353

Hydrogen cryomagnetics: the way forward for superconductivity

Glowacki B. A.

Applied Superconductivity and Cryoscience Group, Department of Materials Science and Metallurgy,University of Cambridge, Pembroke Street, Cambridge CB2 3QZ, England, UKInstitute of Power Engineering, ul Augustówka 6, 02-981 Warsaw, Poland

Approaching the 100-year anniversary of discovery of superconductivity in 2011should be more than just reminiscences how it has happened and who actuallydiscovered superconductivity, it should be the turning point in Applied Super-conductivity and Cryoscience R&D where electric materials engineering, cryogenictechnology combined with construction engineering should path the way to providethe required sustain future energy and cryogenic infrastructure under heading ofhydrogen cryomagnetics where superconductivity interlocked with hydrogeneconomy can be the solution to most of our energy and transport problems.

SUPERCONDUCTORS

Cryomagnetic applications that involve superconducting materials such as Low TemperatureSuperconductors, LTS: NbTi (Tc=11 K), Nb3Sn (Tc=18 K), Medium Temperature Superconductors, MTS:eg. MgB2 (Tc=39 K), ReFeAsO1-x (Tc=55 K) and High Temperature Superconductors, HTS: YBa2Cu3O7

(Tc=91 K), (Bi,Pb)2Sr2Ca2Cu3Ox, (Tc=117 K), (Tl,Hg)Ba2Ca2Cu3Ox (Tc=164 K) require operation at widerange of temperatures from 1.8K to 164K. Such a wide temperature spectrum does include all cryogenicliquids. Due to fact that liquid hydrogen under normal pressure has temperature of 20 K and is only onecryogenic element, which is an energy carrier it may become the actual future choice of indirect or evendirect cooling medium for superconducting devices.

Before reaching more general conclusions on the possible necessity of transition from helium tohydrogen cooling medium let us analyse a few representative examples which may indicate the necessityof such transition.

LOW TEMPERATURE SUPERCONDUCTORS AND HELIUM

There is necessity of direct cooling, of superconducting devices such as MRI or NMR electromagnts andcurrently constructed toroidal and poloidal powerful electromagnets of the ITER fusion reactor, madefrom metallic superconductors such as Nb3Sn and NbTi by liquid helium Fig. 1 [1].

Recently conducted simulation of the global resources of helium shown that there will be a continuousprice increase [2] fig. 2. Fig. 2 indicates that helium wholesale price tends to increase with time although, inprinciple the price may drop if over supply occurs. It was noted from the recent economic slowdown, thatprices stabilise rather than fall and that more specifically the BLM effectively becomes a floor price for themarket. This may happen if the industry over estimates future demand, hence resulting in over-investment.

354

Under the long-term contracts, where wholesalers must supply helium, then over supply is very likely tohappen. Price variation is mainly contributed by two elements. One is the production cost; another is theimbalance between demand and supply, which is called excess demand in the model. This constant rise inprice will put a downward pressure on demand, eventually curtailing market growth as consumers seekalternative technologies to reduce their helium dependence. Because helium is a by-product of natural gasproduction, and its value is very small compared with the value of natural gas (although it costs more perMCF, the raw helium concentration is nearly always very small). Therefore any effort to reduce the end-demand of helium via recycling or alternatives technologies does not significantly change the depletion pathof helium resources. It is happening already in some parts of the world, the low temperature research has tostop due to the very high cost of the liquid helium reaching today ~ 50 £/lHe.

Figure 1. Plot of upper critical magnetic induction, Bc2, versustemperature, T, for low critical temperature superconductors,medium critical temperature MgB2 (Tc=39K) and high criticaltemperature superconductor YBa2Cu3O7. (Magnetic fluxdensity, B, is parallel to crystallographic c-direction of thesuperconducting YBa2Cu3O7 layer)

Figure 2. Helium wholesale price $/MCF (2008)of Monte Carlo simulation, [2]

Preventing loss of helium in the superconducting devices is an important issue in case of MRImagnets and implementation of the zero boil-off technology does provide a solution, but the helium lossduring quench of the fusion reactor magnet section may reach unacceptable cost level, preventingservicing such energy delivery device. It is projected that DEMO fusion reactor and future conceptreactors will require alternative approach where medium temperature and high temperaturesuperconductors will be used to provide high magnetic induction at elevated cryogenic temperatures [3,4]. As it will become obvious in the next chapter, some of the high temperature superconductors, due totheir specific temperature dependent magnetic property can be effectively used at LH2 (20K) or LNe(25K) temperatures [4]. For example a preliminary design of a DEMO, CICC, made with Bi2212 roundstrands and operating at 20 K and 13 T leads to a value of Jcable not very different from a Nb3Sn CICCcable in the range of 54 A mm-2 [5].

