Winner: Superconductors on the High Seas

New ship motors propel a quiet revolution

14 min read

The ocean-going vessel, with its pivotal role in spreading civilizations, sustaining global trade, and settling wars, was as responsible as anything else for the world we have today. So it's not surprising that until a half century ago, humankind applied its most advanced technology to the task of pushing boats around the seas. In the 19th century, when the age of sail was already about 5000 years old, marine architects ushered in an age of engineering genius and romance as they built the great steamers, which later gave way to diesel ships and even nuclear vessels.

Now, a US $78 million U.S. Navy effort is returning ship propulsion to the absolute forefront of advanced technology. In a program with far-reaching implications--not only for future warships but also for the cargo, cruise ship, and conceivably even the electricity-generating businesses--the U.S. Office of Naval Research (ONR) is testing a 5-MW, 23-ton superconductor ship motor and is also already well into the design of a full-scale, 36.5-MW superconductor motor.

"Superconductor technology will help reduce the size and weight of motors, generators, [and] power transmission and supporting components," says Rear Admiral Jay M. Cohen, the Chief of Naval Research. Cohen also expects the technology to kick into high gear the Navy's ongoing effort to move from mechanical-drive to electric-drive propulsion for its ships.

The superconductor motors pack the same amount of power into a package that can be nearly 70 percent smaller and lighter than a conventional motor, even with all of the superconductor's attendant cooling systems figured in. That kind of savings is an enormous advantage on the high seas, where it would translate into more cargo space or ordnance-carrying capabilities, or, on a cruise ship, as many as 20 extra berths. As a bonus, the motor's efficiency also goes up by 1 or 2 percent, depending on the load--an improvement that in a high-power application like ship propulsion means hundreds of thousands of dollars in fuel savings over the course of a year for a typical cargo ship.

The primary contractor for both superconductor motors is American Superconductor Corp., in Westborough, Mass., the maker of the superconductor wire at the heart of the motors. The 5-MW motor, built at a cost to the U.S. Navy of $8 million, was delivered last July and is now sitting in a test bay at the Center for Advanced Power Systems at Florida State University, in Tallahassee [see photo, " Motor Man"]. For the next nine months or so, the motor will be pushed to the limit as tests are conducted in a dynamometer controlled by a real-time digital simulator. The results will be poured into the work now taking place under a $70 million Navy contract for the 36.5-MW motor.

If the tests of these machines fulfill designers' expectations, they could point the way to the first lucrative, large-scale application of the so-called high-temperature superconductors, referred to as HTS, discovered and unveiled in a stupendous worldwide media blitz 17 years ago. Although Sumitomo, Ultera, Pirelli, Southwire, and others have built and tested superconducting electric transmission cables in the years since, commercial sales have been modest so far.

Part of the difficulty for the cable makers stems from the fact that "high temperature" is in this case very much a relative term: it is "high" only in comparison with the near-absolute-zero temperatures that were needed to achieve superconductivity before the 1987 discovery. High-temperature superconductors become perfect conductors, losing absolutely all trace of electrical resistance, at a temperature as high as 100 degrees above absolute zero (100 Kelvin). That's still 173 °C below freezing, but it's balmy in comparison with the 10 K (-263.15 °C) needed to get traditional superconductors to shed their resistance. That enables cooling to be done by relatively inexpensive options like gaseous helium, neon, or liquid nitrogen instead of costly liquid helium. Cooling costs for devices that operate at below 10 K are 10 times those for ones that operate at 100 K.

Still, even the much more favorable costs of liquid nitrogen are a challenge for makers of transmission cables, because the cables must be cooled along their entire length. In a superconductor rotating machine like a motor or generator, on the other hand, only key rotation-producing components, such as the rotor, need to be cryocooled to an operating temperature of about 32 K.

The payoff could be huge. Superconductor motors would be a natural for cruise ships, which started going electric several years ago. Cruise and cargo vessels are already a $400 million annual market for motors and generators, according to maritime consulting company MSCL LLC, in Alexandria, Va. Ironically enough, although the Navy is paying for the tests on the 5-MW motor, superconductor ship propulsion might very well take hold in the commercial world first--a development that experts say would nevertheless hasten military acceptance not only in the United States but also in Britain, France, and Italy. Each of their navies is planning a move to electric drive for next-generation destroyers or frigates.

