initially passed through an impurity trap operating at ~25-30 K to desublimate impurities such as hydrocarbons, argon, oxygen, and nitrogen. The resulting process stream will consist of deuterium, tritium, helium, and neon (if present). The process stream is then pumped by a continuous cryopump known as a “snail pump”. This pump is a steady state continuous cryopump that desubliminates all remaining exhaust gas constituents while allowing helium, a byproduct of the D-T fusion reaction, to pass through. The helium is pumped to the tritium plant for processing while the desublimated material is continuously scraped off, heated up, and diverted to a neon separator (if needed) before transport to the fueling system. This neon separator cryopump will operate in a carefully controlled regime to preferentially pump the neon while the deuterium and tritium pass through. The resulting DT exhaust is then sent to a pump/compressor system and used to supply DT gas directly to the pellet fueling system. The ratio of DT in the resulting fuel stream will be monitored to allow for makeup material to be added to the fuel stream to maintain the desired fuel ratio. This paper will outline the details of this proposed concept and the steps required for the development of this hardware.
The fuel cycle of the European DEMO reactor comprises three loops, where the first two – the Direct Internal Recycling (DIRL) and the Inner Tritium Plant (INTL) Loop – are directly coupled to the reactor. These loops include components that act as actuators on the plasma, as vacuum pumps, pellet injectors or gas injection valves that can and must be controlled on a given timescale.
This paper presents the foreseen actuators in DEMO, its location and the timescales on which control tasks can be executed with the available technology. Therefore, the actuators are listed and achievable timescales will be estimated or calculated. These values will be compared with requirements given by the plasma control systems.
This comparison will lead to the identification of gaps and challenges in the development of actuators or the actuator architecture for a DEMO machine.
The minimization of the Fuel Cycle inventory in a pulsed DEMO-like power reactor can be provided in the Direct Internal Recycling (DIR) concept by adding an additional short-cut between the pumped torus exhaust gas and the fuelling systems [1]. The Fuel Cycle which includes plasma fuelling and exhaust, as well as several exhaust processing and isotope separation processes, is one of the key elements on which the successful operation of a Fusion Power Plant (FPP) will depend.
For steady-state FPP operation it is necessary to remove the reaction products from the tokamak chamber and to compensate for burned-out components of the working mixture of deuterium and tritium. The consumption of fresh fuel is determined mainly by the reactor power and the amount of gas pumped off depends on various factors like reaction-product enrichment or purification conditions. Simple balance considerations can provide estimates and relations to optimize the fuel cycle loads.
In this work the limiting fuel burnup fraction is estimated as a function of the fraction of hydrogen isotopes and helium in the total plasma pressure for different values of the helium-enrichment factor in the mixture pumped off. It is also discussed the required purity of incoming fuel with respect to helium as a function of the allowable concentration of protium in the working chamber for different values of fuel burnup fraction. The required fuel purity with respect to protium, which is virtually independent of the burnup fraction, and the relative helium concentration in the working chamber are estimated.
Calculations show that for DEMO operation a relatively deep purification of the fuel mixture to remove helium is required, while as far as purification from protium is concerned it is sufficient to remove only a small portion of it.
[1] Ch. Day and T. Giegerich, The Direct Internal Recycling concept to simplify the fuel cycle of a fusion power plant, Fusion Engineering and Design 88 (2013) 616-620.
Since the beginning of the nuclear era, Canada has been a leader in the area of deuterium and tritium technologies. CANDU nuclear power stations use deuterium oxide to enable a chain reaction of fission in natural uranium, thereby requiring the development of deuterium oxide production and management. Neutron capture by deuterium leads to formation of tritiated deuterium oxide in CANDU reactors. This, in turn, has driven the development of technologies and best work practices for tritium capture, extraction, and handling. The Canadian nuclear industry is extending those established tritium technologies and experience into nuclear fusion applications by working with fusion organizations to develop solutions at all technology readiness levels.
Mature technologies like tritium capture, water detritiation, tritium handling, getter beds, and tritium monitors and measurements can be used in a DT fuel cycle with no or minimal modifications. Supporting these are fully-developed process simulations, design experience, assets and operating experience that can be immediately used for D-T fuel cycle design. Currently, CNL and other Canadian firms are directly providing these technologies to national and international fusion projects or working with them on how to deploy these technologies themselves.
