Edge Plasma and Divertor Issues in DEMO-FNS Project V.Yu. Sergeev 1, - - PowerPoint PPT Presentation

1/20 1st IAEA Technical Meeting on Divertor Concepts, 29 Sept – 2 October 2015, Vienna, Austria

Edge Plasma and Divertor Issues in DEMO-FNS Project

V.Yu. Sergeev1, B.V. Kuteev2, A.S. Bykov1, V.G. Skokov1, A.A. Gervash3, D.A. Glazunov3, P.R. Goncharov1, A.Yu. Dnestrovskij2, R.R. Khayrutdinov2, A.V. Klishchenko2, V.E. Lukash2, I.V. Mazul3, P. A. Molchanov1,V. S. Petrov2, V.A. Rozhansky1, Yu.S. Shpansky2, A.B. Sivak2, A.V. Spitsyn2

1Peter the Great St. Petersburg Polytechnic University 2National Research Center “Kurchatov Institute”, Moscow 3D.V. Efremov scientific research institute of electrophysical apparatus, St. Petersburg V.Sergeev@spbstu.ru

2/20 1st IAEA Technical Meeting on Divertor Concepts, 29 Sept – 2 October 2015, Vienna, Austria

Challenge:

As simple as As compact as



possible divertor design which provides consistent with stable & steady-state divertor operation reduced heat loads and erosion of solid PFM & high-quality ELM free H mode of plasma core withstanding neutron irradiation

3/20 1st IAEA Technical Meeting on Divertor Concepts, 29 Sept – 2 October 2015, Vienna, Austria

Outline

• 1. DEMO-FNS basic parameters
• 2. Concept of power and particle exhaust
• 3. PFCs design and choice of materials
• 4. FW mock-up testing
• 5. Modelling of DEMO-FNS operation
• 6. Summary

4/20 1st IAEA Technical Meeting on Divertor Concepts, 29 Sept – 2 October 2015, Vienna, Austria

DEMO-FNS basic parameters

[Kuteev B.V. et al 2015 Nucl. Fusion 55 073035] Plasma major radius R (m) 2.5÷2.75 Aspect ratio R/a 2.5÷2.75 Elongation κ

elong

2.1 Triangularity δ 0.5 Plasma current Ip (MA) 5 Magnetic field at geometric axis Bt (T) 5 Volume averaged electron density <n> (1020 m−

3)

1 Neutron load to FW (MWm−

2)

≤0.2 Beam energy Eb (keV) 500 External power PEC+PNBI, МW 6+30 Fusion power PFusion (MW) ≤40 H =τE/τIPB98(y,2) factor 1.0-1.4 Beta normalized β

N

< 3 Non-inductive current fraction 1 First wall area Sw (m2) 162÷188 Plasma volume Vpl (m3) 103÷113 Lifetime (year) 30 Duty factor 1/3

5/20 1st IAEA Technical Meeting on Divertor Concepts, 29 Sept – 2 October 2015, Vienna, Austria

Concept of power and particle exhaust

Ø A Double Null (DN) divertor configuration is chosen since: 1)

Decrease of the divertor heat flux up to1.5 times [Pitcher C.S. and

Stangeby P.C. 1997 PPCF 39 779]

2)

“Good coupling of the radiating divertor with high performance H- mode plasma” [Petrie T.W. et al 2008 Nucl. Fusion 48 045010]

3)

Seems to be more favorable due to more symmetric loads at transients (ELMs, VDE, start-up, ramp-down) [Petrie T.W. et al 2003

• Nucl. Fusion 43 910]

Ø) V-shaped long outer divertor leg is used: 1)

Easier for detachment [Kukushkin A.S. et al 2011 Fusion Eng. and Des. 86

2865; Asakura N. et al 2010 J. Plasma Fusion Res. SERIES 9 136]

2)

Better for coupling with high performance H-mode plasma: “The 1/ ∼ R variation of the parallel heat flux into the detachment zone helps to stabilize the location of the thermal front—keeping it in the divertor volume” [LaBombard B. et al 2015 Nucl. Fusion 55 053020]

6/20 1st IAEA Technical Meeting on Divertor Concepts, 29 Sept – 2 October 2015, Vienna, Austria

Concept of power and particle exhaust(1)

[Greenwald M. et al 2007 Fusion Sci.

• Technol. 51 266]

Ø H factor plotted against outer gap,

showing that confinement does not deteriorate if the plasma wall spacing is larger than 2 to 3 mm. This distance roughly corresponds to the SOL width.

