Design Point for a 1MW Fusion Neutron Source

P. E. Sieck1, S.Woodruff1, P. A. Melnik1, J. E. Stuber1,C.A.Romero-Talamas2, J. O’Bryan2, R. L. Miller3

1Woodruff Scientific Inc, 4000 Aurora Ave N, Suite 6, Seattle WA 98103

2University of Maryland Baltimore County, 1000 Hilltop Circle, Engineering 222Baltimore, Maryland 21250

3Decysive Systems, 813 Calle David Santa Fe, NM 87506

2016 US-Japan CT WorkshopAugust 24th 2016, Irvine, CA

Work supported by DARPA grant N66001-14-1-4044Work completed as part of the IAEA CRP on Compact Fusion

Neutron Sources

Motivation: IAEA CRP Compact Fusion n-Sources

I CRP 2014-1016 aiming forPn=1-100MW (1017 − 1019n/s)design points for wastetransmutation, fuel reprocessing.

I Tokamak, mirror, ST and other DTconcepts considered.

I Ours: magnetically compressedSpheromak with convergence CR

(=R0/Rf ) < 3.

I Based on prior results from SSPX[2] and S1 [3].

I Could readily fall into the ARPA-EIDEAS or ALPHA programs.

Motivation: high nTτ adiabatically with 1m system

I Historical: ATC [4], TUMAN-3M[5]

I Current: ARPA-E, private (FRCs, spheromaks and STs)

I Adiabatic means faster than τE .

Method: 0D, 2D and 3D time-dependent simulation

Tools:

I 0D analytic modeling

I 3D resistive MHD [6] analysis ofcompression to compare with0D analytic modeling

I CORSICA 2D equilibrium andstability code [7]

I Engineering design (CAD andforce/stress analysis)

I ARIES-like systems code writtenfor Matlab [1]

Method:

1. A 0D design point is defined

2. A CORSICA model is definedwith coil positions to supportequilibrium, and for formation.

3. NIMROD is used to test forstability, examine dynamics

4. Given coil currents fromCORSICA, a bank design isdeveloped.

5. Knowing the size of bank andcoils, an engineering designpoint is developed.

6. With all components defined, acost analysis is produced.

Results: 0D modeling for adiabatic compression

I P . V = N . R . T

I For processes that are adiabatic,(γ=5/3): P0 . Vγ0 = Pf . Vγf .

I If the compression is adiabatic, thenPV5/3 = const.

I and for self-similar compression (C =a0 / af )

I P0V0γ=Pf Vf

γ and since Vf =V0C3

then: Pf = P0 . C5 and since P ∼n.T and Tf = T0 . C2

Results: Device design point from 0D modeling

Parameter Symbol Value Units

Radial Convergence Cr 3Volumetric Convergence CV 27Initial Plasma Radius r0 0.5 mFinal Plasma Radius rf 0.166 m

Initial Volume V0 0.55 m3

Converged Volume Vf 0.16 m3

Initial Beta β0 15 %

Initial Density n0 2 ×1020m−3

Initial Magnetic Field B0 0.5 TInitial Temperature T0 244 eVFinal Temperature Tf 2196 eV

Final Density nf 54 ×1020m−3

Final Magnetic Field Bf 4.5 TInitial plasma current I0 0.75 MAFinal plasma current If 6.75 MAFinal Beta βf 45 %Final Magnetic Energy Uf 1.37 MJSteady State Fusion Power from neutrons Pn 1.27 MWSteady State Neutron Rate Γn 5.7E+17 n/s

Results: 3D MHD simulations of compression in cylindricalgeometry

E_tan applied

in bands along Z

I NIMROD simulation campaign with broad range of ICs, 20cmradius can, 5µs compression

I Full Braginskii (T-dependent) 2 fluid transport modelcalibrated to SSPX

Results: 3D MHD results follow 0D adiabatic scaling

2 4 6 8Volume Convergence

2´ 1020

4´ 1020

6´ 1020

8´ 1020

nd

1.5 2.0 2.5 3.0 3.5 4.0Area Convergence

0.1

0.2

0.3

0.4

0.5

bep

2 4 6 8 10 12Volume Convergence

5000

10000

15000

20000

pres

2 4 6 8 10 12Volume Convergence

20

40

60

80

100

tele

Results: CORSICA uncompressed, radial formation

I Injection from outboard side using solenoid and conicalannular electrodes (somewhat like S1)

I T0 = 200eV, and n0= 1e20m−3.

