Detachment Proposals
-
-
Speaker: Dr Thierry Kremeyer (MPG)
Dear Yuhe, Dear Sebastijan, Dear colleagues,
Thank you all for the further comments.
I agree with Yuhe that the points raised by Glen and Fred regarding a density gradient as well as impurity retention are performance issues and do not affect its definition. Nevertheless it's good to know how we qualify the performance of detachment. Currently we discussed:
- neutral pressure at the pumpgap
- compatibility with peaked profiles
- impurity retention by i.e. high SOL density.
As performance parameters.
Is the low target temperature, < 5 eV for W, also a performance issue and should not be part of the definition? If this would also be a performance parameter and not a necessary condition, it would allow us to make the power balance a sufficient condition.
I therefore find Yuhe's definition of detachment appealing as this would allow us to use the power balance as a sufficient condition with f_rad as an analogue metric to measure the degree of detachment within the parameter space.
Sebastijan, what is your physics definition of detachment? I couldn't find a general valid definition.
If we follow Stangeby's definition:
“We are therefore interested in finding a way to decrease the particle flux reaching the target, for given upstream conditions in the SOL. Such a state has been found and is called divertor detachment. […] The characteristic, in fact the defining, observations of detachment are:
- ‘roll-over’ and decrease of the ion saturation current, j+sat, of Langmuir probes built into the inner and outer divertor targets, (d), (f), i.e. the ‘nt -rollover’ seen on ASDEX, figure 16.4;
- the Dα radiation from the target regions does not roll over/decrease, but continues to increase with ne, (e), (g).
A further feature of detachment is that it occurs when the target Langmuir probes indicate low temperatures, Te ≈ a few eV or less “
We never had detachment, because H_a radiation did roll over and decrease. So we only fulfilled one of Stangeby's two necessary conditions. Yet from a practical side we were able to detach the heat and particle loads, which is the crucial necessity for detachment. Not calling it detachment because it lacks this observation from tokamaks while fulfilling the reactor relevant necessity would seem a bit silly to me.
Best,
Thierry
Dear Thierry, dear colleaques:
I missed the discussion last week due to conflict with another session.
Here is my view of 'detachment'.
"Detachment" is a SOL plasma state in which the power balance is governed by radiation and neutral gas ionization. In contrast to other regimes (e.g. linear and high-recycling), the fundamental characteristic of this regime is that both particle and heat fluxes decrease with radiation (or plasma density). Therefore, a rollover in ion saturation currents or downstream density is often considered an experimental sign of the onset of detachment. Detachment is not an operating point but covers a region in the plasma parameter space.
The degree of detachment, its completeness, heat and particle exhaust capability, compatibilty with confinenment and impuriy screening, and reactor-relevance are all performance issues for this regime, not its definition.
best regards,
Yuhe
Hi Thierry,
also to your third question regarding reactor relevance of detachment:
I would add impurity control to the list. We need to understand the sources (already mentioned sputtering but also seeding) and the SOL-transport during detachment.
A reactor will need good impurity retention properties in order to keep the impurity content in the confined region low because of fuel dilution and radiation. At the same time in the SOL these impurities are essential for the detached plasma.
Best,
Frederik
Thierry,
One more point, in a W7-X style machine. Detachment by itself is worthless if we can't simultaneously break the ion temperature clamping at 1.5 keV. This means we have to be able to enhance the density gradient, by any means at hand. If puffing to initiate detachment, or high f-rad, or large recycling coefficient prevents that.....then we don't have a solution.
Glen
Dear Thierry,
I believe that the detachment defintion as physics definition holds and I warn to change this. What you rather mean is the operational window for W7-X to be within the specifications of the components. This is something different.
In view of a reactor-like machine, you need also to stay below the sputtering threshold for W in steady-state including impurities. Thus, you will end-up with max. 5eV.
Regarding the ion clamping, I would rather look on ICRH plasmas before giving up...
Sebastijan
Dear colleagues,
I would like to recap our discussion on the definition of detachment for W7-X. Thank you for your feedback, prior via E-mail, or in the meeting.
