A DOWNMARKET neighbourhood in Detroit is not the most obvious place to
discover a bunch of excited scientists and engineers. But here in Frisbie, we
believe we鈥檙e part of a revolution. From next month, the electricity delivered
to 30,000 homes in Detroit will pass through three superconducting cables each
120 metres long, fitted at the Frisbie substation. It sounds rather modest. But
we expect future generations to talk about 2001 as the year electricity changed
forever.
Superconductivity means no electrical resistance, and that in turn means no
wasted power. Till now, aluminium and copper cables have carried electricity
from the power stations where it is generated to the factories, offices and
homes that use it. By the time mains electricity reaches your home, more than 10
per cent of the power that was fed into the grid by the generating station has
been wasted as heat, produced by the electrical resistance in the transmission
cables.
The superconducting revolution has been a long time coming. Superconductivity
was discovered during Thomas Edison鈥檚 lifetime, so in itself it鈥檚 nothing new.
The first few super-conductors to be found were pure metals. They only lost all
resistance if they were cooled to extremely low temperatures鈥攍ess than 10
kelvin or -263 掳C鈥攁nd to keep them this cold required an array of
unwieldy refrigeration machines. Operating temperature wasn鈥檛 the end of the
problem, either. These superconducting metals can carry only a very limited
current before they lose their superconductivity. Even worse, these materials
lose their superconducting properties if they encounter even moderate magnetic
fields. As big currents inevitably generate big magnetic fields, that seemed to
spell the end for any hope of producing superconducting power cables.
Advertisement
Then, in the 1950s and 1960s, researchers began to find that a number of
metal alloys could superconduct, and would retain their superconductivity when
passing big currents or when subjected to large magnetic fields. And the 1986
discovery of a new class of superconducting materials鈥攖he copper oxide
based 鈥渉igh-temperature superconductors鈥濃�攐pened up even more
possibilities. To this day no one understands how superconductivity occurs in
these materials. But that didn鈥檛 matter: as superconductivity became possible at
less extreme temperatures, people began to dream of materials that had zero
resistance even at room temperature.
But after an initial flurry of improvements, we鈥檝e been stuck for many years
at temperatures little warmer than 130 K. For a while this seemed like an
insurmountable barrier to widespread use. Whatever the application, these
materials would have to be cooled by some complicated means. But in fact, the
real practical barrier has already been overcome, and at much lower
temperatures. The 77 K boiling point of liquid nitrogen, which is a cheap
readily available commodity, provides an easy means to cool superconductors for
practical applications. A number of companies have now made the high-temperature
superconductors into tapes a few millimetres across. The tapes can be wound
around a duct that carries a flow of liquid nitrogen. Layers of thermal and
electrical insulation complete the system, forming a robust power cable
(see Graphic).
These cables are now available from several companies in lengths that are
measured in kilometres rather than metres. Cooled by the circulating liquid
nitrogen, they are capable of transporting currents of 100 amps or more with
zero resistance. In fact, the current-carrying capability of superconducting
cable has improved so fast that it now far exceeds the capacities we envisaged
when we started drawing up the plans for the Frisbie substation. The only way
we鈥檒l be able to test the cables at full capacity will be to persuade all the
local residents to turn on their washers, dryers, indoor lights, TVs,
heaters鈥攅verything and anything that can be connected to the mains.
The switch-on at Frisbie has been four years in the making. The scheme was
conceived when a team comprising Pirelli Cables, Detroit Edison, American
Superconductor, Lotepro (a subsidiary of the Swiss Linde corporation) and my
employer EPRI (the Electric Power Research Institute) proposed to the Department
of Energy that superconductors could now be put to work.
Our superconducting cable provides three times the capacity of a copper cable
the same size. Though it might seem like overkill now, Detroit will one day need
this increased capacity. Forecasts predict that the city鈥檚 demand for
electricity will increase by as much as a 60 per cent by 2010. And if the
Frisbie project saves Detroit, it might also change the delivery of electricity
across the US beyond recognition. America鈥檚 power infrastructure, much of which
is half a century old, is beginning to crumble. The recent blackouts in
California highlighted not so much a lack of generating capacity as bottlenecks
and weak points in the transmission network that delivers the power to where
it鈥檚 needed.
A period of intense renewal is about to begin. Some liken it to the
construction of the interstate highway system in the 1950s and 1960s. Like the
highway network, the new power network will be expected to last for decades. You
don鈥檛 go out and tear up the nation鈥檚 highways just because a better concrete
comes along, and we won鈥檛 be ripping out the transmission grid if a better wire
becomes available.
So the choice of technology at the time of construction is crucial. In the
long run it can pay to use the best, even if it costs somewhat more than the
conventional alternative.
Right now, I believe the best option available for the power network is
superconducting cable, and government officials are beginning to realise this
too. Earlier this year, the US National Energy Policy Development Group鈥檚 Report
recommended that 鈥渢he President direct the Secretary of Energy to expand the
Department鈥檚 research and development on superconductivity鈥�.
Silver lining
But the major argument against starting the superconducting revolution is
still a powerful one: superconducting cable costs far more to buy than the
copper alternative. The superconductor tapes are 70 per cent silver, which is
needed to bind its constituent materials together. The performance of the
superconducting tapes continues to improve, however, and volume production will
bring costs down. We expect that cables manufactured with the tape now available
will bottom out at two to three times the cost of conventional copper cables of
the same power capacity, including the costs of running a refrigeration
system.
