Since 18 of December 2019 uses Nucleus credentials. Visit our help pages for information on how to Register and Sign-in using Nucleus.
10-15 May 2021
Nice, France
Europe/Vienna timezone
The Conference will be held virtually from 10-15 May 2021

The impact of fuelling and W radiation on the performance of high-power H-mode plasmas in JET-ILW

12 May 2021, 08:30
Nice, France

Nice, France

Regular Poster Magnetic Fusion Experiments P3 Posters 3


Dr Anthony Field (UKAEA)



In high-performance, ‘ITER-baseline’ or ‘hybrid’ scenario pulses in JET-ILW with high levels (≲36 MW) of heating power, typically ~20-40% of the input power is radiated, predominantly by W impurities sputtered from the divertor targets that reach the confined plasma. Sustained ELMy H-mode operation at such heating powers in JET-ILW requires gas puffing to increase the ELM frequency $𝑓_{ELM}$ and hence the rate of W ‘flushing’ from the confined plasma, as well sweeping of the strike point to avoid overheating of the target. Core ICRH heating is also used for control of heavy impurity accumulation [1]. A deleterious effect of the gas puffing is to reduce the pedestal temperature $𝑇_{e,ped}$ [2] by increasing the inter-ELM pedestal heat transport [3, 4], which consequently reduces the pedestal pressure [5] and the overall confinemen compared to that of similar JET-C pulses at the same heating power.

ITER-baseline pulses

With gas fuelling alone, if the puffing rate is insufficient, the duration of high-power JET-ILW 'ITER-baseline’ scenario pulses is limited by a gradual increase of the W content and associated radiation, which can cause a transition to L-mode, e.g. as occurs in the gas-fuelled pulse #92432 shown in Fig. 1.

Partially replacing some (~30%) of the gas puffing by injection of small, ELM-pacing pellets, resulting in ~30% less fuelling overall, e.g. in similar pulse #92436, is found both to extend duration of the ELMy H-mode phase and to enhance overall confinement compared to that achieved with gas fuelling alone. As shown in Fig. 1 (d, e), the pacing pellets induce smaller, rapid, more irregular ELMs than those occurring with only gas fuelling. Note that these ELMs are not well synchronised to the pacing pellets, the pellet-triggered ELMs often being followed by bursts of small ELMs, which may be caused indirectly by the pellets, e.g. by changes to the pedestal/SOL characteristics or edge neutral gas pressure. Another beneficial effect of operating at reduced gas puffing rate with pacing pellets is an enhancement in confinement, which is associated with increased toroidal rotation and reduced ion temperature gradient stiffness [6].

Comparison for 3MA/2.7T pulses (with 32MW heating) #92432 (gas fuelling only) and #92436 (pellets & gas) of: (a) line-average density $\overline{n}_e$ and gas fuelling rate $\Gamma_{D}$ (dashed), (b) stored energy $W_{pl}$, (c) radiated power fraction $\mathcal{F}_{Rad}$, (d, e) Be II intensity viewing divertor $I_{BeII}$, and (f) $D_\alpha$ intensity viewing the pellet entry.

W flushing and behaviour

In such high-power, baseline-scenario pulses, the W impurities are mainly localised to the outer, ‘mantle’ region ($0.7 \le \rho_N \le \rho_{N,ped}$) by outward neo-classical convection, e.g. as shown in Fig. 2, which compares the evolution of the pulses #92432 and #92436. By decreasing the net loss power across the separatrix, an increase of this W radiation reduces $f_{ELM}$, thereby decreasing the rate of W flushing from the confined plasma, causing yet more radiation. This ‘vicious cycle’ can be broken by maintaining $f_{ELM}$ using ELM pacing pellets, which is not as deleterious to the overall confinement as is doing so by increasing the gas puffing rate.

*Comparison for pulses #92432 (left, gas-only) and #92436 (right, pellets & gas) of: (a) the signal $P_{Rad}^{Pl}/\overline{n}_e$ (blue) and average W concentration $\overline{C}_W$ (red); (b) relative change in W content $(\Delta \overline{n}_Z/\overline{n}_Z)$ due to ELMs (natural-'green', pellet-'red'), inter-ELM fluence ('blue') and net change per-ELM ('black'); (c) the associated rates of change $\Gamma_W$; (d) flux-surface averaged emissivity $\langle \epsilon_{tot}\rangle$ vs $\rho_N$; and (e) the N-C convection parameter $\zeta_{NC}$ averaged over the mantle ('blue') and pedestal temperature $T_{e,ped}$ ('red').*

A measure of the W flushing efficiency of each ELM $(\Delta\overline{n}_Z/\overline{n}_Z)_{ELM}$, as well as the inter-ELM W fluence $(\Delta\overline{n}_Z/\overline{n}_Z)_{i-ELM}$ can be evaluated from fast bolometric measurements of the radiated power from the confined plasma $P_{Rad}^{Pl}$ and the line-average density $\overline{n}_e$ from the interferometer [7], by availing of the fact that the W emissivity $\xi_W \sim 4.5\times10^{-31}\ Wm^3$ is quite constant over the mantle region, where $T_{e,ped}$ is typically 1-2 keV. By classifying the ELMs as either ‘pellet-triggered’ or ‘natural’ events, the $(\Delta\overline{n}_Z/\overline{n}_Z)$ data, e.g. as shown in Fig. 2 (b), as well as post-ELM profile gradients measured by Thomson scattering and edge CXRS systems, can thereby be conditionally averaged, thereby quantifying the data according to the ELM type.