Greg Yurek, American Superconductor's CEO, told IEEE Spectrum that he expects to "see some 5-MW HTS motors on commercial ships by the first half of 2005, about a year after orders are placed." He declined to name companies, however, citing confidentiality agreements with potential customers.

Attractive though the opportunities are in the marine world, there are even bigger ones in other markets, such as industrial electric motors--a $1.2 billion-a-year industry. And electrical generators, too, could benefit, because the same advantages of small size and weight and high efficiency would apply to generators built with superconducting materials. In a separate project, General Electric Co., in Fairfield, Conn., is working on a $12.3 million project sponsored by the U.S. Department of Energy to build a 100-MVA superconducting generator. "We're watching the technology because it might be suited to shipboard use" once units are in the 20- to 30-MW range, says ONR's director of ship science and technology Scott Littlefield, "but we're focused on the motors."

To understand the Navy's enchantment with superconductor technology, consider the typical U.S. Navy destroyer. It generates power using seven gas turbine engines burning marine-grade diesel fuel. The engines produce pressurized gas that spins the turbine blades to produce power. But only four of those turbines push the ship forward, producing 70­80 MW for the task. These turbines are tied directly to the drive shaft through a gearbox. The other three generate 7.5 MW of electricity for the ship's suite of high-power radars, computers, lighting, appliances, navigation, and communications systems.

When a ship isn't moving at top speed--about 90 percent of the time--all four turbines aren't required for propulsion. Typically, only two propulsion turbines are used for day-to-day operations. But excess power from those idle or underworked turbines can't be put to other uses, such as lights or radar, because the propulsion turbines are all mechanically connected to the ship's huge drive shafts. Add in the space taken up by the turbines, gearbox, and propeller shafts, and you get, well, shafted. Ships must be designed around propulsion shafts; they're over 60 meters long in a destroyer.

Winner: Superconductor Motors

Goal: Build and test full-size superconductor electric motors for ship propulsion

Why it's a Winner: Ships of all kinds are going to electric propulsion. Although the Navy will be first to test a complete superconductor motor, cargo, cruise, ferry, and other big commercial ships will all be prime candidates for the new technology

Organizations: U.S. Office of Naval Research, American Superconductor Corp., Center for Advanced Power Systems at Florida State University

Centers of Activity: 5-MW motor testing at the Center for Advanced Power Systems in Tallahassee, Fla.; 36.5-MW superconductor motor design and development at American Superconductor in Westborough, Mass.

Number of People on the Project: 25 total

Budget: US $78 million for two motors

The Navy already has a detailed vision of what its next generation of destroyers will look like: it's called the DD(X). The vessel will be the first class of ship that the Navy plans to take all electric, in 2011. (Though the Navy calls it the all-electric ship, gas turbines will still be used to generate electricity to power motors and generators.) In the DD(X) destroyers, no turbines will be directly connected to the propeller shafts. Instead, they'll spin generators that will feed motors, and everything else that needs electricity. Electric propulsion allows the turbines to run at their most fuel-efficient speed, and lets all the ship's electricity needs dictate the number of turbines operating at any given time.

Although converting the mechanical power to electricity before it gets to the propellers inevitably means some losses, they can be limited to about 10 percent at full speed, Littlefield notes. And these small losses will be more than offset by the huge advantage of letting the ship's crew divert power to a host of new, electrically intensive systems and weapons that seem pulled from the pages of science fiction--advanced multifunction radars, ultrapowerful microwave defense systems, and lethal lasers, electric rail guns, and electromagnetic launch and recovery of aircraft, says Rear Admiral Cohen.

The Navy hasn't officially decided to use superconducting motors for the DD(X) yet; it is still weighing--literally, in this case--other options, such as motors with permanent magnets. But among the advantages of the superconductor technology is the fact that motors powerful enough to drive ships aren't expected to cost substantially more than ordinary copper-wound motors, according to American Superconductor's Yurek. The main reason is reduced manufacturing costs--assembled superconductor motors can be shipped just weeks before sea trials, rather than being factory built, disassembled, shipped, and reassembled in place during the first months of shipbuilding. It is also difficult, some experts say, to see how the Navy's grand vision could be realized with conventional motors, which in this case would weigh several hundred tons apiece, as opposed to 75 tons for superconducting units, with their cooling hardware included.