Stangeby’s detachment definition is not fully applicable to our detachment, so we’ll need our own. The definition should be robust and practical, with the goal of what we want to achieve in mind.
The necessity of detachment in a reactor originates from two conditions:
- Reduce heat loads to a level <10 MW/m²
- Minimize sputtering to tolerable levels
Additionally, for a degree/level/grade/intensity/… of detachment to be considered “good” or reactor relevant, other conditions have to be added, like enough neutral pressure to get into a continuous flow at the pump gap and a SOL that will reduce neutral penetration so that recycling impurities will not make it back into the core. A high SOL density can also increase momentum losses and therefore minimize sputtering, so it’s desirable for two reasons.
“Detachment” (noun), in the widest sense is a disconnection, disengagement, or a “de-coupling of things” as Ralf said.
In our case the detachment of plasma with the divertor target.
This is not a sudden mode switch, but rather a continues transition process on a spectrum with “attachment”, where the plasma is in full contact with the divertor, on the other side. Along this transition, the radiation- and ionization-front movements play a crucial role.
However this continuous behavior doesn’t rule out that there are certain milestones or key events along the spectrum that cause drastic changes. I.e. when a front crosses the LCFS.
Detachment can be divided into energy, particle, and momentum detachment.
We also discussed that full detachment is not necessarily beneficial from a reactor perspective, so our ideal working point will be somewhere along the way. We therefore need an analog metric that depicts the entire spectrum and ideally can be used as a control parameter. From an engineering and control aspect, it is advantageous if this value doesn’t have to be hit exactly, but it has to be a value where we know in which direction we need to “turn the knob” in order to get to our desired position on the spectrum.
We don’t want to know our altitude on a nanometer scale, we want to fly an airplane.
In regards to our goal of reducing heat loads by intercepting the heat flow on its way to the targets, energy and particle detachment are closely linked by our global power balance (without volume recombination and charge-exchange):
[Feng, NF, 2021]
https://iopscience.iop.org/article/10.1088/1741-4326/ac0772/pdf
As our pumping is insignificant in regards to our recycling flux, there are no pumps included in this modeling study. Therefore the ion influx on the target is interchangeable with the total recycling flux. The first term is the energy that the ion flux transports onto the target with T_t being an average value for the ion and electron temperatures at the targets, and gamma: the total sheath transmission factor. Epsilon_i is the total energy cost of electrons per ionization event of the recycling flux.
This demonstrates why many of us already use f_rad as a key metric for energy and particle detachment as it covers the entire spectrum from fully attached at f_rad = 0, to fully detached at f_rad = 1. It therefore covers our entire spectrum and is an excellent candidate to quantify our degree/level/grade (we should agree on a term for this!) of detachment. With feedback gas fueling on f_rad in OP2.2, it can also be used as a control parameter.
We didn’t discuss much the minimization of sputtering by momentum detachment. A critical metric is decreasing the edge temperature, T_e < 5 eV. Are there other hard requirements in regard to sputtering, particle momentum losses, or maybe something else that we have good diagnostic coverage of? How critical is n_e at the target?
As not everyone was in the meeting, I’d like to ask everyone, from PhD student to professor, for your opinion again:
- When you talk about detachment, what are your necessary conditions? What has to be true for you, to consider a discharge to be called detached?
- As we already identified 2 necessary conditions (reduced heat loads and low target temperatures), it will be difficult to find a sufficient condition for detachment. But maybe we can think of detachment being a sufficient condition for something else: If a discharge is detached, then what else is true?
- Beyond the fundamental definition of detachment: What has to be true for reactor relevant detachment from the edge perspective? So far we have high neutral pressure and high SOL density.
Following the basic rules of open communication, there are no wrong questions and everyone’s opinion on this matters! Therefore any other feedback on this recap or our discussion is highly welcome.
