Use of He Gas Cooled by Liquid Hydrogen with a 15-T Pulsed Copper Solenoid Magnet

K.T. McDonald

Princeton University, P.O. Box 708, Princeton, NJ 08544, USA

M. Iarocci and H.G. Kirk.

Brookhaven National Laboratory, P.O. Box 5000, Upton, NY 11973, USA

G.T. Mulholland (deceased)

Applied Cryogenics Technology, P.O. Box 2158, Ovilla, TX 75154, USA

P.H. Titus.

Princeton Plasma Physics Laboratory, P.O. Box 451, Princeton, NJ 08543, USA

R.J. Weggel

Particle Beam Lasers, Inc., 18925 Dearborn Street, Northridge, CA 91324,USA

International Cryogenic Engineering Conference 23 – International Cryogenic Materials Conference 2010 (Wroclaw, Poland, July 20, 2010) K. McDonald ICEC23-ICMC2010, Wroclaw July 20, 2010 1

Cool Magnets to Lower Their Resistance – and Their Power Consumption

We considered a 15-T, 20-cm-diameter, warm bore, pulse copper solenoid.

Would require 70 MW to operate at room temperature.

Favorable to operate at ~ 30 K, to reduce resistance by a factor of 30.

If go below 30 K, the very low heat capacity of copper leads to rapid temperature rise.

 , c p and  /c p for High-Purity Copper (  =0.05  cm below 20 K)  , c p and  /c p for High-Purity Copper at Very Low Temperature 3.6

1.8

0.5

3.2

Heat capacity, C P 2.8

c p 1.6

1.4

0.3

 /C P Heat capacity, C P Heat capacity, c p [J/cm 3 K] 2.4

1.2

Resistivity,  [  cm] 0.2

2.0

1.6

 1.0

Resistivity,  0.8

0.14

Ratio,  /c p 1.2

0.8

0.4

0 30 0.6

0.10

 /c p 0.4

 /C P 0.07

Resistivity,  0.2

60 90 120 150 180 Temperature [K] T (K) K. McDonald 210 240 270 0 300 0.05

20 ICEC23-ICMC2010, Wroclaw 30 40 50 Temperature [K] T (K) July 20, 2010 60 70 80 90 100 2

Cooling Concept: He gas + LH

2

Heat Exchanger

The concept is simple – and we foresaw low-cost implementation using recycled components.

Vent H 2 gas to atmosphere “Weathered” 20,000 liter LH 2 Dewar Concept based on direct cooling of aluminum and copper magnet coils by liquid hydrogen and liquid neon in the late 1950’s.

Laquer, RSI 28, 875 (1957) Surplus heat exchanger (from the SSC) Circulate helium gas thru magnet to cool it 15-T pulsed copper magnet 20-cm-diameter warm bore (new) K. McDonald ICEC23-ICMC2010, Wroclaw After the success of large, high-field superconducting magnets in early 60’s, this concept was largely forgotten.

July 20, 2010 3

Choice of Cryogens

Only candidates are H 2 , He and Ne.

Magnets sometimes catch fire  don’t cool directly with hydrogen.

Heat capacity per liter same for He and Ne gas, so use cheaper He gas.

Quality factor Q for the refrigeration of the circulating gas via liquid cryogen consumption (boiling in the heat exchanger) was defined as Q (kJ/$US) =  H V   L  (1 m 3 /1000 liter)  (liter/$US).

That is, Q is a kiloJoule of heat-of-vaporization/$US at T NBP . Fluid T NBP  H V  L Cost Q K kJ/kg kg/m 3 $US/liter kJ/$US (Costs from 2002) He 4.2

20.3

124.9

3.00

0.85

H 2 Ne 20.3

27.1

446.0

85.8

70.8

1207.0

0.53

173.00

59.58

0.60

N 2 77.3

199.0

808.0

0.07 2297.03

An operational cycle of the system involved a 10-s-long pulse of the 15-T magnet during which 18 MJ = 18,000 kJ of energy was generated, followed by a 30-min cooldown.

LH 2 Cooling Cost = 18,000 / Q = $300 per pulse.

LHe Cooling Cost = (60/0.85)  (LH 2 Cooling Cost) = $21,000 per pulse.

LNe Cooling Cost = (60/0.60)   (LH 2 Cooling Cost) = $30,000 per pulse.

Clearly, liquid hydrogen is favored economically.

K. McDonald ICEC23-ICMC2010, Wroclaw July 20, 2010 4

What Came of This?

We developed a PI diagram and presented it to the Lab Safety Committee.

But when an 8-MW power supply became available, we used it, along with liquid nitrogen cooling of the magnet.

(Thanks to F. Haug for the LN 2 cryo system of the CERN MERIT Experiment.) K. McDonald ICEC23-ICMC2010, Wroclaw July 20, 2010 5