Results: CORSICA peak compression

I Plasma current increases to 6.75MA

I Fusion power of 1.3MW at 5keV

Results: Pressure optimization with shaping

I At peak compression, the shape of the plasma dictates theattainable beta (stable to Mercier (pressure-driven) modes).

Results: PSpice design point for coil banks

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

00

0

0

0

0

0

0

0

I

I

I

I

IC

20k

+

R8

1e-5

V4

TD = 100m

TF = 1n

PW = 1

PER = 2

V1 = 0

TR = 100n

V2 = 5

+-

+-

S3

S

VON = 1.0V

VOFF = 0.0V

R11

2.8

K K47

COUPLING = .104

K_Linear

V8

TD = 100m

TF = 1n

PW = 1

PER = 2

V1 = 0

TR = 100n

V2 = 5

R9

2

C60

5m

+-

+-

S1

S

VON = 1.0V

VOFF = 0.0V

R3

.1m

R6

1e-5

R2

.1m

K K27

COUPLING = .113

K_Linear

K K45

COUPLING = .0452

K_Linear

L4

946u

R12

9

V5

TD = 2.3m

TF = 1n

PW = 1

PER = 2

V1 = 0

TR = 1n

V2 = 5

+

-

+

-

S6

S

VON = 1

VOFF = 0

V6

TD = 100m

TF = 1n

PW = 1

PER = 2

V1 = 0

TR = 100n

V2 = 5

K K67

COUPLING = .0322

K_Linear

L10

42n

IC

20k

+

L13

42n

K K12

COUPLING = .248

K_Linear

C1

1m

V1

TD = 1.5m

TF = 1n

PW = 1

PER = 2

V1 = 0

TR = 1n

V2 = 5

K K25

COUPLING = .0379

K_Linear

K K36

COUPLING = .0452

K_Linear

L2

792u

K K17

COUPLING = .113

K_Linear

K K56

COUPLING = .0487

K_Linear

K K16

COUPLING = .0379

K_Linear

L8

42n

R15

9

+-

+-

S5

S

VON = 1.0V

VOFF = 0.0V

C2

1m

K K15

COUPLING = .0489

K_Linear

K K23

COUPLING = .106

K_Linear

K K34

COUPLING = .0806

K_Linear

K K14

COUPLING = .106

K_Linear

C61

5.16e-8

+

-

+

-

S8

S

VON = 1

VOFF = 0

R16

.1m

V7

TD = 1n

TF = 1n

PW = 1

PER = 2

V1 = 0

TR = 1n

V2 = 5

R1

.1m

R18

1e-5

+-

+-

S7

S

VON = 1.0V

VOFF = 0.0V

L6

574u

IC

20k

+

K K13

COUPLING = .174

K_Linear

K K46

COUPLING = .131

K_Linear

+

-

+

-

S2

S

VON = 1

VOFF = 0

K K26

COUPLING = .0489

K_Linear

L3

946u

R5

1e-5

C59

5.16e-8R14

2.8

R7

1

R10

1.9

L7

1.23109e-3

+

-

+

-

S4

S

VON = 1

VOFF = 0

C55

5.16e-8

IC

40k

+

V3

TD = 1.9m

TF = 1n

PW = 1

PER = 2

V1 = 0

TR = 1n

V2 = 5

K K37

COUPLING = .104

K_Linear

R20

2.8

R17

1e-5

L9

42n

R4

1e-5

K K57

COUPLING = .0322

K_Linear

K K24

COUPLING = .174

K_Linear

V2

TD = 100m

TF = 1n

PW = 1

PER = 2

V1 = 0

TR = 100n

V2 = 5

L5

574u

C58

5.16e-8

C3

0.05m

R13

1.9

L1

792u

K K35

COUPLING = .131

K_Linear

Results: PSpice design point for banks

I Fast compression: 1ms rise for coils

Time

0s 1.0ms 2.0ms 3.0ms 4.0ms 5.0ms

I(L2) I(L6) I(L4) I(L7)

-5KA

0A

5KA

10KA

15KA

Results: Engineering design point for banksI Bank engineering point design based on WSI prior design [8]I Shown are two 120kJ modules, 20kV caps. $200k per set.