HYDROGEN FOR MEDIUM / HIGH TEMPERATURE SUPERCONDUCTORS

Demand for hydrogen is growing every year and for example only in USA the actual mass growth rate ispredicted to increase dramatically from n = 2 to n = 6, see figure 3.

355

Figure 3. Predicted mass production demanded by the USAeconomy versus time. It is evident that the differences for twoscenarios: normal development and demand driven by green-house gas are almost identical until 2050. The n value of thecalculated time = massn may increase from n = 2 to n = 6 overnext ten years

Recent study of applicability of 1 GW class hybrid energy transfer line of hydrogen and electricitywas conducted in details by Yamada at al. [6]. Current plans are based on high temperature fission reactorto produce H2, liquefier it at nearby liquefaction plant, and transfer of LH2 by superconducting dc cablemade of MgB2 through the shipping Mutsu-Ogawara port for wide distribution. Target distance of hybridenergy transfer line is 1000 km. Hydrogen refrigeration station is placed on every 10 km of the unitsection. Capacity of the liquid hydrogen transportation is 100 tons per day where a typical hydrogen LH2

plant delivers 3 tonnes/day.Considering liquid hydrogen safety, a direct cooling, can be only handled by highly specialised

organisations and companies, but a indirect liquid hydrogen cooling, iLH2, can be an viable option fordecentralised economy. In iLH2 installations a helium gas exchanger can be used, transferring coolingpower of the hydrogen bath at ~ 20 K to the desired cryo-magnetic installation [4, 7]. Economicalcalculations of cooling efficiency of the large 15 T electromagnetic non-superconducting device,supported by testing study for the scheduled purpose build installation in USA conducted for LHe, LH2

and LN2 shown very clearly that cooling directly by helium or neon is 70 times and 100 times moreexpensive than indirect cooling by LH2 [7]. Taking to account price of 1l of LH2 ~ £0.5, LHe ~ £3) andNe ~ £150 and also the fact that hydrogen is the only element that is an energy carrier, choice of LH2 as acryogenic cooling medium for cryomagnetic applications is rather apparent.

Similar concept can be implemented in smaller hydrogen cryomagnetic installations such asdecentralised hydrogen energy generation, storage and use, where superconducting energy devices [8] canbe indirectly cooled and operate, as presented in figure 4 [9] and figure 5 [10].

Figure 4. Simplified layout of thedecentralised liquid hydrogen cycle [9].There is apparent reason for integration ofthe hydrogen generation, storage and usewith cooling of superconducting devices

356

Figure 5. Indirect cooling of an fully superconductingMgB2/YBaCuO bearing [10] using a liquid hydrogendewar and gHe CryoFan heat exchanger. Helium gas ispumped around a closed loop with one end cooled to20K via a heat exchanger connected to a liquidhydrogen dewar. This setup is under design andconstruction stage and is going to be a part of a liquidhydrogen laboratory in the new Department ofMaterials Science and Metallurgy building, Universityof Cambridge (completion in 2012)

In fig. 5 the novel concept of a magnetic bearing which uses the difference in irreversibility field oftwo superconducting components, see figure 1., to allow one superconducting component to be field-cooled in the field originating from the other component which is first magnetised at a highertemperature can be exploited thanks to the liquid hydrogen indirect cooling system [4, 7, 9, 10]. Alsoprogress of the manufacturing technology of large bulk superconducting plates of MgB2 (Tc ~ 40 K)which do dot suffer from the granularity problem such as HTS may become a very important magneticbearings material for the flywheel considering that the temperature provided by liquid hydrogen will beat the level of 15-20 K [11].

Figure 6. For small basic liquefaction plants the energyneeded to liquefy hydrogen may exceed the HHV of thegas. But even with the largest plants (10,000 kg/h) about30% of the higher heating value (HHV) energy is neededfor the liquefaction process. The total energy needed togenerate and compress hydrogen at filling stationsexceeds the HHV energy of the delivered hydrogen by atleast factor 1.5. The vertical broken line marks a very lowproduction of 1 litre LH2/h. It is obvious that the standardH2 liquefaction process requires dramatic efficiencyimprovement if it is going to provide energy carrier andcoolant for superconducting devices

Concept of the small volume decentralised liquid hydrogen generation devices [12] face the majorengineering and materials challenges because of the process efficiency, where even with the largest plants(10,000 kg/h) about 30% of the HHV energy is needed for the liquefaction process, figure 6. The solutionto this challenging complex problem may only come from a new design of the compact liquefactionsystem where an initial cooling stage of the H2 gas will come form the mixed gas portable coolersoperating at the 60 - 80 K temperature range [13] but the final liquefaction stage will be reinforced withmagnetic refrigeration [14] and expansion turbines [15] that can dramatically improve the efficiency ofthe decentralised small liquefiers. Ceramic polycrystalline magnetic refrigerant such as for exampledysprosium-gadolinium-aluminium-garnet enables achieve condensation at the level of 90% Carnotefficiency and the liquefaction efficiency reaching 50% of the Carnot efficiency [14]. There is also urgent

357

need for research on more effective hydrogen generation, purification, compression and ortho-paraconversion to improve liquefaction efficiency and long-term storage.

