Neutral recirculation the key to control of divertor operation A.S. - - PowerPoint PPT Presentation

Neutral recirculation  the key to control

• f divertor operation

A.S. Kukushkin1,2, H.D. Pacher3

1Kurchatov Institute, Moscow, Russia 2NRNU MEPhI, Moscow, Russia 3INRS‐EMT, Varennes, Québec, Canada

Presented at the 1st IAEA Technical Meeting on Divertor Concepts, Vienna, 29 Sept – 2 October 2015

Introduction

A.S. Kukushkin, H.D. Pacher. 1st IAEA Tech. Meeting on Divertor Concepts, Vienna, 29 Sept – 2 Oct. 2015

Reactor plasma controls: Force balance (pressure, magnetic field, currents) Power balance (heating, radiation, transport) Particle balance (fueling, pumping) Fueling affects all the controls Pressure: p = n∙T Fusion reactions: Pfus  nD∙ nT∙F(T) Radiation: Prad  ne∙ nI∙L(T) Transport: n < ∙nG limitation For steady state, fuel = pump (neutrals!)  Neutral recirculation affects everything (unless Li walls are employed!)

2

Introduction

A.S. Kukushkin, H.D. Pacher. 1st IAEA Tech. Meeting on Divertor Concepts, Vienna, 29 Sept – 2 Oct. 2015

Divertor functions: Impurity screening (plasma‐surface interaction further from the core) Pumping (neutral compression) Improving core confinement (experimental observation, unexpected) Side effect: Power flux concentration on targets  severe operational constraint  Need to control  neutrals! ITER design studies 1995‐2014: neutral transport in the edge is the most important issue This presentation: review of those studies

3

SOLPS4.3 – computational model

A.S. Kukushkin, H.D. Pacher. 1st IAEA Tech. Meeting on Divertor Concepts, Vienna, 29 Sept – 2 Oct. 2015

SOLPS4.3 development: primary attention to neutral transport Reasonably simple plasma model ‐ no drifts, no currents in the SOL ‐ flux‐limited  and  , constant  , D Monte‐Carlo model for neutrals ‐ m‐i collisions, including c‐x (“MAR”) and elastic ‐ n‐n collisions ‐ elastic He0 – D+ collisions

4

puff to pump

 important for ITER (also for C‐mod [S. Lisgo, JNM 2003] ) Scope: SOL & divertor plasma, target loading, particle fluxes, pumping, … Interaction with the core: feedforward PSOL, core, Prad_core (for surface loading) Methodology: density scans, matching detachment state for comparisons

FED‐2012

Core‐SOL model integration

A.S. Kukushkin, H.D. Pacher. 1st IAEA Tech. Meeting on Divertor Concepts, Vienna, 29 Sept – 2 Oct. 2015

Direct coupling impractical – time scales too different “Mediated approach”: parameterization of SOLPS results  BC for the core model  Core solutions compatible with the edge Edge‐imposed restrictions for the core  operational window for whole plasma Realisation: ICPS model based on 1D ASTRA code for the core [G.W. Pacher et al.] Scope: operational flexibility, effect of divertor variations and operation mode

5

Methodology: Vary core inputs and adjust puffing rate to keep qpk  qmax , find the area in parameter space satisfying all known physics and design constraints

NF‐2003 PSI‐2010

Edge and core fueling: practically independent

A.S. Kukushkin, H.D. Pacher. 1st IAEA Tech. Meeting on Divertor Concepts, Vienna, 29 Sept – 2 Oct. 2015

6

Gas puff efficient for qpk control  Nearly independent controls for

• Core: pellets
• Divertor: gas puff

To be tested in experiment (ITER?) Indications from C‐mod

PSI‐2014

High‐power, big size machine (e.g. ITER): Increase of the puffing rate  saturation of nsep at rather low level depending on PSOL  strong neutral screening in the SOL  low neutral influx to the core  poor core fueling  Additional means of core fueling are needed (pellets)

Power dissipation in divertor

A.S. Kukushkin, H.D. Pacher. 1st IAEA Tech. Meeting on Divertor Concepts, Vienna, 29 Sept – 2 Oct. 2015

Divertor detachment: the path to reach acceptable qpk Energy dissipation disproportionately high in near SOL Radiation loss  n2  effect of neutral fueling from PFR

• Stronger with high pn

Longer dome (i‐n contact area ): higher qpk for same pn

7

To Pump

Dome To Pump

PSI‐2012, PSI‐2006

 higher pn needed Detachment (Isat rollover) at the same pn  Narrower window in pn (or qpk) if limited by detachment (core confinement)