For areas with low technology readiness levels, Canada is leveraging its assets to help develop the new technologies required by fusion projects. For example, CNL’s tritium facility is licensed to handle up to two million Curies of gaseous tritium (200 grams) and 100’s of Curies/kg of aqueous tritium in the lab. This ability means that CNL’s tritium facility is an excellent place to perform highly tritiated tests, such as tritium permeation through materials for fusion energy. CNL is planning molten metal test loops for tritium breeding and extraction technology tests and demonstrations, as well as developing a cost competitive and compact Thermal Cycling Adsorption Process (TCAP) unit, and a variety of installations of Combined Electrolysis and Catalytic Exchange (CECE) systems and equipment.
This presentation will detail some of the above activities in Canada.
Recently, as many countries are developing their demonstration fusion plant, Korea has also begun developing the Korean-style demonstration fusion plant. For the fuel cycle design, the handling of a large amount of tritium is an essential problem to be solved. Several activities related to process modeling and simulation are in progress for process design optimization and tritium inventory analysis of sub-systems in the fusion fuel cycle.
First, pure component physical properties of liquid and gaseous hydrogen isotopes are investigated to be applied to the process simulation. They include vapor pressure, density, heat capacity, the heat of formation, and virial coefficients. Collected data are used to obtain an appropriate equation of state and property calculation methods. Peng-Robinson equation with twu-alpha function shows accurate results for fugacity, density, and enthalpy of hydrogen and deuterium compared with the NIST REFPROP database. The properties of other isotopes are estimated with the same EOS and calculation methods. Second, experiments for a permeator for hydrogen separation have been conducted to assess the developed palladium membrane model. Single and double-stage palladium membranes are tested under several feed conditions. The one-dimensional palladium membrane is modeled using Aspen Custom Modeler. The simulations and experiment results show similar recovery rates at 3.5E-5 kmol/(m·hr·bar0.5) hydrogen permeability. Third, an optimization framework for distillation configuration is developed by establishing superstructures for the distillation system using Python. The goal is to minimize tritium inventory by adjusting the positions of the feed stream, positions of catalytic reactors, and the number of column stages. Bayesian optimization is applied to solve the problem. As a result, an optimized design that reduces tritium hold-up compared with the existing distillation configuration is obtained.
In this presentation, the progress and detailed information for the above activities will be summarized. Through above studies, it is expected to contribute to improving and developing the Korean demonstration fusion fuel cycle.
Significant production of radioactive metal dust, mainly due to erosion of plasma facing components, is expected to be present in the vacuum vessel of DEMO while it is operating. A relevant tritium content is expected to be present in this dust. Therefore, the removal of tritium and the management of this radioactive dust for safety reasons and for eventual remanufacturing or appropriate disposal are key issues.
Today several techniques are available for producing pure tungsten components, or mixtures of tungsten and other metallic and non-metallic materials. Basically, the tungsten powder metallurgy follows the sequence Mixing, Pressing and Sintering. These techniques could be used to the re-fabrication or the safely appropriate disposal of tungsten eroded dust.
The present work describes the study to select and optimise techniques for the detritiation of the DEMO W dust, able to produce massive, mechanically stable components to be further processed for disposal or possible recycling. These selected strategies and techniques are aimed at decreasing the presence of mobilisable tritium on site during operation and reducing the quantities and/or classification levels of radioactive waste. The development of technological solutions and the design of the operating conditions to facilitate the next handling and destination of the materials toward disposal, recycling or re-fabrication avoiding any problem of radioactive dust dispersion have been studied.
RINA Consulting CSM has done during 2021 experimental activities for Tungsten powders hydrogenation starting by ultra-pure commercial powders with an average granulometry of 20 m. The 2022 sintering tests with Tungsten “polluted” powders to simulate exogenous species present in the DEMO W dust are described, to evaluate the feasibility of sintering and de-hydrogenation. Results will be discussed in the paper.
When a plasma experiment using deuterium (D) gas is conducted in a large fusion test device, a small amount of tritium is produced in the plasma. The produced tritium can be used to evaluate tritium behavior and inventory in fusion systems as a tracer because of its small amount. As one of the large fusion test devices, the deuterium plasma experiment with the Large Helical Device (LHD) have started on March 7th, 2017. From the viewpoint of public acceptance, the exhaust detritiation system (EDS) was installed in LHD to remove tritium in the vacuum exhaust gas or purge gas in vacuum vessel during maintenance activity. The EDS is connected to all the exhaust systems of the LHD, and tritium monitoring at the inlet of the EDS is suitable for evaluating the tritium balance. From the start of the deuterium experiment to the present, tritium released from the vacuum vessel has been continuously observed both during the plasma experiment and during vacuum vessel maintenance activities. In this study, the tritium release behavior from LHD was investigated during the mid-term deuterium plasma experiment.