[ASDEX Team 1989 Nucl. Fusion 29 1959]

Ø

“Non-ideal” divertor configuration in which (3050%) of plasma power exhausting through separatrix is distributed onto the first wall is considered for DEMO-FNS [Sergeev V.Yu. et al 2015 NF 55 (accepted)].

Ø

Plasma-separatrix gap of DEMO-FNS should be 23 cm since q

• 5 mm [Eich T. et al 2011
• Phys. Rev. Lett. 107 215001] 5/2 q
• 1.3 cm of the SOL temperature and density e-folding

lengths and deviation SP ( 1 cm) due to ELMs at JET. Control system should provide accuracy of plasma position < 1 cm during steady-state operation.

Ø

Experiments with reduced plasma-wall gaps would be interesting on contemporary tokamaks (JET, Asdex Upgrade, D-IIID, TCV) where this is technically possible.

7/20 1st IAEA Technical Meeting on Divertor Concepts, 29 Sept – 2 October 2015, Vienna, Austria

Concept of power and particle exhaust(2)

Ø Impurity puff is being tested as a tool for obtaining the detachment

conditions in simulations [Kukushkin A.S. et al 2011 Fusion Eng. and Des. 86 2865;

Asakura N. et al 2010 J. Plasma Fusion Res. SERIES 9 136] and experiments on

JET [Giroud C. et al 2012 NF 52 063022], Asdex Upgrade [A. Kallenbach 2015 NF

55 053026], D-IIID [Petrie T.W. et al 2008 Nucl. Fusion 48 045010]. Local neon

puff is being considered for DEMO-FNS.

Ø ELMs are the concern of divertor and FW designs as well. According to

JET report [Loarte A. 2006 Proc. 7th ITPA Divertor and SOL Physics] 2000 Type I ELMs with energy content  1 MJ each caused erosion of  3 mm of Be wall which is unacceptable for the steady-state operation.

Ø External injection of lithium as pellets, droplets or liquid jets into the edge

plasma could control ELMs and/or H-mode regime as it was demonstrated using Li dust and granules injection in NSTX [Bell M.G. et al 2009 PPhCF 51

124054], EAST [Hu J.S. et al 2014 Fusion Eng. Des. 89 2875] and DIIID [Osborne T.H. et al 2015 NF 55 063018]. We rely on such external injection of lithium to

mitigate or better to suppress ELMs in DEMO-FNS.

8/20 1st IAEA Technical Meeting on Divertor Concepts, 29 Sept – 2 October 2015, Vienna, Austria

PFCs design

Sketches of PFC elements for DEMO-FNS: (1) heat carrier pipe (left) and plate (right) made of chromium–zirconium bronze, (2) sectioned beryllium coating, (3) brazing layer, (4) stainless-steel base (vacuum vessel), (5) pipe for heat carrier (water) flow.

2 3 1 4 Divertor (ITER like, pipe-type) First wall (thin, arched plate-type)

1 2 5 4 3 [Sergeev V.Yu. et al 2012 Plasma Phys. Rep. 38 521] [Sergeev V.Yu. et al 2015 NF 55 (accepted)]

Ø

ANSYS code simulations with heat transfer coefficient values of about 80-90 kW/(m2K) was done. It reveals that both designs can withstand heat loads up to 10 MW/m2.

Ø

Further development of the shaping procedure for FW design instead of the cutting procedure of the target is in progress, which will allow manufacturing curved surfaces of the first wall panels. 10 mm 6 mm 10 mm 16 mm 16 mm 3 mm

9/20 1st IAEA Technical Meeting on Divertor Concepts, 29 Sept – 2 October 2015, Vienna, Austria

Choice of materials

Ø

Evaluated temperatures satisfy temperature limitations described above for the 5 10 MW/m2 heat loads at Be, CuCrZr bulks and Be-CuCrZr interface.

Ø

For heat loads lower than 1 MW/m2 CuCrZr heat-sink material can be replaced to SS.

Ø

CuCrZr/Be/SS keep their properties at 5/10/30 dpa correspondingly. Neutron irradiation of PFC  20 dpa during the DEMO-FNS lifetime.