Results: Engineering design coils

I Coils to be actively cooledin vacuum.

I Design point current forcopper at 20C, althoughcooling will allow bank tobe reduced in size.

I Outer-most coil will useHT superconducting tape(REBCO), since itremains energized.

Results: Full assy for in-vacuum componentsI Coil casings, windings, support for in-vacuum operation.

Results: full system engineering design point

I Test compression and optimize neutron production

I Banks, chamber and diagnostics would fit in a 2000sqft lab

Results: Diagnostics requirements for system

I Thomson, interferometry,magnetics, electrostaticprobes.

I Engineering design pointscompleted as part ofPhase II SBIR

Results: costing for n-Source prototype over 5 years

I Forecast costs by category(materials, labor andoverhead), based onexperience at WSI.

I Build-out over Y1-Y3.

I Labor Y3-Y5.

I Prototype costs are $10M

I Full neutron source:$50M

Further Work

I Neutronics analysis with MCNP6 awaits - coming up inSeptember.

I Design point can be further optimized for efficiency, and forrep-rate.

I CORSICA analysis can be performed to address transport andshape optimization during run-in - this would form part of aSBIR proposal.

I Concept fits into ARPA-E IDEAS program, seeking to usefusion to resolve fission issues.

Summary

I Design point for a n-Source prototype has been examinedanalytically and numerically, using state-of-the-art tools (2Dand 3D time-dependent MHD).

I A full engineering point design has been developed, based onphysics design point.

I Cost of prototype source would be $10M, full source likely$50M.

References

S. Woodruff, R. L. Miller Cost sensitivity analysis for a 100 MWe modular power plant and fusion neutron

source Journal of Fusion Engineering and Design V90, P7-16, (2015)

E. B. Hooper; R. H. Bulmer; B. I. Cohen; D. N. Hill; C. T. Holcomb; B. Hudson; H. S. McLean; L. D.

Pearlstein; C. A. Romero-Talams; C. R. Sovinec; B. W. Stallard; R. D. Wood; S. Woodruff SustainedSpheromak Physics Experiment (SSPX): Design and physics results Plasma Physics and Controlled Fusion,Volume 54, Issue 11, article id. 113001, 26 pp. (2012).

M. Yamada, T. K. Chu, R. A. Ellis, A. C. Janos, F. M. Levinton, R. M. Mayo, R. W. Motley, M. Nagata, Y.

Ono, N. Satomi, and Y, Ueda Experimental investigation of magnetic compression of a spheromak plasmaPhys Fluids B 2 (12) 1990

K. Bol, R. A. Ellis, H. Eubank, H. P. Furth, R. A. Jacobsen, L. C. Johnson, E. Mazzucato, W. Stodiek, and

E. L. Tolnas Adiabatic Compression of the Tokamak Discharge Phys. Rev. Lett. 29, 1495 (1972)

S V Lebedevdag, M V Andrejkodag, L G Askinazidag, V E Golantdag, V A Kornevdag, S V Krikunovdag, L

S Levindag, B M Lipindag, G T Razdobarindag, V A Rozhanskyddag H-mode studies on TUMAN-3 andTUMAN-3M Plasma Physics and Controlled Fusion, Volume 38, Number 8 (1996)

C.R. Sovinec, A.H. Glasser, D.C. Barnes, T.A. Gianakon, R.A. Nebel, S.E. Kruger, D.D. Schnack, S.J.

Plimpton, A. Tarditi, M.S. Chu and the NIMROD Team, ”Nonlinear Magnetohydrodynamics withHigh-order Finite Elements,” Journal of Computational Physics, 195, 355 (2004).

J. A. Crotinger, R. H. Cohen, S. W. Haney, L. L. LoDestro, A. I. Shestakov, G. R. Smith, L. D. Pearlstein,

T. D. Rognlien, A. G. Tarditi, X. Q. Xu CORSICA: a comprehensive tokamak simulation code, Proceedingsof the 1994 International Sherwood Fusion Theory Conference Pages 3 - 47 (1994)

http://woodruffscientific.com/bankdesign