MkI Divertor Research Articles

Reassessment of tritium content in CFC tiles exposed to the JET D-T campaign in 1997

• In 2019 two MkIIA divertor tiles (6IN3 and 4BN4) exposed during DTE1 were retrieved at CCFE for Thermal Desorption Spectroscopy (TDS) and pyrolysis analyses. A set of samples were prepared using a coring technique. The highest tritium (T) inventories were found in the shadowed corner of the inner divertor due to asymmetric deposition. TDS analyses indicated that T is desorbed at rather high temperatures with maximum release peaks at ∼590 and 820 o C. T amounts were also investigated as a function of depth from the tile surface up to a depth of ∼4.5mm and T was detected at these depths. Comparison was also made with old results obtained with full combustion technique and PIN-diode method a few years after the DTE1 experiment, allowing for the natural decay of T. Our results agree within a factor of ∼3 with these results. In 2019 two MkIIA divertor tiles (6IN3 and 4BN4) exposed during DTE1 were retrieved at CCFE for Thermal Desorption Spectroscopy (TDS) and pyrolysis analyses. A set of samples were prepared using a coring technique. The highest tritium (T) inventories were found in the shadowed corner of the inner divertor due to asymmetric deposition. TDS analyses indicated that T is desorbed at rather high temperatures with maximum release peaks at ∼590 and 820 o C. A few samples were reannealed at 850 o C using the same heating procedure and it turned out that a further ∼40-50 % of T was still released indicating that the annealing procedure used does not empty the sample completely. Pyrolysis results for thin disks cut from the surface of the tile were somewhat higher than the corresponding TDS results. T amounts were also investigated as a function of depth from the tile surface up to a depth of ∼4.5mm and T was detected at these depths. Comparison was also made with old results obtained with the pyrolysis technique and a PIN-diode method a few years after the DTE1 experiment, allowing for the natural decay and off-gassing of T. Our results agree within a factor of ∼3 with these results.

EDGE2D-EIRENE simulations of the impact of poloidal flux expansion on the radiative divertor performance in JET

In the past at JET, with the MkI divertor, a systematic study of the influence of X-point height and poloidal flux expansion has been conducted [1,2] showing minor differences in the radiation distribution, whereas in [3] experiment and simulations have shown enhancement of detachment as the flux expansion was increased. More recently at JET, equipped with the ITER-like wall (ILW), radiative seeded scenarios have been studied and a maximum radiation fraction 75% has been achieved. EDGE2D-EIRENE [4–6] simulations [7,8] have already shown that the divertor heat fluxes can be reduced with N2 injection, qualitatively consistent with experimental observations [9], by adjusting the impurity injection rate to reproduce the measured divertor radiation. In this paper we will present edge predictive simulations on the assessment of effects of poloidal flux expansion and recycling on radiation distribution and X-point peaking on JET-ILW nitrogen seeded plasmas.

Tritium distribution measurement of JET Mk II SRP divertor tiles

Tritium surface distribution on the JET Mark II Septum Replacement Plate (Mk II SRP) divertor tiles was measured by imaging plate technique. It was observed that areal tritium concentration was higher at the entrance of inner/outer pumping slots (so called ‘shadowed area’). The tritium distribution profiles were similar to those obtained in the Mk IIA divertor which was exposed to a series of D–T plasma operation (DTE1). Tritium concentration of the plasma facing surface was lower compared to that of the shadowed area. Particularly, it was very low at the outer divertor surface. The inner divertor surface also showed low level of tritium retention, though it was covered by the thick carbon deposition on that. This could be caused by tritium release due to the temperature rise when the inner strike point was on the tiles. On the plasma shadowed area like tile gaps, high tritium retention owing to the codeposition was observed.

Detailed Tritium Distribution on the JET MK IIA Divertor Tiles

Tritium surface distributions on the plasma-facing surface and four sides of JET Mk IIA divertor tiles employed in the D-T operation phase of JET were measured by Tritium Imaging Plate Technique (TIPT). Tritium distribution on the plasma-facing surface was consistent with carbon deposition profiles and asymmetric in both poloidal and toroidal directions. The toroidal asymmetry was attributed to the alignment of the tiles preventing direct impact of flux lines to tile edges. Accordingly, no significant carbon deposition or tritium accumulation was observed on two sides facing the toroidal direction. As already reported, heavy codeposition retaining high levels of tritium was observed on the plasma-shadow area of the horizontal target tile surface and the bottom side of the vertical target tile of the inner divertor region where it was kept relatively cool by water coolant. In addition, TIPT has clearly distinguished at least two different carbon deposition layers with different tritium retention in poloidal direction, showing that the poloidal asymmetry on the horizontal target tiles is due to the different carbon deposition properties in the poloidal direction. All the results suggest that tritium retention in the divertor area, which was determined by the carbon/hydrocarbon distribution, correlates closely with divertor geometry and surface temperature.