ACKNOWLEDGMENT

This work was carried out under SUPERGEN 14 project ‘H-delivery’.

REFERENCES

[1] http://www.iter.org/mach/Pages/Magnets.aspx[2] Cai, Z, Clarke, R. H., Glowacki, B. A., Nuttall, W. J. and Ward, N., Ongoing ascent to the helium production plateau -

insights from system dynamics, Resources Policy (2010) 35 77–89.[3] Glowacki, B. A., Nuttall W. J., Assessment of liquid hydrogen cooled MgB2 conductors for magnetically confined fusion,

Journal of Physics: Conference Series (2008) 97 012333.[4] Clarke, R. H. and Glowacki, B. A., Indirect hydrogen versus helium or nitrogen cooling for fusion cryogenic and magnet

systems, presented at ICEC 23 - ICMC 2010, 19-23 July 2010, Wrocław, Poland (to be published in Adv. Cryo. Eng.).[5] Duchateau, J. L. and Hertout, P., Which superconducting material for the toroidal field system of the fusion DEMO

reactor? Journal of Physics: Conference Series 97 (2008) 012038.[6] S. Yamada, S., Hishinuma, Y., Uede, T., Schippl, K., Yanagi, N., Mito, T. and Sato, M., Conceptual design of 1 GW class

hybrid energy transfer line of hydrogen and electricity, Journal of Physics: Conference Series 234 (2010) 032064.[7] McDonald, K. T., Iarocci, M., Kirk, H. G., Mulholland, G. T., Titus, P. H. and Weggel, R. J., Use of He Gas Cooled by

Liquid Hydrogen with a 15-T Pulsed Copper Solenoid Magnet, Technical / Economic Report August 15, (2002) Page 1-5E951; also presented at ICEC 23 - ICMC 2010, 19-23 July 2010, Wrocław, Poland (to be published in Adv Cryo. Eng.).

[8] Janowski, T., Głowacki, B. A., Wojtasiewicz, G., Kozak, S., Kozak, J., Kondratowicz-Kucewicz, B., Majka, M. andWoźniak,, M., Fault current limitation in power network by superconducting transformers made of 2G HTS, presented atASC 2010 Washington, USA, (to be published in IEEE Trans. Appl. Superconductivity).

[9] Glowacki, B. A., Liquid Hydrogen as a energy vector and coolant for superconducting devices, Biuletyn PolskiegoStowarzyszenia Wodoru i Ogniw Paliwowych, (2010) No.5 13-21.

[10] Patel, A., Palka, R. and Glowacki, B. A., New fully superconducting bearing concept using the difference in irreversibilityfield of two superconducting components, presented at PASREG July 29 – 31, 2010, Washington, USA ; (to be publishedin SUST special issue).

[11] Giunchi, G., Ripamonti, G., Perini, E., Ginocchio, S., Bassani, E. and Cavallin, T., Highly densified MgB2 bulks byreactive Mg liquid infiltration, Advances in Science and Technology (2006) 47 7-16.

[12] Piotrowska-Hajnus, A., Tomov, R., Chorowski, M. and Glowacki, B. A., Comparative efficiency analysis of low – ratehydrogen liquefaction cycles, presented at ICEC 23 - ICMC 2010, 19-23 July 2010, Wrocław, Poland (to be published inAdv Cryo. Eng.).

[13] Gong, M., Wu, J. and Qinguo Hu, Q. Development of a -186oC cryofreezer based on a dual mixed-gases Joule-Thomsonrefrigeration cycle, presented at ICEC 23 - ICMC 2010, 19-23 July 2010, Wrocław, Poland (to be published in Adv. Cryo.Eng.).

[14] Kamiya, K., Takahashi, H., Numazawa, T., Nozawa H. and Yanagitani, T., Hydrogen liquefaction bymagneticrefrigeration, Cryocoolers, eds Miller, S. D. and Ross, Jr. R.G., (2007) 14 637-644.

[15] Clarke, E. and Gardner, J. B., Refrigeration with expansion turbines, Contemp. Phys., (1976) 17 No. 6, 507-528.