Tritium was released from the vacuum vessel either directly by vacuum pumping under the plasma discharges or by wall conditioning operations or by air exposure operations due to opening the vacuum vessel to the atmosphere. Tritium in the exhaust gas was measured by an ionization chamber for real-time monitoring and a water bubbler system with chemical form discrimination for accumulated monitoring. On the other hand, the tritium produced in the deuterium plasma was estimated by neutron measurements, assuming that the number of neutrons is the same as the tritium produced by the D-D fusion reaction. Figure 1 shows the amount of released and produced tritium, accumulated tritium exhaust rate, and tritium inventory in LHD from 2017 to 2021. The tritium inventory shown in Figure 1 (c) takes into account tritium decay with a half-life of 12.3 years. Tritium released from the vacuum vessel was mostly during D and hydrogen (H) plasma discharges but was also observed when the vacuum vessel is opened to the atmosphere. The accumulated tritium exhaust rate, defined as the ratio of the amount of produced tritium to the amount of released tritium since the start of the deuterium plasma experiment, decreases when deuterium plasma experiments are conducted, indicating that some of the tritium produced remains in the vacuum vessel. The tritium analysis of the plasma-facing material after the 1st deuterium plasma experiment suggests that most of the tritium is retained in the divertor plates. However, during H plasma discharges and when the atmosphere is opened, the exhaust rate increases and approaches a constant exhaust rate (~44%). Since the annual tritium exhaust rate is about 40~50%, the tritium inventory in the vacuum vessel is increasing year by year. The tritium in the vacuum vessel is not saturated in the mid-term of the deuterium experiment, probably due to the low tritium yield.
The Diagnostic Residual Gas Analyzer (DRGA), an integrated, multi-sensor diagnostic system, will access and sample the ITER sub-divertor region, in the ducts of the cryogenic pumps, out-of-site of the main plasma chamber. It will deliver time resolved neutral gas composition measurements directly related to fuel cycle processes in the core plasma, in plasma-wall equilibration timescales [1, 2]. The system will be capable of simultaneously resolving hydrogen and helium isotopic composition, with detection limits as low as 1% and 0.1%, respectively, in terms of isotopic species concentration [3]. These are critical capabilities for the ITER research program, including the pre-fusion (pre-DT) plasma operation phases. As such, they are explicitly called out in the latest version of the ITER Research Plan [4]. The ability to carry out such measurements without the need for direct access to/through the main chamber wall, or a blanket module, makes this diagnostic approach attractive for use in future, post-ITER, burning plasma fusion devices. In this work, concepts for generalization of such a diagnostic system for next step fusion devices will be discussed. The possibility to use it for fuel-cycle process control will also be explored, both for ITER and for next generation burning plasma devices.
A future fusion reactor based on the tokamak concept in particular will need to employ methods to mitigate both edge localized modes (ELMs) and disruptions. Both of these unwanted plasma events can lead to high heat fluxes that can damage internal plasma facing components. The mitigation of these events in the plasma relies on the injection of material into the plasma to trigger a plasma response that will pre-empt otherwise naturally occurring events with more intensity that can lead to damage that will interrupt fusion energy output. The material injection for ELM mitigation is largely deuterium into the edge plasma in small pellets of sufficient size to trigger small ELMs on demand at a much higher rate than they would otherwise naturally occur. This will lead to extra plasma exhaust gas in the form of deuterium. Disruption mitigation is now envisioned to be accomplished with the deep injection of very large cryogenic pellets that are shattered upon entry into the plasma that contain hydrogenic species mixed with some higher Z impurity such as neon to efficiently radiate the plasma thermal energy.