Tritium Retention in the Gap between the Plasma-Facing Carbon Tiles Used in D-T Discharge Phase in JET and TFTR

Tritium accumulating in codeposits in the gaps between plasma facing components is a safety concern in next step fusion machines as suitable removal techniques have yet not been developed. We report on Imaging Plate measurements of the tritium areal distribution on the side surface of graphite/CFC tiles installed in the TFTR bumper limiter and JET Mk IIA divertor, both of which were exposed to D-T discharges. The tritium profiles on the four sides of TFTR tiles showed a short- and long-range decay pattern. In case of JET divertor tiles, only a small amount of tritium retention was detected on the tiles side facing the toroidal direction, while tritium retention was very large on the side facing the poloidal direction. These retention properties showed that the orientation or alignment of plasma facing component plays important role on the tritium retention in the gaps of those machines.

Modelling of carbon transport in fusion devices: evidence of enhanced re-erosion of in-situ re-deposited carbon

The paper presents new Monte-Carlo transport simulations of methane 13CH 4 injected through a hole in a testlimiter and exposed to the edge plasma of TEXTOR. The results show that the spatial distribution of 13C re-deposited locally on the testlimiter surface can be modelled if the parameter S for the sticking of returning hydrocarbons 13CH y is set to zero or almost zero. This is interpreted as a negligible effective sticking of the returning hydrocarbon radicals due to the instantaneous re-erosion caused by the hydrogen carried with the CH y radicals (`self re-erosion'). However, the calculated local deposition efficiency of 13C, remains too high compared with the observed value. Therefore, in addition an enhanced yield for chemical erosion caused by the background hydrogen for the fresh re-deposits has to be assumed. Similar assumptions can reproduce also the high amount of carbon deposition found on the inner louvers in the MkIIa divertor configuration of JET and on the plasma-shadowed areas of the MkIIGB divertor.

Modelling of carbon transport in fusion devices: evidence of enhanced re-erosion of in-situ re-deposited carbon

The paper presents new Monte-Carlo transport simulations of methane 13CH4 injected through a hole in a testlimiter and exposed to the edge plasma of TEXTOR. The results show that the spatial distribution of 13C re-deposited locally on the testlimiter surface can be modelled if the parameter S for the sticking of returning hydrocarbons 13CHy is set to zero or almost zero. This is interpreted as a negligible effective sticking of the returning hydrocarbon radicals due to the instantaneous re-erosion caused by the hydrogen carried with the CHy radicals (`self re-erosion'). However, the calculated local deposition efficiency of 13C, remains too high compared with the observed value. Therefore, in addition an enhanced yield for chemical erosion caused by the background hydrogen for the fresh re-deposits has to be assumed. Similar assumptions can reproduce also the high amount of carbon deposition found on the inner louvers in the MkIIa divertor configuration of JET and on the plasma-shadowed areas of the MkIIGB divertor.

Hydrocarbon transport in the MkIIa divertor of JET

This contribution presents two-dimensional Monte-Carlo simulations of the local transport of hydrocarbons that are chemically eroded in the JET MkIIa divertor. The effect of a given background carbon flux flowing from the main plasma down to the divertor is also taken into account. The influence of local plasma temperatures and densities and the influence of different assumptions for the sticking of re-deposited particles hitting the tiles is analysed. Even under the assumption that the sticking of re-deposited hydrocarbon fragments is zero, a large amount (about 75%) of the eroded particles is re-deposited in the form of ionized carbon on the tiles. A reasonable agreement between the simulation and the experimental observation of the carbon deposition at the inner louvers in the MkIIa JET divertor is achieved, if high chemical erosion yields of about 20% together with a negligible effective sticking of hydrocarbons are assumed. Although there are experimental observations indicating such high erosion yields, such yields cannot be applied as a stationary value for the effective erosion since it would turn the inner divertor into a net erosion area in contrast to the experimental findings. The possibility of applying different erosion rates for (re)-deposited layers and the substrate material is discussed. To further improve the modelling a better knowledge about the formation and erosion of re-deposited layers is required. A first conclusion concerning the erosion and re-deposition behaviour in ITER will be drawn.