And with an astonishing discovery made earlier this year, the cost of
superconducting cable might soon plummet further. It鈥檚 hard to know how everyone
missed it in the past, but it turns out that the simple compound magnesium
diboride becomes superconducting near 40 K. It鈥檚 cheap, and available almost off
the shelf. It鈥檚 also easy to turn into wire. Already tens of metres have been
produced in several laboratories worldwide, despite the fact that its
superconductivity was only discovered nine months ago. It isn鈥檛 upset by high
currents or magnetic fields. Though it would have to be cooled by liquid
hydrogen or helium (which could make refrigeration up to seven times more
expensive) it already looks like it could significantly reduce the cost of
future superconducting power cables.
But even without this advance, superconducting cables might already make an
economic alternative to existing cables where these don鈥檛 have the capacity to
meet the demand for electricity. The key factor here is that it will be far
easier and cheaper to install them than to add more copper cables. A
superconducting cable can carry at least three times the power of the copper it
replaces, so a straight swap automatically increases power capacity. Frisbie has
shown how easy it can be to fit superconductors. Here, 9 old cables containing
over 8 tonnes of copper in all have now been replaced by a mere 110 kilograms of
superconducting wire in three cables. Yet in terms of capacity, the new power
cables match the old ones. Their light weight made it easy to pull them through
the existing underground ducts, despite the five 90-degree bends that had to be
negotiated. Initial measurements on the cables indicate that they survived
manufacturing, spooling, shipment from Pirelli Cable鈥檚 factory in Milan, and
installation with only a 1 per cent loss in the maximum current they can
carry.
Going all the way
In the US, about 80 per cent of all high-voltage underground cable鈥攁nd
that means just about everything but the lines from the local substation to your
house鈥攊s contained in ducting, and thus can be upgraded in the same way.
In theory, a network of superconducting transmission cables could take
electricity all the way from the power plant to the suburbs.
Promising as they are, superconducting cables aren鈥檛 perfect. One snag is
that they do not totally banish power losses. While the superconducting tape can
carry very high currents and withstand large magnetic fields and still retain
its zero resistance, this is only true with direct current. With the alternating
current that flows through the power distribution system, the superconducting
cables do suffer some losses of energy. This is because the magnetic field
created by the current penetrates the superconductor in particular regions,
creating islands or 鈥渧ortices鈥� of normal conductivity. As the alternating
current cycles back and forth, the resulting variation in the magnetic field
sloshes these vortices around within the superconductor, losing energy as heat
as they do so. Higher voltage cables, which carry lower currents, lose less
power, but the average is around 1 per cent loss.
Even so, these wires still lose only 1/200th as much power as equivalent
copper conductors. Copper loses so much of the electrical power as heat that
many transmission wires and transformers have to be cooled by circulating oil,
just as in a car engine. That creates its own problems: fires and oil spills are
a common hazard with standard electric power equipment. In the superconducting
cables, however, the liquid nitrogen removes any heat that鈥檚 generated. And if
there鈥檚 a spill, the nitrogen simply evaporates and returns harmlessly to the
air .
The other remaining problem is that the refrigeration technology needed to
keep the superconducting cable supplied with liquid nitrogen is not really ready
for commercial use. The cryogenics shed at Frisbie is extremely reliable, but it
was put together specifically for this pilot. It has lots of built-in back-up
systems to ensure that the project doesn鈥檛 fail simply for lack of coolant,
which means it wouldn鈥檛 be economic for general use.
But there should be no problem separating nitrogen from the air, and
liquefying it economically and efficiently. People have been doing this for
decades, and I know of one company that is already developing a system suitable
for supplying a substation. There is no fundamental reason why we should not be
able to set up refrigeration stations to cool ducted wire every 500 metres. As
we have seen with other applications of superconductors (see 鈥淕lobal cooling鈥�),
if the end result is desirable enough, efficient and economical ways of
cooling the materials will be developed.
None of the problems of superconducting cables now outweigh the benefits. We
are at the point where it is worth running superconductors, and everyone is
beginning to realise it. In May, a Danish project began sending power through
superconducting cables to 150,000 homes in Copenhagen. Tests are also under way
in Tokyo, though they are not yet delivering power to customers. We are
expecting the US Department of Energy to announce three new superconducting
cable projects and the development of an all-superconducting substation very
soon. The message is simple. The superconducting revolution is here. Resistance
is useless.
You may already be using superconductors without knowing it. In many mobile
phone masts, parts of the electronics are superconducting, operating at hundreds
of degrees below freezing. Working without electrical resistance allows the
masts鈥� signal filters to pick up extremely weak signals, which increases the
number of channels available for use. And if you鈥檝e had a hospital MRI (magnetic
resonance imaging) scan recently, that may well have been powered by
superconducting magnets that significantly boost the sensitivity of the
device.
In both these cases there鈥檚 no obvious evidence that the components are
working at such low temperatures. The mechanical cryocoolers attached to these
devices make it possible to cool something to within 20 degrees of absolute zero
using nothing more exotic than mains electricity. These impressive fridges are
often no bigger than a beer can. There are no messy cryogenic fluids involved:
they compress a gas鈥攊t could be helium, neon or argon鈥攁nd then allow
it to expand. During this cycle, the gas absorbs heat from the system to be
cooled and then ejects it into the atmosphere. To anyone using the system, these
boxes appear to be little more than another component to be connected up.
It wasn鈥檛 always that way. Doctors using MRI machines used to have to
negotiate vats of helium and various types of complex refrigeration. The cooling
technology for delivering superconducting electrical power may still be rather
rudimentary, but it could quickly become highly sophisticated.