In both of these mitigation techniques the injected material ends up as exhaust gas that must be pumped out of the reactor vessel and removed from the continuous fuel cycle. In the case of pellets for ELM pacing this would occur continuously during the plasma burn phase and require an enhanced pumping capacity and ability to remove the excess deuterium from the recirculating DT that ends up in the fueling subsystem. In the case of a disruption, the pre-emptive mitigation will result in significant amounts of gas that will cause the plasma pumping to be interrupted and switched over to a roughing pump system that can handle tritiated gas at a high pumping speed to quickly recover back to an operating reactor. In this presentation the implications of additional exhaust gases from both mitigation techniques on the fuel cycle are described and possible methods to efficiently handle the unwanted perturbations on the operation of the fuel cycle.
There is a strong relationship between fusion machine physics characteristics and the associated fuel cycle systems. As these grow in size and features, so may the facility hazards and the need for hazard mitigation. This talk will introduce the consideration of the need for hazard mitigation alongside the physics/fuel cycle relationship.
The fuel cycle of future demonstration and fusion power plants is a complex and highly dynamic system by nature, resulting from the pulsed operation of the tokamak as well as from a number of cyclic operations employed within its processing systems. The fuel cycle nevertheless has to guarantee the availability of fuel in the right quantities and composition to the plasma fueling systems, while continuously removing ash and unburnt fuel. Next to these tasks directly servicing the plasma chamber, the fuel cycle also processes tritium extracted from the breeding blankets or recovered in detritiation systems.
The strongest design driver hereby is the minimization of tritium inventories in the employed processing systems to allow safe operation of the plant and due to its limited availability. Next to design tools based on in-depth, steady state modeling, integrated process simulation of the full fuel cycle is also required to investigate and ultimately confirm its dynamic operability.
A holistic, transient process simulation of the EU-DEMO fuel cycle is therefore under development within the work package Tritium – Matter Injection – Vacuum of the European Fusion Programme. This talk gives an introduction to the EU-DEMO Fuel Cycle Simulator, outlining the modeling approach and simulation methodology, as well as presenting simulation results for the dynamic characteristics of the inner fuel cycle. It is found that transients in the inner fuel cycle are driven by the time dependent fueling requirements to the plasma in its different phases (ramp-up, flattop, ramp-down), as well as the dwell pumping phase. This sequence of different operation modes leads to pressure oscillations in the gas distribution system, which have to be buffered against to ensure the continuous fuel supply to all systems. A design study showcasing how dwell bypasses can be used to minimize the required size of buffer vessels will be presented.
Fueling, exhaust, breeding, and processing of large amounts of tritium is one of the significant technical challenges facing future deuterium-tritium fusion reactors. The propensity of tritium (as other hydrogen isotopes) to permeate through metals and other structural materials is a significant complicating factor. Successful closure of the D-T fusion fuel cycle requires that tritium losses via this mechanism remain small, and safety and environmental considerations demand that they are far smaller still. This talk will review physical phenomena relevant to the migration of tritium throughout the fusion core and beyond, their measurement and uncertainty, and predictive models and codes developed to simulate them.
Central requirements for DEMO are production of net electricity and operation with a closed fuel cycle. Thus, the machine will be equipped with a primary and a secondary coolant loop for heat extraction and energy conversion, and with a breeding blanket and a fuel cycle respectively for tritium production and processing. Such configuration inevitably opens a path to tritium migration because heat removal and tritium production occur both in the blanket region where the presence of high temperatures, large metallic surface areas and high tritium concentrations facilitate tritium permeation from blanket to primary coolant. Once in the primary coolant circuit, tritium permeates either across coolant tubes and into secondary coolant loop from which it easily reaches external environment. Obviously, the entire phenomenon is of safety relevance for DEMO and for the future fusion power plants, therefore a series of activities are ongoing with the intent to identify and implement effective strategies to mitigate tritium permeation. Along the years, simulation tools have been developed to evaluate tritium migration and inventory in blanket, operational coolants, structural materials, steam generator etc., as well as tritium losses into environment under different operating conditions. Results of these simulation analyses have been used to guide the design activities towards the definition and development of possible strategies to mitigate tritium permeation. Particularly tritium migration analyses have identified, as countermeasure to permeation, the chemistry (mainly H2 and H2O addition) in the blanket and coolant side, the use of effective anti-permeation barriers and the appropriate dimensioning of a coolant purification system.