Non-destructive tritium measurements of Mk IIA divertor tile by BIXS

β-Ray-induced X-ray spectrometry was applied to determine the amounts and depth profiles of tritium captured by graphite tiles used in the PTE1 campaign and by carbon fiber composite tiles in the DTE1 campaign of JET. Although the X-ray spectra observed for both carbon tiles consisted of characteristic and bremsstrahlung X-rays, the intensities and shapes of the X-ray spectra were quite different between these tiles. The analyses of the X-ray spectra showed that the amount of tritium retained in the latter tile was on average 15 500 times greater than that in the former tile, and the tritium depth profile extended to 60 μm for the latter tile compared to 5 μm for the former tile. In addition, characteristic X-rays corresponding to Cl, Cr, Fe, Ni and Mo appeared in the latter tile, where Ni showed the highest concentration, indicating that considerable erosion and deposition of vacuum vessel materials took place in the DTE1 campaign.

Advances in the modeling of chemical erosion/redeposition of carbon divertors and application to the JET tritium codeposition problem

We have improved the modeling of chemically eroded carbon transport and applied this to JET and ITER. New features are: (1) coupled REDEP and ERO-JET impurity transport code calculations for sputtered wall/divertor carbon, (2) MolDyn molecular dynamics calculations of carbon/hydrocarbon particle reflection at hydrogen-saturated carbon surfaces, (3) ADAS full collisional radiative carbon ion recombination rate coefficients. At low incident particle energies relevant to chemical-erosion (∼0.1–15 eV), we predict high reflection coefficients (∼20–100%), implying more net erosion and T/C codeposition than for full-sticking models. Calculated tritium codeposition rates for the JET MkIIA divertor, using ‘reference’ chemical erosion yields of order 1% – while higher than previously estimated are well short ( X∼1/40) of published data. Possible explanations include much higher chemical erosion yields.

Tritium off-gassing trials on dust and flakes from the JET MKIIA divertor

In 1998, 100 g of highly tritiated carbon dust and flakes containing an estimated 1 g of Tritium were removed from the JET machine using a vacuum cleaner with a cyclone dust separator during the Remote Exchange of the divertor tiles. Two pots of flakes were filled and transferred, in drums, to the Active Gas Handling System where they were stored and ventilated to the Exhaust Detritiation System (EDS). A ventilated glovebox allowed the material to be handled safely and undergo calorimetric tritium assay. The glovebox had an air atmosphere with a continuous purge to the EDS with a facility for argon purging of isolators within the glovebox. Design features of the glovebox, operational experiences in handling tritiated carbon with a specific activity in excess of 1 TBq/g and results of the tests are discussed.

A report on the first remote handling operations at JET

The first fully remote handling operations on a fusion Tokamak have been completed at JET. The JET MkIIa divertor has been successfully removed and replaced with the MkII gas box divertor configuration using only remote handling methods inside the JET torus. This shutdown required the handling of over 450 components over a 15 week period operating 20 h per day and 6 days per week. The JET remote handling equipment functioned and performed almost flawlessly with only two significant faults. The total operational time lost due to equipment problems was less than 3 days. The operations proceeded safely and without significant problems. No time was lost due to either operator induced equipment problems or operational errors. A number of new tasks were successfully included in the operations at short notice. Some new tasks were identified immediately before the shutdown but with sufficient time available to test their feasibility using the mock-up facility. Others were only evident after the shutdown had started. The successful implementation of the shutdown and in particular of the unprepared tasks has been due to the high degree of adaptability provided by the JET man-in-the-loop remote handling approach.

Development of popup Langmuir probe system for the JET MkIIa divertor

The successful operation of a popup Langmuir probe system, which was installed in the JET MkIIa divertor, is described. The system utilises the ambient magnetic field in tokamak plasmas to act on a current carrying coil and pop up a rail containing Langmuir probes. Measurements were made using pin-plate probes which, owing to their relatively large exposed area, are ideally suited for use with such a system. Details of the design, testing, measurements and potential applications of JET’s popup system are given.