Overfuelling: discharge collapse

A.S. Kukushkin, H.D. Pacher. 1st IAEA Tech. Meeting on Divertor Concepts, Vienna, 29 Sept – 2 Oct. 2015

Typical in detachment modelling: Isat rollover as puffing rate increases (pn goes up) Then at some point the edge plasma collapses

8

baffled normal

• no flux to targets
• recycling inside separatrix
• Power loss and recombination zones

also move inside the core Stronger baffling delays the collapse  Collapse is caused by neutrals escaping from the divertor to the core  Neutral re‐circulation determines the limits of stable divertor detachment

PSI‐2014

Impurity seeding: Increase of divertor radiation  reduction of nsep and qpk Prerequisite for qpk: high density (fueling!) OK for divertor However, detrimental for the core plasma performance at high Q

Gas puffing vs. impurity seeding

A.S. Kukushkin, H.D. Pacher. 1st IAEA Tech. Meeting on Divertor Concepts, Vienna, 29 Sept – 2 Oct. 2015

9

Integrated modelling: Operational window decreases in Q as cNe grows (more Paux is needed)  puff with minimum seeding (to control nsep ) is preferable for qpk control

PSI‐2014

In‐out asymmetry

A.S. Kukushkin, H.D. Pacher. 1st IAEA Tech. Meeting on Divertor Concepts, Vienna, 29 Sept – 2 Oct. 2015

Neutral exchange between the inner and outer divertors: Outer divertor: more power (geometry effect) Inner divertor: cooler plasma, higher density  more neutrals neutrals flow from inner to outer divertors, reducing the asymmetry

10

EPS‐2005

When this neutral flow is impeded (reducing gas conductivity beneath the dome)  Stronger in‐out asymmetry, higher qpk (outer)  Transparency of dome supporting structures important

Transparency 

Role of divertor dome

A.S. Kukushkin, H.D. Pacher. 1st IAEA Tech. Meeting on Divertor Concepts, Vienna, 29 Sept – 2 Oct. 2015

11

PSI‐2006, PSI‐2010

0.01 0.1 1 0.1 1

353_iDomGrid_A_s_z10_2

Epump_He [eV] 

Becomes critical in ITER

• r reactor: DT > min, He Pfus

Major effect: cooling down the neutrals thus condensing the flow From previous slides: Reduction of the dome furthers detachment Free inner‐outer neutral exchange beneficial weaker asymmetry, lower qpk Why do we need the dome at all? – For pumping! If dome were removed, need a factor 10 higher pumping speed to keep the same He upstream or the same fuel throughput

1 2 4 6 8 10 30

366_iLk4C_z2_5

qpk [MW/m ] m=-1.17 pDT [Pa]

qpk [MW/m2]

10

Effect of neutral leaks in structures

A.S. Kukushkin, H.D. Pacher. 1st IAEA Tech. Meeting on Divertor Concepts, Vienna, 29 Sept – 2 Oct. 2015

Previous slides: divertor structures assumed vacuum tight. What if they leak? Simplified model with leaks: pv >> pc >> pb ; pd > pc  qpk the same  He removal better: leak (v) enhances pumping, He from bypass (b) recycles further from the separatrix  inter‐cassette gaps acceptable

12

FEC‐2006

return flo w to pump throughput

b v

d v b b n n i i p

Cassette body

pd pc pv pv pc pc pc pc pb pb v c

Conclusions

A.S. Kukushkin, H.D. Pacher. 1st IAEA Tech. Meeting on Divertor Concepts, Vienna, 29 Sept – 2 Oct. 2015

Conditions of neutral re‐circulation in tokamak affect virtually all plasma controls SOLPS4.3 integrated with ASTRA  edge (neutrals) effect on total performance Lessons learnt from modelling and implemented in ITER design

• Core and edge fueling schemes are ~ independent  separate control possible
• Neutral penetration from PFR provides detachment control even if q ~ 1 mm
• Neutral confinement in divertor determines stability of detachment
• Gas puffing for divertor control

is better for overall performance than impurity seeding

• Free neutral exchange between divertors reduces in‐out asymmetry (qpk )
• Divertor dome is essential for efficient pumping (neutral compression)
• Neutral leakage through divertor structures does not degrade performance

 does not affect divertor control via neutral re‐circulation

13

Disclaimer

A.S. Kukushkin, H.D. Pacher. 1st IAEA Tech. Meeting on Divertor Concepts, Vienna, 29 Sept – 2 Oct. 2015

In this paper, we use results of calculations done previously for ITER conditions and published elsewhere (see, e.g., the references in the paper) to discuss more general considerations resulting from those studies. We make neither recommendations for, nor assessments

• f, ITER here and our conclusions imply no liability

for the authors, nor for their affiliated organizations.

14