Don’t hold back to share your thoughts, you don’t have to be right. Feel free to also send me your reply privately if the “Reply all” button puts too much pressure on you ;)
Best,
Thierry
PS:
This email chain and my slides are stored on indico for documentation:
https://indico.euro-fusion.org/event/2015/
_________________________________
Dr. Thierry Kremeyer
Max Planck Institut für Plasmaphysik
Stellarator - Rand - und Divertorphysik (E4)
7.0-013
Wendelsteinstr. 1
17491 Greifswald
Germany_________________________________
Tel +49 3834 88 1496
+49 3834 88 2377
Mail thierry.kremeyer@ipp.mpg.dehttps://orcid.org/0000-0002-6383-944X
Von: Marcin Jakubowski <marcin.jakubowski@ipp.mpg.de>
Gesendet: Dienstag, 12. April 2022 09:40
An: 'Felix Reimold' <flr@ipp.mpg.de>; 'Glen Anthony Wurden' <wurden@lanl.gov>
Cc: 'Thierry Kremeyer' <thierry.kremeyer@ipp.mpg.de>; 'Arturo Alonso' <arturo.alonso@ciemat.es>; 'Arun Pandey' <arun.pandey@ipp.mpg.de>; 'David Bold' <dave@ipp.mpg.de>; 'David Ennis' <dae0005@auburn.edu>; 'Dhard, Chandra Prakash' <chandra.prakash.dhard@ipp.mpg.de>; 'Dorothea Gradic' <dorothea.gradic@ipp.mpg.de>; 'Erik Flom' <erik.flom@ipp.mpg.de>; 'Frederik Henke' <frederik.henke@ipp.mpg.de>; 'Georg Schlisio' <georg.schlisio@ipp.mpg.de>; 'Gregor Pechstein' <gregor.pechstein@ipp.mpg.de>; 'Kenneth Hammond' <khammond@pppl.gov>; 'Maciej Krychowiak' <maciej.krychowiak@ipp.mpg.de>; 'Matt Kriete' <kriete@auburn.edu>; 'Matthias Otte' <matthias.otte@ipp.mpg.de>; 'Oliver Schmitz' <oschmitz@wisc.edu>; 'Ralf Konig' <ralf.koenig@ipp.mpg.de>; 'Sehyun Kwak' <sehyn2000@gmail.com>; 'Stepan Serdeda' <stepan.sereda@ipp.mpg.de>; 'Thomas Sunn Pedersen' <tspe@ipp.mpg.de>; 'Valeria Perseo' <valeria.perseo@ipp.mpg.de>; 'Victoria Winters' <victoria.winters@ipp.mpg.de>; 'Wang, Erhui' <e.wang@fz-juelich.de>; 'Xu, Shuai' <s.xu@fz-juelich.de>; yu.gao@ipp.mpg.de; 'Yuhe Feng' <yuhe.feng@ipp.mpg.de>; 'DIETER BOEYAERT' <boeyaert@wisc.edu>
Betreff: RE: [EXTERNAL] Detachment definitionI think Felix is right. There are certain features of exhaust that we want to achieve while detaching:
- Bening heat fluxes and reduced particle fluxes
- Enough recycling particles downstream, to provide a sufficient amount of neutrals at the pumping gap
- Keep the target plasma so cold that it keeps the impurity sources there at a minimum (<5 eV for tungsten).
In this sense, you can have detached plasma (e.g. small plasmas), which is not fulfilling all our needs. So maybe it is better to define features of good quality detachment and work in this direction.