Interpretative modelling of JET's thermal helium diagnostic

Electron density, n e, and temperature, T e, in the MkIIa divertor have been measured using neutral helium line ratios. Experimental evidence indicates that the spectroscopic signal for helium injected from the vertical target is dominated by recycling. When helium is injected from the horizontal target it is possible to differentiate between injected and recycled spectral sources. To determine the spatial location of the neutral helium emission, and hence of the n e, T e measurements, an interpretative model has been developed which uses the impurity/neutral codes, DIVIMP/NIMBUS, the atomic database ADAS and the diagnostic lines of sight. Measurements and simulations have been performed using the various injection points over a range of target conditions. Comparisons with other divertor diagnostics, such as the divertor interferometer and target Langmuir probes are presented.

Dust and flakes in the JET MkIIa divertor, analysis and results

After operation of the JET machine with the MkIIa divertor, samples of dust and flakes/deposits have been removed from the divertor region for analysis. Flakes/deposits up to 40 μm thick have been found in the inner leg of the divertor. The flakes/deposits formed on cold surfaces shadowed from the plasma. The flakes were found to be mostly carbon saturated with deuterium to a D/C ratio of 0.4. Analysis of the material removed has shown that it contains significant fractions (∼4%) of the tritium (produced by D–D reactions) and deuterium introduced in the JET machine during this period of operation. Dust material has also been collected. Vacuum cleaning collected very little material but a significant quantity of material was collected by smears. The analysis of this smeared material has shown it to be of a very similar chemical composition to that of flakes but to contain a significantly lower inventory of tritium. The dust has an average particle size of 27 μm and a specific surface area of 4 m 2/g. The vast majority of this material is associated with co-deposited layers in the divertor region. The dust is on average present at a level of 120 μg/cm 2.

The role of recycling and impurity production in JET hot-ion H-modes

The effect of impurity production and recycling on JET hot-ion H-mode discharges is investigated for the MkI and MkII divertors. Analysis of the best of these discharges has been performed with the EDGE2D/NIMBUS codes and the results compared with a wide range of diagnostics. For the very low recycling hot-ion regime, the code predicts that Z eff at the edge (in the upstream SOL) increases with SOL density, and is higher for the MkII than for the MkI divertor. These predictions have been confirmed by the experimental data. The significance of the edge density and edge Z eff for the hot-ion regime is that the loss power has been predicted from neo-classical theory to scale as: P loss ∼ n edge 2 Z eff,edge. It has now been demonstrated that this is consistent with both the MkI and MkII experimental data. It has been observed that the carbon sources in the MkII divertor are increased relative to those in the MkI divertor. These data are compared with the modelled results from the EDGE2D code, taking into account the change in the chemical sputtering yield with the different base temperatures of the MkI (40°C) and MkII (270°C) divertors.

SOL width scaling from consideration of edge transport in tokamaks

A simple two-point representation of the tokamak scrape-off layer (SOL), based on the parallel and perpendicular energy balance equations, has been used to develop heat flux width scalings for a wide variety of χ ⊥ models in terms of B φ , n sep, q 95, P SOL, R and a. These scalings have been tested against databases of SOL parameters established on COMPASS-D (which exhibits a collisionless SOL), JET Mk IIa divertor and Alcator C-MOD (both of which exhibit strongly collisional SOLs). The better fit scalings to the collisionless dataset all show a positive dependence of SOL width on q 95 and an inverse dependence on n sep. Four of the better fit scalings also show a positive dependence on P SOL and a strong inverse dependance on B φ . In the collisional case a group of 5 models, each with a χ ⊥∝ T 1/2/ n dependence, provide the better fit scalings to both collisional datasets. The derived scalings exhibit a strong inverse dependence on P SOL but are independent of B φ , unlike scalings derived from the commonly used Bohm diffusion model, and four of the five are also independent of n sep.

Overview of probe measurements at JET

This paper presents an overview of the advance in probe design and measurements made at JET during the MkIIa divertor experimental campaign. In addition to the fixed probe system, new probes have been implemented specifically to investigate the effect of plasma resistivity on probe derived electron temperatures and to measure turbulence in the SOL.

Erosion/redeposition at the JET MkI divertor

Erosion/redeposition at the JET MkI divertor

Measurement and analysis of radiated power components in the JET MkI divertor using VUV spectroscopy

Measurement and analysis of radiated power components in the JET MkI divertor using VUV spectroscopy