This contribution provides a general description of the problem of tritium transport in two blanket concepts, the Helium Cooled Pebble Bed (HCPB) and the Water Cooled Lithium Lead (WCLL), highlights the impact of most relevant parameters and main results of the sensitivity analysis. In addition, the solutions currently considered for the Coolant Purification System (CPS) are illustrated and the status of their design is presented. For the case of water-CPS, a review of the existing and under development facilities for water detritiation was carried out during the pre-conceptual design phase, posing emphasis on dimension and complexity. Due to the high tritium permeation rate in water blanket concept, both strategies require the adoption of anti-permeation barrier on blanket walls. Although off-line approach presents relevant advantages such as plant simplification and less stringent issues related to the size of CPS technologies, a sensitivity analysis has demonstrated the need of an on-line CPS to reduce tritium inventory in primary coolant. For this reason, the on-line strategy has been considered in the presented design activity. For the case of helium, two processes were identified in the pre-conceptual design phase: one based on conventional technologies, such as CuO and molecular sieve beds, and one on novel getter materials for the direct adsorption of hydrogen isotopes. Recent analysis suggests that a major parameter to be considered in the definition of the helium CPS design is the H2 content in primary coolant.
Mechanisms underlying tritium retention in chamber materials can be roughly divided into two groups: trapping in deposition layers and that in bulk of materials. The contribution of trapping in the bulk to the total tritium retention could be larger in DEMO than that in existing fusion devices due to far longer discharge pulse that allows diffusion of tritium into deeper region of the materials.
Significant increase in fuel retention in W was observed after neutron irradiation to 0.01-1 dpa due to trapping effects by radiation-induced defects [1,2]. Nuclear reaction analyses of deuterium (D) profiles in neutron-irradiated D-plasma-exposed W showed that D concentration increased with decreasing temperature and reached ~1 at.% at 200 ˚C after irradiation to 0.3 dpa, Penetration depth of D into neutron-irradiated W was proportional to the square root of plasma exposure time [3,4]. The rate of penetration depends on damage level (trap concentration), temperature and hydrogen isotope flux (H/trap ratio). Penetration depth was ~50~100 micrometers after plasma exposure for 3 h at temperature of 500 ˚C and flux of ~1021 D m-2s-1 [5]. The TDS measurements for neutron-irradiated W showed broad peaks extending from plasma exposure temperature to ~1000 ˚C [2]. The apparent trapping energy was 1.4-2eV [2,5,6]. Because of the relatively large trapping energy, the fuel release at a moderately elevated temperature (~300 ˚C) was very slow [7]; the tritium removal by bake out process should be very difficult. If W monoblocks with cooling channels are used in DEMO, tritium should penetrate to cold regions around the cooling channels and be accumulated there. Nevertheless, He seeding in D plasma resulted in drastic reduction in D retention in neutron-irradiated W [8], Alloying with Re and Cr (and probably accumulation of Re by transmutation) significantly enhance annihilation of vacancy-type defects and consequently reduce fuel retention after irradiation [9,10].
Tritium trapping in deposition layers were examined via post-mortem analysis of W and Be tiles used in JET ITER-like wall experiments under the Broader Approach Activities [11]. The summary of the analyses will be given in the presentation.
[1] Y. Hatano et al 2013 Nucl. Fusion 53 073006. doi.org/10.1088/0029-5515/53/7/073006
[2] Y. Oya et al 2020 J. Nucl. Mater. 539 152323. doi.org/10.1016/j.jnucmat.2020.152323
[3] Y. Yajima et al 2019 Nucl. Mater. Energy 21 100699. doi.org/10.1016/j.nme.2019.100699
[4] Y. Yajima et al 2021 Phys. Scr. 96 124042. doi.org/10.1088/1402-4896/ac2c20
[5] Y. Hatano et al 2013 J. Nucl. Mater. 438 S114–S119. doi.org/10.1016/j.jnucmat.2013.01.018
[6] M. Shimada et al 2018 Fusion Eng. Design 136 1161-1167. doi.org/10.1016/j.fusengdes.2018.04.094
[7] V. Kh. Alimov et al 2020 Nucl. Fusion 60 096025. doi.org/10.1088/1741-4326/aba337
[8] Y. Nobuta et al 2021 Fusion Sci. Technol. 77 76–79. doi.org/10.1080/15361055.2020.1843314
[9] Y. Nobuta et al 2022 J. Nucl. Mater. 566 153774. doi.org/10.1016/j.jnucmat.2022.153774
[10] J. Wang et al 2022 J. Nucl. Mater. 559 153449. doi.org/10.1016/j.jnucmat.2021.153449
[11] S. E. Lee et al 2021 Nucl. Mater. Energy 26 100930. doi.org/10.1016/j.nme.2021.100930
In fusion DEMOs, tritium (T) decontamination scenario before maintenance begins is a key issue. Hence, it is important that T decontamination under vacuum conditions before opening the plasma vacuum vessels. Currently, JA-DEMO team has not yet determined the allowable value of residual T in the vacuum vessel, but it is necessary to indicate a candidate T decontamination technique. Furthermore, the construction of a short-term maintenance scenario that includes the T decontamination process after plasma operation is stopped, is also important for fusion DEMOs.
