FIRE Divertor Design

March 31, 2004

M. UlricksonPresented at the FIRE Physics Validation Review

Sandia is a multiprogram laboratory operated by Sandia Corporation, a Lockheed Martin Company,for the United States Department of Energy’s National Nuclear Security Administration

under contract DE-AC04-94AL85000.

MAU 2 5/20/2003

Outline

• Overview• Plasma flux structure• Plasma Heat Loads (Results from T. Rognlien)• Design Description (Dan Driemeyer, Boeing)• Disruption Analysis (Input from C. Kessel)• Main Issues to be addressed• Summary

MAU 3 5/20/2003

OverView

MAU 4 5/20/2003

Power Balance Assumed

• Heating Power: 30 MW• Plasma Gain: 5• Fusion Power: 150 MW• Alpha Power: 30 MW• Total Plasma Heating: 60 MW• Core Radiation: 10% 6 MW• Power to SOL: 54 MW (split evenly up/down)• Radiation in the SOL: 20% 11 MW• Radiation in divertor:10% 4 MW• Total Power to the Divertor plates: 40 MW

MAU 5 5/20/2003

Overview

• We assumed a double null divertor because:– We can use active heat load balancing (active cooling)– There is some evidence that double null mitigates ELMs– The active area of the divertor is increased, lower heat

loads (see Rognlien results)• The increase from 2 to 2.14m was done without a

proportional increase in height which meant the distance from the x-point to the plates decreased, but the heat is spread more.

• The plasma current increased which may make the eddy currents worse (see Disruption section)

MAU 6 5/20/2003

Plasma Flux Geometry in FIRE

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

1.8

2

1 1.2 1.4 1.6 1.8 2 2.2 2.4 2.6 2.8 3

R (m)

Z (m

)

outxpinxpout05out10out20out30out40in05in10in20innerbaffleouter

MAU 7 5/20/2003

Divertor Heat Flux

MAU 8 5/20/2003

UEDGE is a plasma/neutral fluid code

• Features of UEDGEPhysics:– Multispecies impurities; var. n, u||, Te,i, φ– Flux-limited kinetic corrections– Reduced Navier-Stokes neutrals or Monte Carlo

coupling– Multi-step ionization and recombination; sputtering

MAU 9 5/20/2003

FIRE is designed for a double-null divertor

Doublenull

Single null variant

• FIRE divertor must tolerate 28 MW into the SOL (DN)

• For 150 MW fusion power, helium must be removed at a rate of 5x1019 particles/sec

• Edge density is set to 3x1020 m-3

• Unity recycling with PF pumping

MAU 10 5/20/2003

UEDGE Modeling Results

• Core-edge temperatures are inconsistent with that required for good core confinement

• A single-null FIRE variant has more than 2 times the peak heat flux of the double-null

• Neon injection can induce partial detachment• Helium pumping in the private flux region

appears adequate• Peak power scales nearly inversely with density

and with anomalous diffusion coeff.• Midplane profiles show scaling with core-edge

density and transport coefficients

MAU 11 5/20/2003

Power Balance is Sensitive to DN Balance

• UEDGE now treats range from single-null to double-null

• Double nulls reduce peak heat flux, but balance is delicate

• ExB flows and currents produce asymmetries

• Code reproduces measured DIII-D heat flux imbalance

drSEP - distance between two separatrices at outer midplane

DIII-D heat-flux asymmetry betweenupper & lower divertors - T. Petrie

- 0.5

0.0

0.5

1.0

- 4 - 2 0 2 4drSEP (cm)

ion ∇B driftdownward

- 1.0

UEDGE

(q - q )up low

(q + q )up low

MAU 12 5/20/2003

A number of uncertainties remain

• Peak heat flux scales inversely with unknown anomalous transport coeff.

• Reabsorption of hydrogen radiation at high densities & stability of detachment

• Maintaining double-null power balance• Redeposition/removal of beryllium to/from

surfaces• Consistency of pedestal temperature with good

core confinement• Size and impact of ELMs

MAU 13 5/20/2003

SOL Heat Loads from UEDGE

0

2

4

6

8

10

12

14

-0.04 -0.02 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16

Distance Along Plate (m)

Heat

Flu

x (M

w/m

^2)

Flux InFlux Out

MAU 14 5/20/2003

Heat Loads on the Divertor

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7Distance Along Plate (m)

0

2

4

6

8

Hea

t Flu

x (M

W/m

2 )

0

100

200

300

400

500

Tem

pera

ture

(C)

Outer FluxOuter TInner FluxInner T

FIRE Divertor Flux and Temperature

MAU 15 5/20/2003

Engineering Design of FIRE Divertor (Boeing)

MAU 16 5/20/2003

FIRE Divertor Design

Pumping Slot

Water Supply

Vessel Attachments

Coolant Pass to Inner

MAU 17 5/20/2003

Disruption Cases

MAU 18 5/20/2003

Disruption Modeling with TSC

• Data provided by C. Kessel, PPPL• Two Scenarios considered:

– Vertical disruption with maximum Ip dot– Vertical disruption with maximum halo currents

• Current versus time was provided for about 900 filaments representing the plasma

• This plasma model was input to OPERA to drive a transient eddy current calculation

• 1/16th of FIRE was modeled in OPERA

MAU 19 5/20/2003

Disruption Cases from TSC

1.00E+04

1.00E+05

1.00E+06

1.00E+07

0.61 0.62 0.63 0.64 0.65 0.66 0.67 0.68

t (sec)

Cur

rent

(A)

IPFastIHaloMax

MAU 20 5/20/2003

Idot in Disruption Cases

-4.00E+09

-3.00E+09

-2.00E+09

-1.00E+09

0.00E+00

1.00E+09

2.00E+09

0.615 0.620 0.625 0.630 0.635 0.640

t (sec)

dI/d

t (A

/s)

I Halo MaxIpfast

MAU 21 5/20/2003

OPERA Model Of FIRE

MAU 22 5/20/2003

OPERA Model of FIRE Sector

MAU 23 5/20/2003

Opera Model of FIRE Divertor

MAU 24 5/20/2003

Induced Currents in IPmax VDE

MAU 25 5/20/2003

Induced Currents in IPmax VDE

MAU 26 5/20/2003

Induced Currents in IPmax VDE

MAU 27 5/20/2003

Induced Currents in IPmax VDE

MAU 28 5/20/2003

Results of OPERA Model

• In some small regions of the new divertor components the induced forces are about 2 times higher than for the 2 m design.

