Detachment control

Summary and motivation

Trade off between problems for the divertor and problems for the core.

Detachment and heat exhaust control systems aim to meet the requirements of prolonging the lifetime of plasma facing components, particularly in the divertor, while avoiding excessive use of the actuators used to achieve and maintain detachment, which can have harmful side effects. That is, there is an optimal degree of detachment or heat dissipation that protects the plasma facing components while minimizing problems in the core.

Possible problems associated with insufficient detachment and heat dissipation:

• Thermal stress on plasma facing components due to high heat flux
• Melting due to high heat flux
• Sputtering of wall material due to high electron temperature $T_{e}$ next to wall

Possible problems associated with excessive detachment / impurity content:

• Reduced performance due to suboptimal scenario properties / excess density
• Reduced energy confinement time due to excess core radiation
• Fuel dilution due to extrinsic impurity seeding used to promote detachment
• Excitation of various MHD instabilities, reducing fusion performance further
• Increased effective charge state $Z_{eff}$ and resistivity and therefore more difficult current drive and potentially shorter pulse length
• H-L back transitions due to higher H-mode power threshold at high density and/or power loss via core radiation
• Density limit disruptions
• MARFEs
• Radiative collapse disruptions

Basic technique

The problem with an attached plasma with low radiation is that heat and particle exhaust out of the core plasma becomes concentrated in a narrow part of the chamber wall, usually in the divertor. In a tokamak, this takes the form of a narrow annulus next to the magnetic strike point. To avoid this concentration and distribute the heat exhaust load over a larger area, the flow of energy and particles through the Scrape-Off Layer (SOL) is interrupted by activating dissipation processes like radiation and charge exchange. Low $Z$ impurities like neon, nitrogen, and carbon are efficient radiators at low $T_{e}$ but less so at high $T_{e}$, which reduces their ability to cool the core plasma. To ensure that the SOL is cold enough for low Z impurities to radiate, edge density can be increased by puffing in additional hydrogenic (H,D, or T) gas. With an appropriate combination of density and impurity content, detachment can begin.

The plasma state is measured with some set of sensors connected to the Plasma Control System (PCS) to transmit data in real-time. For example, Langmuir probes can be used to estimate degree of detachment, triple-tipped Langmuir probes or divertor Thomson scattering can be used to measure $T_{e}$, or bolometers can measure radiated power. Data from the chosen sensor(s) is formulated into one or more control variables. A target or reference value is set for each control variable in the PCS, and a control policy such as PID compares the measurement to the reference to decide on a command to one more actuators.

Possible actuators are gas valves for adding fuel or impurities, impurity powder droppers, or pellet launchers. Tests so far have used gas valves.

Advanced techniques

Detachment control is fundamentally a tool for integrating core and edge scenarios. Thus, it is natural to try to combine basic detachment control with other requirements of an integrated scenario, such as ELM removal and wall conditioning.

At ASDEX-Upgrade, a detachment control system to also control impurity-induced ELM suppression. In this case, the control variable is the the height of a local radiation centroid above (in a lower null plasma) the magnetic X-point. It was found that this is first of all a viable control variable that is useful even when measurements at the divertor plate are saturated at low levels in deep detachment, and furthermore that positioning the radiator a specific distance above the X-point results in ELM suppression. Somewhat related work at DIII-D has achieved ELM mitigation by impurity seeding, but without the sophisticated X-point radiator height controller.

Another promising discovery is that injection of low-Z powdered materials such as lithium, boron, and boron nitride powders are not only useful for wall conditioning but also can result in heat flux mitigation, detachment, and ELM removal.

History

Radiated power control is the most commonly deployed control system related to dissipation and heat exhaust handling. These systems use foil or UV photodiode bolometers to measure radiated power. The first prototype was demonstrated and published in 1995 at ASDEX Upgrade, with a demonstration at DIII-D reported in 1997. Radiated power controllers have been demonstrated on CMOD JT-60U JET, and EAST, and progress has continued on ASDEX Upgrade and DIII-D.

Divertor power loads assessed with shunt resistors have been used as a control variable at ASDEX Upgrade and reported in 2010.

Electron temperature $T_{e}$ measured with divertor Thomson scattering was used as a control variable at DIII-D, and $T_{e}$ from triple-tipped Langmuir probes was used for detachment control at EAST.

Heat flux from surface thermocouples was used for feedback control at Alcator CMOD, heat flux as calculated from Langmuir probes was used at COMPASS, and a model for heatflux was used for control at DIII-D.

Attachment fraction, based on ion saturation current $I_{sat}$ measurements from Langmuir probes, has been used as a control variable at JET, EAST, DIII-D, and KSTAR. While many other control systems have developed semi-independently, the JET design was the direct basis for the successors at EAST and DIII-D. The KSTAR implementation was also a result of this lineage, but with modifications resulting from lessons learned while operating with the JET design.

The position of the detachment front along the divertor leg (between the X-point and the divertor target plate) has been controlled on TCV using the MANTIS camera to view C-III emission (peaks at about 8-10 eV).

The position of a radiation centroid relative to the magnetic X-point has been controlled at ASDEX Upgrade.

In 2020, ITPA DSOL 43 was formed to coordinate global efforts to develop detachment control systems.