Moving More Juice through the Wires

Superconductivity could shave billions off energy bills

Jackhammers chewing through pavement. Barricades around trenches forcing two lanes of traffic to merge into one. Those are two of the ever-present growing pains that all cities suffer--but they could become less irksome because of what's about to happen in Detroit.

Perhaps as soon as summer, Detroit Edison Co. will launch service to 14,000 downtown customers through three 120-meter-long power cables, the likes of which have never before been used by an urban utility in the U.S. The cables are made not with copper wire but with a flat, ceramics-based wire that turns superconductive when the cables are cooled by liquid nitrogen flowing through their cores. One superconducting cable can carry three to five times more power, so power utilities can meet new demand just by pulling out a copper cable and inserting a ceramic one--without ripping up streets.

Moreover, electricity zips from end to end of a superconducting cable with essentially zero resistance. Copper cables, by comparison, lose roughly 7% of the power they transmit due to electrical resistivity. Tack on similar losses from electrical motors, generators, and transformers, and more than 8% of all generated power simply vanishes--dissipated as heat.

While 8% may not seem like much, given America's enormous appetite for power, it's worth billions--and billions more in Europe and Japan. That's why the Energy Dept. and foreign governments have been pumping millions into research on superconductivity since shortly after IBM researchers discovered the first high-temperature superconducting (HTS) materials in 1986. These oddball ceramics sparked plenty of dreamy-eyed promises in the late 1980s. Now, it seems they could soon start coming true.

Detroit is just one example of the major progress in HTS technology over the past two years. Southwire Co. got the ball rolling in January, 2000, by switching on the first industrial HTS transmission system at its home base in Carrollton, Ga. For 12,700 continuous hours, three 30-meter-long HTS cables have funneled all the power to the company's three Carrollton factories, with nary a glitch. Southwire made the cables with HTS wire from Intermagnetics General Corp. and technical help from Oak Ridge National Laboratory.

Last May, Copenhagen beat Detroit to the punch by "energizing" the first HTS cable in a utility grid. A 30-meter cable from Denmark's NKT Group now provides electricity to 150,000 residents served by Copenhagen Energy. Some Paris customers of Electricité de France are slated to get power from a similar HTS project this year.

Also in May, Tokyo Electric Power Co. turned on its third HTS project, using 100 meters of cable from Sumitomo Electric Industries Ltd., and the utility is planning a 300-meter operation for 2005. By far the longest installation anywhere--800 meters--is being designed by a group led by Pirelli Cables & Systems and KeySpan-Long Island Power Authority. It is due online in 2004. And next year could see the start of two more urban tests--one by Edison of Milan, Italy, and another in Columbus, Ohio.

These milestones all have been heavily subsidized by government agencies, but HTS experts assert that business could be reaping payoffs by 2004 or 2005. Then, motors made with HTS materials could more efficiently pump oil through pipelines, drive factory machinery, and propel large ships. Huge magnetic-resonance imaging systems could shrivel in size and cost, moving MRI out of hospitals and into doctors' offices. Electronic signals could zip through telecommunications equipment and mainframe computers up to 1,000 times faster. And there would be no need to burn fossil fuels to generate the juice that's now wasted.

Because U.S. utility execs tend to be very conservative, Energy is counting on Detroit "to establish the viability of superconducting cables in crowded urban locations," says David K. Garman, an assistant secretary at the Energy Dept. That's important because cities are one place where the high cost of HTS cables--some 50 times that of copper--can be justified. To satisfy growing demands for power, urban utilities have two options, notes Paul M. Grant, head of superconducting efforts at the Electric Power Research Institute in Palo Alto, Calif., the utility industry's research arm: "Either you tear up the streets, which can cost a small fortune, or you put in new cables that can carry three times the power of existing cables."

Detroit Edison chose the latter. It grabbed the ends of nine copper cables weighing 18,000 pounds and hauled them out from their underground ducts. Then it threaded HTS cables weighing only 250 pounds into three of the ducts, leaving six ducts for future expansion or rental to telecommunications providers. To braid the cables, Pirelli used 29 kilometers of wire from American Superconductor Corp., the leading supplier of HTS wire.