Averaging data from four high-power 3 MA/32 MW baseline pulses, with either gas fuelling alone (#92432/3) or with gas and pellet fuelling (#92434/6) we find: As a result of the pellet ablation, the amplitude, expressed in terms of $(\Delta\overline{n}_e/\overline{n}_e)$, of the pellet-triggered ELMs is about half that of the natural ELMs occurring with both fuelling methods. The W flushing efficiency of the ELMs in the gas-fuelled pulses is ~5 times than that for both ELMs types in the pulses with pacing pellets. However, the shorter inter-ELM periods for the natural ELMs in the pellet pulses result in smaller inter-ELM W fluences into the plasma, resulting in similar net rates of change $\Gamma_{W}$ of the W content (see Fig. 2 (c)).

Relative, ELM-sputtered Be and W fluences $\Phi_{Be,W}$ can be estimated from time-integrated Be II (527 nm) and W I (401 nm) line intensities from a multi-chord spectrometer viewing the outer divertor targets. Similarly conditionally-averaged fluences $\Phi_{Be,W}$ are larger ($\times1.8$ and $\times1.4$) for the natural ELMs in the gas-fuelled pulses than for both ELM types in the pellet fuelled pulses.

During the sustained ELMy H-mode phase of pulse #92432, outward neo-classical convection, which is proportional to the parameter $\zeta_{NC} = R/2L_{T_i}-R/L_{n_i} > 0$ [7], helps concentrate the W in the mantle region. However, after a transition to L-mode,the convection changes sign ($\zeta_{NC}<0$) causing the W to accumulate in the core (see Fig. 3 (e)). Although the inward neo-classical convection across the pedestal is always strong ($\zeta_{NC} << 0$), it is significantly reduced ($\times0.7$) following the pellet-triggered ELMs. Neo-classical screening ($\zeta_{NC} > 0$) of the W from the core by the mantle gradients is however weaker ($\times0.7$) following the pellet-triggered ELMs, likely due to changes to the mantle gradients caused by the subsequent burst of smaller, natural ELMs.

W behaviour in 'hybrid' pulses

Strong radiation from W impurities has a similar influence on the evolution of high-power, ‘hybrid’ scenario pulses (2.2MA at 2.8 or 3.4T), with higher $\beta_p \sim 1.3$ (c.f. ~0.75 in the 'baseline' scenario pulses #92432/6) and a broader, elevated q-profile ($q_0 > 1$). The more peaked density profiles of ‘hybrid’ than baseline pulses often reduces localisation of W impurities to the mantle region. Local flattening of the temperature gradient by core MHD modes (NTMs), which occur as the $q$-profile evolves, can also trigger W accumulation. However, this mechanism can be avoided by operating at higher toroidal field (3.4T) and hence higher safety factor $q$. Results illustrating these processes will be included in the full paper.

[1] Garzotti L. et al., IAEA FEC ‘18
[2] Giroud C. et al., Nucl. Fusion 53 (2013) 113025
[3] Field, A. R. et al., EPS 2018
[4] Hatch D. et al., Nucl. Fusion 57 (2017) 036020
[5] Maggi C. et al., Nucl. Fusion 55 (2015) 113031
[6] Kim H-T et al., Nucl. Fusion 58 (2018) 036020
[7] Fedorczak, N. et al., JNM 463 (2015) 85-90
[8] Angioni, C. et al., Phys. Plasmas 22 (2015) 055902

This work has been carried out within the framework of the EUROfusion Consortium and has received funding from the Euratom research and training programme 2014-2018 and 2019-2020 under grant agreement No 633053 and from the RCUK Energy Programme [grant number EP/T012250 /1]. The views and opinions expressed herein do not necessarily reflect those of the European Commission.

Country or International Organization United Kingdom
Affiliation UKAEA

Primary author

Dr Anthony Field (UKAEA)


Clive Challis (Culham Centre for Fusion Energy) Luca Garzotti (United Kingdom Atomic Energy Agency - Culham Centre for Fusion Energy) Athina Kappatou (Max-Planck-Institut für Plasmaphysik) Hyun-Tae Kim (EUROfusion Consortium JET) Morten Lennholm (European Commission) Ernesto Augusto Lerche (LPP-ERM/KMS) Costanza Maggi (CCFE) Dr Andrew Meigs (UKAEA) Dr Fernanda Rimini (UKAEA) Colin Roach (Culham Centre for Fusion Energy) Dirk Van Eester (LPP-ERM/KMS) Dr Gabor Szepesi (UKAEA) Dr Marco Sertoli (Max-Planck-Institut für Plasmaphysik)

Presentation Materials

There are no materials yet.