Three kinds of candidate techniques of T decontamination are considered in the vacuum conditions; 1. Temperature control by decay heat and baking/cooling, 2. Active wall conditionings, such as glow discharge, ion cyclotron wall conditioning, and electron cyclotron wall conditioning, 3. A selection of working gas and vacuum pressure. Mainly retained tritium on the surface of materials is important for T decontamination. Since T is easily replaced with hydrogen, it is well known that the process of replacing H with T and desorption. Therefore, the ratio of water molecules present in the space is thought to greatly influence T decontamination.
In this presentation, several T decontamination techniques are shown. The fundamental techniques required for decontamination are surface T replacements from H to T in the water of the atmosphere and temperature control. The candidate for T decontamination technologies under DEMO vacuum conditions is discussed based on the results, such as the isothermal desorption, glow discharge cleanings, and T reduction to compare with and without air contaminations, and so on.
This work is supported by JSPS-CAS Bilateral Joint Research Projects (GJHZ201984 and JPJSBP120197202).
JET is the largest tokamak in use and currently the only one capable of handling radioactive tritium (T). It operates since 2011 with the ITER-like wall (ILW), which consists of a tungsten (W) divertor and a beryllium (Be) main chamber. Following preparatory campaigns in deuterium (D), hydrogen (H) then T, JET has operated the second Deuterium Tritium Experimental campaign (DTE2, after DTE1 in 1997 with carbon based Plasma Facing Components, PFCs), aiming at answering urgent ITER needs [1].
About 1 kg Tritium has been supplied by the JET Active Gas Handling System (AGHS) during the T and DT campaigns, ran on cycles of three-four weeks of operation followed by one week of tritium reprocessing and accounting. A global gas balance was performed to assess the in-vessel T retention after one day of operation [2] by subtracting the amount of actively pumped neutrals from the amount of injected T through the gas injection systems. Long-term outgassing experiments completed the study, evidencing a faster decay of the T partial pressure compared to D. The so-determined long-term T retention will be compared with the global T accountancy by AGHS, still in progress.
After DTE2, a sequence of complementary fuel recovery methods was successfully operated to remove T from the PFCs. It consisted of baking the main chamber under vacuum at 320ºC, followed by isotopic exchange with Ion Cyclotron and Glow Discharges in D2 at this temperature, preferentially accessing T retained in the main chamber. 20 seconds long diverted plasmas with up to 16 MW NBI and ICRH power were then operated in different magnetic configurations with the main chamber at 200°C, targeting in particular the inner divertor baffle region, where the majority of the retained fuel is known to reside in thick Be deposited layers [3].
At the end of the sequence, the plasma isotopic ratio T/[H+D+T] inferred from neutron spectroscopy, was found to be ~10-4 in H-mode D plasmas, well below the 1% target set by the allocated 5∙1019 14 MeV fusion neutrons budget for the following D campaigns [4]. In total, about 4∙1023 T atoms were removed from JET PFCs, among which ~45% by baking and 50% by ICWC and GDC, the remaining 5% being removed afterwards from PFCs already depleted from T by limiter and diverted plasmas. Similar results had been obtained in a qualification experiment prior to the T campaign, where D was removed from PFCs using baking, as well as ICWC, GDC and plasma operated in H [5]. Though removal by baking seems to be less effective for T than for D, ICWC and GDC promote extra removal in both cases. Access to T buried in co-deposited layers at the upper part of the inner divertor was clearly evidenced from the increased neutron rate and elevated surface temperatures above 1200°C in plasmas that had the inner strike point raised onto this area. Still, the likely re-deposition of released material has to be assessed. For this high fidelity numerical simulations are on-going using the Monte-Carlo erosion/migration code ERO2.0 [6].