search

ukaea / PROCESS

PROCESS is a systems code at UKAEA that calculates in a self-consistent manner the parameters of a fusion power plant with a specified performance, ensuring that its operating limits are not violated, and with the option to optimise to a given function of these parameters.
https://ukaea.github.io/PROCESS/
MIT License
36 stars 11 forks source link

X-point radiator and ELM-free operation #589

Closed jonmaddock closed 1 year ago

jonmaddock commented 7 years ago

In GitLab by @mkovari on Nov 27, 2017, 11:51

DEMO may have a high radiation zone just inside the separatrix near the X-point, known as an X-point radiator. This is not consistent with a one-dimensional model, of course. It may be possible to insert it by hand in an ad hoc way.

In AUG the X-point radiator has been observed to radiate up to one third of the heating power. The location of the X-point radiator can be controlled in real time by varying the heating power or the seeding level. Remarkably, ELMs disappear when the location of the radiator is more than 7cm above the X-point, virtually eliminating the temperature pedestal. Experiments have simultaneously achieved full detachment and the following parameters, with nitrogen fraction in the confined plasma ~ 2 – 2.5 %. This mode is promising for reactor operation (with an impurity other than nitrogen, probably).

The X-point radiator is a strong poloidal asymmetry, and raises questions about the accuracy of any models based on 1D transport such as Plasmod. However, it lies entirely outside the 95% flux surface, so it does not need to be included in a core transport model.

The X-point radiator will have the following effects.

  1. It will impose a large heat load on the bottom part of the main chamber, and on nearby parts of the divertor assembly.
  2. It will allow a large radiation fraction even with low Z impurities, because the local electron temperature is strongly reduced to a regime in which the seeded impurity is able to radiate efficiently. Effectively it creates an additional low temperature zone, not taken into account by the 1D models.

Bernert, M., et al. Power exhaust by SOL and pedestal radiation at ASDEX Upgrade and JET (2017). See also ‘X-point radiation and detachment control at ASDEX Upgrade’, talk presented at 3rd IAEA Technical Meeting on Divertor Concepts. Vienna, 2019. Stuart Henderson at UKAEA has been involved in this work.

Comments @jmorris-uk @ajpearcey @skahn @stuartmuldrew ?

jonmaddock commented 7 years ago

In GitLab by @hlux on Nov 27, 2017, 11:54

FYI: The X-point radiator is already taken into account in the default values of the peaking factors for the radiation wall load estimate. (The arguments are captured in Ronny's wall loading paper.)

I think before asking what parameters we would need to specify I would ask which models to we want to get effected by the X-point radiator?

jonmaddock commented 7 years ago

In GitLab by @mkovari on Nov 27, 2017, 13:48

The X-point radiation might affect the overall power balance and the separatrix power pdivt if it is strong enough.

jonmaddock commented 7 years ago

In GitLab by @hlux on Nov 27, 2017, 14:32

Yes, it is a very good question where you take it into account in your power balance. For the radiation wall loading work, I have associated it to the SoL radiation instead of the radiation inside the separatrix. But that was simply because our model was only calculating the symmetric radiation inside the separatrix and hence all other parts were additional. My impression is though that even the most recent experiment are not able to give you any predictions of X-point radiation. So we can include placeholders to estimate uncertainties, but a proper model is not possible at the current stage.

jonmaddock commented 7 years ago

In GitLab by @hlux on Nov 27, 2017, 14:33

Also it is highly unclear to me whether this would affect Psep as we use that for the L-H threshold etc. Maybe worth chatting to Matthias Bernert/ Arne/Marco Wischmeier unless Mattia already has some comments.

jonmaddock commented 7 years ago

In GitLab by @hlux on Nov 28, 2017, 11:32

Here is a copy of the e-mails exchanged to keep a clear record in the issue:

Mattia wrote: 

Dear Michael, Hanni,

For some reason I cannot reply on GitLab to the discussion about X-Point radiation – although I can read it. I’ll try to solve the problem in the following.

Talking to Marco or Matthias is always a good idea.
However, as Arne was mentioning, the situation is quite unclear about the X-point radiator. I think that the peaking factor in the radiation distribution is the best we can do at this stage. A previous calculation of Bernhard Sieglin was in any case indicating such clusterization not to be an issue for the DEMO wall.

Concerning PLH, I don’t think the X-point radiator will be such an issue. In DEMO, the problem is to keep Psep low enough, not high enough. If some power is lost through the X-point radiator, it is sufficient to reduce a bit the Xe-in the core, and should be fine. Also – and this is surprising – there is no experimental evidence of lethal repercussions of the X-point radiator on the overall stability.

By the way, an X-point radiation ensures a well-detached divertor, thus it is conservative not to take it into account.

Clearly, if you want to interact with the expert, I’ll be pleased to be involved (not because I don’t trust you, but because I’m interested in the topic as well :) ).

Have a nice afternoon – and apologize again for not using GitLab,

Cheers

Mattia

Michael wrote:

Nevertheless, if the X-point radiation is a significant fraction of the total, then it should be included.
Of course, the same is true of radiation from the SOL, and of radiation from the edge of the pedestal.

All the best,

Michael
jonmaddock commented 4 years ago

In GitLab by @stuartmuldrew on Feb 6, 2020, 11:56

Hanni and I started to look at this with Marco Wischmeier. We had a meeting with Marco, Mattia and Emiliano; but I ran out of time to drive it on. We thought about combining it with Richard's L-mode DEMO work.

We didn't make any changes to the code but initially started playing around with the confinement to see what it would do to DEMO. Marco was of the view it could be as low as H~0.7-0.8. Palermo et al. (2019) do something very similar with ASTRA, but for a different reason, which made me question if the PROCESS version was needed.

I've attached the scans I did: DEMOconfinement.pdf

I'm happy to go back to Marco to discuss this if someone has time to do the work.

jonmaddock commented 4 years ago

In GitLab by @ajpearcey on Aug 14, 2020, 16:25

Michael, you say we could add something in an ad hoc to model these x-point radiators, I think it's possibly worth exploring the model presented in Sensitivity of detachment extent to magnetic configuration and external parameters

The paper gives a 1D model of the SOL from outer mid plane to the divertor target which is used to compute position of a thermal front region, which contains the largest contribution to emitted radaiton. Equations 18, 24 and 26 give us all the ingredients needed to implement this as a constraint in PROCESS. Of course this wouldnt cover the physics of this radiating region inside the seperatrix, but it could help generate PROCESS soultions that capture the expected detached divertor conditions required.

jonmaddock commented 2 years ago

In GitLab by @wg7580 on Jul 21, 2022, 16:18

Hi @mkovari Could you assign this to someone who might have a good idea as to how to solve/progress this forward- thanks

jonmaddock commented 2 years ago

In GitLab by @mkovari on Jul 22, 2022, 08:32

As far as I can see the X-point radiator is not being considered as an attractive reactor option. I suggest we close this issue. Any comments @ajpearcey @wg7580 ?

jonmaddock commented 2 years ago

In GitLab by @mkovari on Jul 22, 2022, 08:32

closed