[Sergeev V.Yu. et al 2015 NF 55 (accepted) and references in]

10/20 1st IAEA Technical Meeting on Divertor Concepts, 29 Sept – 2 October 2015, Vienna, Austria

FW mock-up: design and manufacturing

Sketch (a) and photo (b) of mock-up of the PFC element for FW [Mazul I. et al. 2012

Fusion Eng. Des. 87 437]: (1) heat carrier plate made of chromium–zirconium bronze,

(2) sectioned beryllium coating, (3) brazing layer, (4) stainless-steel base (vacuum vessel), (5) pipe for heat carrier (water) flow.

1 2 5 4 3

11/20 1st IAEA Technical Meeting on Divertor Concepts, 29 Sept – 2 October 2015, Vienna, Austria

Mock-up: experiments on Tsefey- M

Ø Photo (a) and sketch (b) of test-bed experiments with the mock-up of the

PFC element on the electron beam facility Tsefey-M [M.Rodig et al 2000 FED

51–52 715] at Efremov Institute.

Ø The mock-up has 2 sets of Beryllium tiles: 3х16х16 mm – 18 pieces and

3х7х16 mm – 6 pieces.

12/20 1st IAEA Technical Meeting on Divertor Concepts, 29 Sept – 2 October 2015, Vienna, Austria

Mock-up: results of experiments

Ø Parameters of test-bed experiments

§ Area of heat loads

50 cm2

§ Range of heat flux 5-10.5 MW/m2 ± 5% § Cross section of water cooling 204 mm2 § Water velocity

7 m/s

§ Water rate

1.4 kg/s

§ Water pressure at the entrance

2 MPa ± 5%

Ø 1 stage: Power load was lasting 30 sec and was incremented up to 5 MW/m2

(1.0, 2.0, 3.0, 3.5, 4.0, 4.5, 5.0 MW/m2). Then, the cycle load (15 sec load, 15 sec pause) was done at 5.0 MW/m2. The mock-up had successfully sustained during 1000 cycles. The surface temperature was about 260 C.

Ø 2 stage: Power load was increased with a step of 0.5 MW/m2 to reach the

surface temperature of 600 C. It is occurred at 10.5 MW/m2. Then, the cycle load (15 sec load, 15 sec pause) was done at 10.5 MW/m2. The mock-up had successfully sustained during 100 cycles.

13/20 1st IAEA Technical Meeting on Divertor Concepts, 29 Sept – 2 October 2015, Vienna, Austria

Temperature fields of the mock-up

Ø Temperature (260 600 C) of tiles demonstrate the uniform radiation (IR) Ø No tiles lost a thermal contact at both at 5 MW/m2 (left – shot #120) and at

10.5 MW/m2 (right – shot # 95)

14/20 1st IAEA Technical Meeting on Divertor Concepts, 29 Sept – 2 October 2015, Vienna, Austria

Divertor cassette design

The DEMO-FNS divertor cassette design (a) and drawing with magnetic equilibrium map (b) : (1) outer divertor plate, (2) dome, (3) inner divertor plate, (4) cassette body, (5) outer target, (6) water cooling pipes, (7) outer pumping slot, (8) inner pumping slot, (9) vacuum vessel, (10) FW panels, (11) Ne injection.

15/20 1st IAEA Technical Meeting on Divertor Concepts, 29 Sept – 2 October 2015, Vienna, Austria

B2SOLPS5.2 simulations

[Rozhansky V.A. et al 2001 Nucl. Fusion 41 387]

Ø Lithium source in vicinity of

separatrix at middle plane

Ø Neon source at outer divertor

region

Ø Fluid description of neutrals Ø No drifts included

For sputtering, impact energy Eimpact=3Te+2Ti

[Stangeby P.C. 2000 The Plasma Boundary of Magnetic Devices (New York: Taylor and Francis), p. 641]

16/20 1st IAEA Technical Meeting on Divertor Concepts, 29 Sept – 2 October 2015, Vienna, Austria

Modeling of heat loads

Parameters I II III D+T puff rate (1022 part sec-1) 5.0 5.0 5.0 Li injection rate (1022 part sec-1) 0.8 6.5 Ne injection rate (1022 part sec-1) 0.3 <Zeff> at separatrix 1.5 1.3 2.0 Inner plasma-wall gap in (cm) 2 2 2 Outer plasma-wall gap out (cm) 2 2 5 Power balance Input power (MW) 40.0 40.0 40.0 Upper/Bottom outer DP (MW) 3.9 0.3 6.1 Upper/Bottom inner DP (MW) 0.6 0.3 0.8 Outer first wall (MW) 15.1 16.1 5.5 Inner first wall (MW) 5.1 5.1 5.0 Upper/Bottom private region (MW) 0.4 0.1 1.4 Total radiation power (MW) 10.0 17.4 12.9 OSP heat flux (MWm-2) 5.6 0.7 7.2

Ø

Lithium injection (Zeff 1.5) corresponds to OSP heat flux  8 MW/m2 (high- recycling regime with target plasma temperature  20 eV). Erosion problem (see below)!!!