HTS cables may make economic sense under city streets, but nowhere else. A meter of HTS wire that can carry a 1,000-ampere current costs about $200, vs. $4 for copper. Still, that's a fraction of what it cost in 1995: $1,000 a meter. American Superconductor says costs will gradually drop to $50 after it kicks off production at its new $85 million factory. The highly automated plant can spew out 20,000 kilometers of HTS wire annually--40 times current output.

Producing HTS wire seems simple enough: Fill a silver tube the diameter of a tennis ball with "bisco," the nickname for ceramic compounds consisting of bismuth, strontium, calcium, and copper oxide (BSCCO). Then heat the tube and pull it through a series of progressively smaller holes until it's about 1 millimeter thick. Finally, roll it flat. "Basically, we take spaghetti and make fettuccini," quips Gregory J. Yurek, president of American Superconductor.

However, developing that recipe took countless hours of trial-and-error tinkering--in part because scientists still can't explain why HTS materials do what they do. So theoretical models offer little guidance in optimizing material compositions and processing. IBM's 1986 discovery stood earlier theories on end. Before, superconductivity had been mainly a frigid curiosity: Metals turned superconductive only when cooled with costly liquid helium to temperatures well below -400F. Physicists can explain that. However, even after 15 years of struggling, they are still puzzled by superconductivity in ceramics at relatively balmy temperatures above -300F.

For engineers, though, HTS materials were a revelation. Liquid nitrogen can reach such temperatures easily, and it costs less than bottled water. So harnessing superconductivity suddenly looked feasible. Then they learned that making HTS wire requires exquisite care. The microscopic grains within the ceramic's crystalline structure must line up with atomic-level precision. Any gaps wider than 2 nanometers, or about three atoms, will impede the superconductive rush of electrons.

Much more tinkering lies ahead. One challenge: HTS wires lose their vaunted high-temp powers in the presence of magnetic fields, although superconductive properties can be restored with further cooling. Since all motors and generators produce magnetic fields, HTS models using bisco wire will have to be cooled with expensive helium--until newer, second-generation HTS materials can be produced in volume. That work is underway at Energy's national labs, as well as a score of universities in Europe, Japan, and the U.S.

HTS experts also hope the new materials can slash costs to $10 a meter, which is considered the trigger price for widespread use. Today's bisco wire will never get that low, because "it just uses too much mint-grade sliver," says Philip J. Pellegrino, president of IGC-SuperPower, a subsidiary of Intermagnetics General. The most common second-generation recipes substitute a nickel alloy for the silver, and YBCO ceramics--short for yttrium-barium-copper oxide--for bisco.

The trouble is, producing YBCO wire is proving far tougher than bisco. Expensive coating methods, adapted from the semiconductor industry, deposit several layers of superthin materials--almost as if the flat wire were a superlong chip. Doing this for a meter is no sweat, says EPRI's Grant, "but even 10 meters is a tour de force today." Grant is cautiously optimistic that solutions for volume production will be found. Meanwhile, Yurek is working with researchers at Massachusetts Institute of Technology on a process that mimics the production of photographic film.

Because the new materials are relatively immune to magnetic fields, they are key to HTS motors and generators, says Martin P. Maley, chief supercondutor scientist at Los Alamos National Laboratory. Big motors gobble up lots of wire. The rotor inside a prototype 1,000-horsepower HTS motor, built in 2000 by Rockwell Automation's Reliance Electric unit, contains several kilometers of HTS wire. Such monster motors consume 25% to 30% of all the electricity generated in the U.S., so industrial owners will have a potent incentive to consider HTS models.

Physicists may remain stumped by HTS for a while, but that's nothing new. "Edison was ahead of the science of his time, too," says James Daley, Energy's superconductivity program manager. And today's HTS engineers are determined to follow in his footsteps.

By Otis Port in New York

Before it's here, it's on the Bloomberg Terminal.