Best,
Marcin
From: Felix Reimold <flr@ipp.mpg.de>
Sent: Montag, 11. April 2022 20:55
To: Glen Anthony Wurden <wurden@lanl.gov>
Cc: Thierry Kremeyer <thierry.kremeyer@ipp.mpg.de>; Marcin Jakubowski <marcin.jakubowski@ipp.mpg.de>; 'Arturo Alonso' <arturo.alonso@ciemat.es>; 'Arun Pandey' <arun.pandey@ipp.mpg.de>; 'David Bold' <dave@ipp.mpg.de>; 'David Ennis' <dae0005@auburn.edu>; 'Dhard, Chandra Prakash' <chandra.prakash.dhard@ipp.mpg.de>; 'Dorothea Gradic' <dorothea.gradic@ipp.mpg.de>; 'Erik Flom' <erik.flom@ipp.mpg.de>; 'Frederik Henke' <frederik.henke@ipp.mpg.de>; 'Georg Schlisio' <georg.schlisio@ipp.mpg.de>; 'Gregor Pechstein' <gregor.pechstein@ipp.mpg.de>; 'Kenneth Hammond' <khammond@pppl.gov>; 'Maciej Krychowiak' <maciej.krychowiak@ipp.mpg.de>; 'Matt Kriete' <kriete@auburn.edu>; 'Matthias Otte' <matthias.otte@ipp.mpg.de>; 'Oliver Schmitz' <oschmitz@wisc.edu>; 'Ralf Konig' <ralf.koenig@ipp.mpg.de>; 'Sehyun Kwak' <sehyn2000@gmail.com>; 'Stepan Serdeda' <stepan.sereda@ipp.mpg.de>; 'Thomas Sunn Pedersen' <tspe@ipp.mpg.de>; 'Valeria Perseo' <valeria.perseo@ipp.mpg.de>; 'Victoria Winters' <victoria.winters@ipp.mpg.de>; 'Wang, Erhui' <e.wang@fz-juelich.de>; 'Xu, Shuai' <s.xu@fz-juelich.de>; yu.gao@ipp.mpg.de; 'Yuhe Feng' <yuhe.feng@ipp.mpg.de>; 'DIETER BOEYAERT' <boeyaert@wisc.edu>
Subject: Re: [EXTERNAL] Detachment definitionHi Thierry,
I think there is no good common definition of detachment even in tokamaks that is universal. And there are many different flavors depending on what aspect is emphasized (particle vs power, complete vs. partial and power starvation vs. pressure loss). I think as long as one clearly states what is discussed this is alright. As for the degree of detachment this mostly used to describe the deviation of the target flux scaling from the high recycling scaling of the TPM (n**2). This is neither easy to measure (integral flux to target) nor is the expected scaling clear for W7-X.
In my mind detachment just a term to describe that the 'expected' power and particle fluxes to the target are decreasing and that volume losses are important.
As Glen states the operationally relevant aspect is just the heat flux density limit (10 MW/m2) and the sputtering threshold (5eV for W) as a target survival constraint. In contrast to the tokamaks we need to look more closely to the neutral pressure in the divertor for the required particle exhaust.
Frad will indeed characterize the power aspect of it as long as Te, target > 2.5 eV. The particle flux reduction for us is not a critical aspect so far and would even is even detrimental for pumping. Hence, I think that frad is currently the most relevant detachment quantifier (with detachment being complete - i.e. no hot spots).
Felix
BlueMail for Android herunterladen
Am 11. Apr. 2022, um 19:34, "Wurden, Glen Anthony" <wurden@lanl.gov> schrieb:
In a tokamak reactor design, the predicted heat loading for a plasma attached to the divertor is thought to be too high. Achieving a stable detached state (ie, lowered heat load in the divertor) is thought to be essential.... especially for characteristically narrow strike lines. So, yes, the goal of detachment is to reduce the heat loads (W/m^2) to a tolerable level at the divertor. At the same time, sufficient edge densities are required so that the average energy of the impacting particles is low enough to have a tolerable residual sputtering levels. Having a "too hot" edge plasma with a low overall heat load is therefore also not desired.
So two points are needed...... say a 10x reduction in peak divertor strike line heat loads, and at the same time, high enough SOL densities to keep sputtering under control (high Z impurities should therefore be avoided, not just because of their increased radiation losses, but because of their ability to sputter).
Glen
From: Thierry Kremeyer <thierry.kremeyer@ipp.mpg.de>
Sent: Monday, April 11, 2022 8:51 AM
To: 'Marcin Jakubowski'; Wurden, Glen Anthony; 'Arturo Alonso'; 'Arun Pandey'; 'David Bold'; 'David Ennis'; 'Dhard, Chandra Prakash'; 'Dorothea Gradic'; 'Erik Flom'; flr@ipp.mpg.de; 'Frederik Henke'; 'Georg Schlisio'; 'Gregor Pechstein'; 'Kenneth Hammond'; 'Maciej Krychowiak'; 'Matt Kriete'; 'Matthias Otte'; 'Oliver Schmitz'; 'Ralf Konig'; 'Sehyun Kwak'; 'Stepan Serdeda'; 'Thomas Sunn Pedersen'; 'Valeria Perseo'; 'Victoria Winters'; 'Wang, Erhui'; 'Xu, Shuai'; yu.gao@ipp.mpg.de; 'Yuhe Feng'; 'DIETER BOEYAERT'
Subject: [EXTERNAL] Detachment definitionDear Detachment colleagues,
How do we define detachment at W7-X?