[Sergeev V.Yu. et al 2015 NF 55 (accepted)] I II

Ø

Neon injection (Zeff 1.3) decreases heat flux to  0.7 MW/m2 (detached regime with low outer target plasma temperature  1 eV).

17/20 1st IAEA Technical Meeting on Divertor Concepts, 29 Sept – 2 October 2015, Vienna, Austria

Beryllium sputtering yield

ØMonte-Carlo calculations are reduced to reproduce experimental

• bservations

[Nishijima D. et al 2009 J. Nucl. Mater. 390–391 132] [J. Wesson, Tokamaks, 3d Ed, 2004, 467]

18/20 1st IAEA Technical Meeting on Divertor Concepts, 29 Sept – 2 October 2015, Vienna, Austria

Steady state erosion

Li (a) and Be (b) loss rate in DEMO-FNS. Solid curves ( 1.2 1020 cm-2s-1) at OSP, dash-dotted curves ( 7.2 1016 cm-2s-1) at FW middle plane. Gray shadow areas depict a range of Li Li (a) and Be Be (b) loss rates at FW middle plane and OSP.

Ø

Sputtering of Be at FW is low (< mm per year). D=>Be sputtering normally dominates in comparison with Be=>Be one due to bigger particle flux in comparison with impurity flux.

Ø

Sputtering of Be at OSP corresponds to  200 mm Be layer thickness per year!

Ø

Erosion of Li requires narrow temperature window due to ablation. Li does not help to protect Be at divertor if Li is injected externally. Either liquid lithium flow on divertor plate

• r detachment may help.

[Sergeev V.Yu. et al 2015 NF 55 (accepted)] in = 2 cm, out = 5 cm (run III, Li, high-recycling)

19/20 1st IAEA Technical Meeting on Divertor Concepts, 29 Sept – 2 October 2015, Vienna, Austria

Estimations of steady-state net Be loss rate

Be loss in DEMO-FNS regimes at 0.3 duty factor, mm/year High recycling regime at OSP with Li injection (I) (in = 2 cm, out = 2 cm) Detached regime at OSP with Ne injection (II) (in = 2 cm, out = 2 cm) Outer middle plane 0.17 0.01 Inner middle plane 0.10 ~0 Outer strike point 83 ~0 Inner strike point 19 12

ØAcceptable level of Be sputtering of FW in the middle plane region. ØHigh sputtering is modelled at SPs for high recycling regimes of divertor. ØDetachment of plasma at OSP with Ne injection reduces of sputtering

• essentially. At ISP is still high sputtering that requires Ne injection into

this region as well.

ØBe erosion during ELMs Type I is a concern. Evaluations of these

phenomena is in progress. For a moment, we rely on mitigation or suppression of ELMs.

20/20 1st IAEA Technical Meeting on Divertor Concepts, 29 Sept – 2 October 2015, Vienna, Austria

Summary

Ø

Double null, long external leg (~ R/2) with V-shape corner divertor is

• chosen. Small 23 cm gap between separatrix and FW is planed.

Ø

Modelling DEMO-FNS operation suggests that outer divertor targets has 7-9 MW/m2 load in the high-recycling regime. Up to 50% of input power is transferred to the first wall with heat load less than 1MW/m2. Experiments with reduced plasma-wall gaps would be interesting on contemporary tokamaks (JET, Asdex Upgrade, D-IIID, TCV) where this is technically possible.

Ø

To prevent steady-state Be erosion at divertor plates either the full detached

• peration of both outer and inner divertor or arrangement of the renewal

lithium flow on targets without degradation of core confinement should be investigated.

Ø

Beryllium erosion at divertor during transients is a major concern. Either mitigation or suppression of ELMs is being considered to prevent it.

Ø

Mock-up of the water cooled first wall element of DEMO-FNS with Be tiles was successfully tested. No tiles lost a thermal contact at both 5 MW/m2 (1000 cycles) and at 10.5 MW/m2 (100 cycles) were observed.