Practical Superconductors Around The Corner?

The problem was the materials themselves. They are ceramics, and therefore brittle. They don’t form wires easily; lengthy strands tend to break and become highly resistive. They are not three-dimensionally symmetric – the superconductivity has a much higher current capacity in one plane than in others, and grains are likely to be randomly oriented unless deposited with great care. The properties are also heavily dependent on composition and a certain level of defects (missing oxygen atoms, for example). And as type-II superconductors, high currents bring with them magnetic flux vortices that can generate their own resistive forces, unless the vortices are pinned by further defects.

And of course the increasing critical temperatures never reached room temperature, like we all half expected.

However, there have been some small-scale applications. Superconducting quantum interference devices can now be made smaller and operated more inexpensively; some RF applications have been put in place as well. Superconducting power cables have even been tried in real life with a 30 meter cable now supplying some power to Copenhagen.

According to the New York Times article, and SuperPower itself, practical large-scale power cables are just around the corner. We’ll see. Apparently SuperPower has hit on a promising technique; using a long metallic substrate tape, polished very smooth, to lay down layers that produce well-connected superconductor along the length of the tape. The process is still a bit on the slow and expensive side though.

Is there something we should be learning from the difficulties here? We’ve spent hundreds of millions of dollars investigating these high Tc compounds – but are we underfunding the applied science that could actually make them practically useful? It’s a little disturbing to your faith in science when something with such seeming promise goes so far awry…

3 thoughts on “Practical Superconductors Around The Corner?”

  1. Ahem.
    One might point out that superconducting MRI machines are today a $4b a year business – from nothing in 1981.
    There is also a very significant NMR instrument market.

    What the author wants is electrical power applications I think: with AC motors, powerlines, transformers, fault current limiters etc. and not (just) DC magnets; and he wants them to use the HTS materials not other materials (which are actually better in many cases).

    One might point out that MagLev trains don’t need superconductors anyway; that transformers are already very cheap machines, that most of the cost of electrical delivery is nothing to do with the current carrying capacity of the wires… I could go on.
    There are exciting, useful, valuable applications for new superconductors, but they are probably not going to be the things that people were dreaming of: because the things most people were dreaming of are not actually worth doing.
    Things that are worth doing are controlling the phase on the electricity grid, controlling harmonics in industrial and marine electrical power systems, limiting fault currents to reduce capital expenditure, making lighter and smaller heavy electrical motors, and things which we haven’t thought of yet.

  2. As far as I can tell, all MRI’s currently in use are based on old-style superconductors, not the new high temperature ones. They are still immensely bulky and require liquid-helium coolant. And have you noticed how expensive they are? While that is an example of actual use of superconductors (just as is their use in several high energy physics accelerators) it’s not quite what we had in mind by “practical”, widespread use. And those applications would be there with or without the new materials.

  3. While an MRI machine may cost $1m, the magnet generally costs only about $30k – including the manufacture.
    They have to be big enough to fit a person inside, and they need to be in a screened room anyway. So there are hard size limits. Look at the size of an xray CT scanner room for comparison. Or look at the size of a 0.2T non-superconductor MRI machine.

    So, simplistically, an improved superconductor would make almost no difference to the cost of an MRI machine. Actually, you save on the cost of the liquid Helium (about $5k a year in the UK – also not significant compared with the cost of radiographers running the thing, despite the relatively low NHS wages) and can reduce the cryogenic cooler capital cost a little bit.

    The superconducting material (NbTi for most MRI) only costs about 1 $/kA.m whereas HTS generally cost 100 – 200 $/kA.m.

    For DC magnets, NbTi/He is actually not a bad technology at all. For AC machines, you really have to be operating above 20-25K because the cryogenic cost is wildly more important (thermodynamics; not materials). That’s where the HTS and MgB2 can become important: see these four meetings about this. [The Niobium alloys stop working above 10 – 15K, depending on the alloy and the magnetic field.]

    There is an OST HTS MRI project: running without a liquid cryogen reduces the mechanical engineering and manufacturing costs a useful amount, and it is easier to make a split machine which is essential for very fat people – an increasing problem especially since very fat people don’t x-ray well either. The OST machine uses B2212 tape which is probably even more expensive than the 200 $/kA.m B2223 tape. The design works, but they will need tape at 5-10 $/kA.m to make it commercially viable.

Comments are closed.