Even though we’ve started this discussion multiple times, we haven’t agreed on a clear definition, yet.
After this question came up again today in the “SOL journal club”, where we currently discuss Yuhe’s 2021 paper, perhaps we can start this discussion one last time.
I would like to collect your feedback in this regard via email, before we then discuss this in the detachment meeting on April 21st
Background:
Stangeby, Plasma Boundary Book IoP 2001:
“We are therefore interested in finding a way to decrease the particle flux
reaching the target, for given upstream conditions in the SOL. Such a state has
been found and is called divertor detachment. […]
The characteristic, in fact the defining, observations of detachment are:
• ‘roll-over’ and decrease of the ion saturation current, j+sat, of Langmuir
probes built into the inner and outer divertor targets, (d), (f), i.e. the
‘nt -rollover’ seen on ASDEX, figure 16.4;
• the Dα radiation from the target regions does not roll over/decrease, but
continues to increase with ne, (e), (g).
· A further feature of detachment is that it occurs
when the target Langmuir probes indicate low temperatures, Te ≈ a few eV or
less [16.23, 16.30]”
So according to this, in Detachment we get less particles neutralizing on the target, but the amount of particles re-ionizing, for which H_a is a proxy, continues to increase with n_e, while we get lower temperatures on the strike line from the Langmuir probes.
In practice the applied definitions differ a bit, but commonly used is a decrease of peak j+sat at the Langmuir probes and a peak temperature below 10 eV or even 5 eV.
In his 2018 publication, Leonard divides detachment in momentum, particle, and energy detachment,
and links each kind of detachment with which experimental observations are possible in tokamaks.
https://iopscience.iop.org/article/10.1088/1361-6587/aaa7a9/pdf
Requirements:
For W7-X we need a robust and practical definition. It has to be based on one or multiple metrics that are applicable to all magnetic field configurations. A definition based on Langmuir probes is difficult to apply, if the strike line peak is not covered.
Point 2 of Stangeby’s detachment definition is not fulfilled at W7-X detachment, as our H_a radiation in the divertor actually decreases by 35% when we detach.
I also believe that the detachment question should not be answered digitally in a “yes or no” fashion, but that we need an analog value, as I understand our detachment more as a spectrum. I guess in tokamaks this has been described as the “degree of detachment”.
When looking at this spectrum, we have the plasma in direct contact with the divertor target plates under fully attached conditions where all the power arrives at the target. As we transition into detachment, more and more power is radiated away before particles hit the targets. Heat and particle loads are reduced. The further the radiated power increases, the more the ionization front shifts poloidally inward. There is less plasma in the SOL, the SOL becomes more and more transparent for neutrals, until we eventually have our “tiny plasmas” at the other end of the spectrum.
Due to this, many people have been using the fraction of radiated power f_rad as a metric to differentiate these different scenarios.
I see many advantages in using it, as it is readily available, covers the entire spectrum from “all heat load on the target” to “no heat load on the target”, and is applicable to any magnetic field configuration.
Your opinion:
So what do you think?
Is there another reason besides heat load reduction, why we need detachment in a reactor?
What is important for our detachment definition?
What are our key requirements?
What is practical?
Which metric covers our entire spectrum of “degree of detachment”?
Do we need to start separating momentum, particle and energy detachment already in our base definition or should that come a step later?
Best,
Thierry
_________________________________
Dr. Thierry Kremeyer
Max Planck Institut für Plasmaphysik
Stellarator - Rand - und Divertorphysik (E4)
7.0-013
Wendelsteinstr. 1
17491 Greifswald
Germany_________________________________
Tel +49 3834 88 1496
+49 3834 88 2377
Mail thierry.kremeyer@ipp.mpg.de
-