• On average the induced forces are less than 20% higher than for the 2 m design.

• Since the moments and loads on the supports depend sensitively on the location of the forces, we cannot tell whether the design has adequate margin for disruptions.

• The transient nature of the forces must be taken into account in detailed design.

MAU 29 5/20/2003

ELM Effects

MAU 30 5/20/2003

Estimate of ELM Heat Loads

• Best understanding of ELM energy losses at Snowmass indicated 3.6 MJ in a typical Type I ELM

• All of the ELM energy goes to the outer divertor• The effective area of the outer divertor is 2.4 m2

• The energy deposition is 1.5 MJ/m2

• The melting threshold is between 0.5 and 1.5 MJ/m2 depending on the ELM duration (0.1 or 1.0 ms)

• Type I ELMs are a life limiting event for the divertor.

MAU 31 5/20/2003

Type II ELMs

• At high density (nped >70% nGR), ELMs losses can become purely convective with ∇TELM~0

• Conditions of access vary: high δ is required (possibly q95 >3.5)

• High βp (JT60-U) and proximity to DN (ASDEX-U, JT60-U?)

• Type II ELMs in ETB H-modes so far observed for pedestal parameters near the Type I-III transition (ASDEX-U, and mixed ELM regime in JET and DIII-D QDB)

• DIII-D Locked Mode coils reduce ELM size without affecting confinement

From Saibene, EPS

MAU 32 5/20/2003

High Triangularity Effect on ELMs

3 4 5 6 70.1

0.2

0.3

0.4

0.5

0.6

JT60-U

JET (high Pin)

ASDEX-U

JET

<δ >

q95

, Type I (high βp & std)

, Type II & mix (high βp)

Type III (std)

From Saibene, EPS

MAU 33 5/20/2003

Type II in ASDEX-U: Quasi DN configuration

3.0 3.5 4.0 4.5 5.00.25

0.30

0.35

0.40

0.45

0.50

ASDEX-U

q95

vs average δ

<δ >

q95

Type I Type III Type I+II Type II

-3.5 -3.0 -2.5 -2.0 -1.5 -1.0 -0.5 0.00.25

0.30

0.35

0.40

0.45

0.50

ASDEX-U

Type I Type III Type I+II Type II

<δ>

∆X separatrices midplane (cm)

•Type II require high δ (as JET and JT60-U) and ne

•Proximity to DN configuration is essential

From Saibene, EPS

MAU 34 5/20/2003

Allowed ELMs on FIRE

0.0 0.5 1.0 1.5 2.0Energy Density (MJ/m2)

0

1000

2000

3000

Tem

pera

ture

Ris

e (C

)

1.0 ms0.1 ms

4.5 MW/m2

9 MW/m2

Normal Heat Flux

2% ELM Loss

5% ELM Loss

MAU 35 5/20/2003

Remaining Open Issues

• The disruption forces on the new design need to be compared to the forces in the 2.0 m design. If there is a large difference the supports for the divertor will have to be refined. (will be done by 3/30)

• The effect of deposited Be on the performance of the W divertor need to be determined (upcoming TPE experiments).

• The pumping speed of the new configuration needs to be calculated (should be better than the old design because there are fewer obstructions.

MAU 36 5/20/2003

Remaining Open Issues

• Analysis of the divertor loads on the vacuum vessel needs to be done (local reinforcement may be needed).

• Thermal stress analysis for the new design needs to be done (heat loads are lower than the old design unlikely there is a problem). (Baxi is working on this)

• ELM erosion testing should be done (starting 3/31 at SNL).

MAU 37 5/20/2003

ELM Simulations on W Rods

• EB1200 at SNL has two independent e-guns• We will use one to apply steady state heat flux

(FIRE and ITER like) and the other (unscanned) to simulate ELMS (up to 1.5 MJ/m2 in <1ms)

• This testing will be conducted at a few Hz and result in thousands of ELMS in less than a days operation.

• Damage to the W surface will be monitored during cycling.

MAU 38 5/20/2003

ITER Overlap

• FIRE peak heat flux is very similar to the ITER peak heat flux

• The FIRE duration is long enough to require active cooling of the divertor (steady state is possible)

• The Be first wall is the ITER choice also (First Wall heat loads are similar).

• ITER is considering an all W option and a mixed W/Be option because of the continuing problem with T retention in C.

MAU 39 5/20/2003

Conclusions

• The divertor design has been updated to the new major radius (2.14 m)

• Heat flux to the divertor has been updated using UEDGE and the newest edge transport parameters.

• Heat flux is reduced because of lower fusion power, greater flux spreading, and increased transport.

• The divertor operating temperatures are reduced and it is not as necessary to use impurity radiation to spread out the power.

MAU 40 5/20/2003

Conclusions II

• Two disruption cases are being used to determine disruption forces: maximum current decay rate and maximum halo currents

• Analysis of disruption induced eddy currents is being conducted using the OPERA code.

• Because of the lower heat fluxes the margins for ELM heat deposition have been increased.