To grasp the advantages of superconductor motors, some background may be helpful. Motors convert electrical energy into mechanical energy. They do it by spinning a rotor inside a typically nonmoving structure called a stator. Both the rotor and the stator are magnetic, with pairs of north-south poles arranged around their circumferences. North poles on the rotor are attracted to south poles on the stator, and vice versa, causing the rotor to spin. As the poles approach, a control system swaps the stator poles at just the right instant, keeping the rotor going.

Simple motors have two pairs of poles, the 5-MW motor has three pairs, and the 36.5-MW motor will have eight rotor pole pairs. In the electric drive system of a ship, the rotor is connected directly to the propeller or drive shaft.

Like all electromagnets, those in the stator and possibly in the rotor are coils (some motors use permanent magnets for the rotor). In a big conventional motor, these coils are usually copper; in a superconducting motor, the rotor coils are made of a high-temperature superconductor and the stator coils are copper. The material of choice is bismuth strontium calcium copper oxide (BSCCO--pronounced bisco), a ceramic that can be cooled with off-the-shelf, closed-loop refrigerators that use gaseous helium.

The basic advantage of superconducting rotor coils is that their wires can carry over 100 times more current than a copper wire of the same cross section does. So the coils can produce much stronger magnetic fields in the same space. Stronger fields in the motor mean more torque, or rotational force. A side benefit: the high efficiency of superconducting motors is constant across most of their power range.

The two basic types of electric motors are synchronous and induction. In a synchronous motor, the stator interacts directly with the rotor's magnetic field. In an induction motor, on the other hand, the rotor does not have fixed magnetic poles, and it spins because the stator sets up a moving magnetic field, which in turn induces a secondary magnetic field in the rotor. Superconductor motors are synchronous machines, because the rotor currents have to be direct current. Both synchronous and induction motors are used in the electric drives of commercial cruise ships today, but the more costly synchronous motors dominate because they are more efficient.

Conventional electric motors have changed relatively little in decades, in terms of efficiency. For superconductor units, even when power losses from the motor cooling systems are figured in, the motors can best conventional power plants by about 1 percent. Crucially, that high efficiency is constant, regardless of the ship's speed. As it happens, the variable that has the biggest impact on efficiency is the rotor-coil winding material, and that's where superconductors come in.

In the 5-MW motor, the stationary stator winding is made of copper, but the rotor winding is made of BSCCO [see diagram, " Special Rotor"]. The BSCCO tape is wound into an oval racetrack shape. Yes, BSCCO is called tape, because it's thin, wide, and flat; the material isn't cylindrical like metal wire.

Another advantage of superconductor motors is that they have no iron teeth in their stator windings. A conventional motor has these teeth to strengthen the magnetic field and help restrain the stator conductors. But the superstrong field in the stator of a superconductor motor would simply saturate iron teeth. Getting rid of the teeth makes the motor not only smaller and lighter but much, much quieter.

Still, building a superconducting motor has its own unique challenges, and perhaps the biggest is chilling the coils of the spinning rotor. American Superconductor did not put the cryocoolers on the rotor itself; it would be too hard to maintain them if they were inside the motor. Instead, the cooling system circulates frigid helium gas around the rotor, carrying away the heat through tubes to the stationary cryocooler nearby.

The cryocoolers can be stocked and replaced as needed even while the motor is running, Yurek says. They're also a little smaller than a two-drawer filing cabinet, so stocking spares aboard a ship doesn't eat up the space you've just gained by switching to a superconductor motor.

At American Superconductor's wire manufacturing plant in Devens, Mass., David Paratore, senior vice president of the company's wires business, leads a plant tour. It's an attractive place, if you like blue. BSCCO starts out as five salts, measured and mixed. After the first round of processing, it's a royal blue powder, Paratore explains. And as it happens, almost all the plant's wire-making machinery is painted a matching bright, deep blue, purely for aesthetic reasons, Paratore assures me.

To drive home his earlier point about the current-carrying prowess of BSSCO, Yurek hands me a bundle of copper wires, as thick as my forearm, that can carry 1200 amperes. Then he hands me a bundle of BSSCO strands that can carry the same current: they're like nine strands of stainless steel linguine.

Though it is adept at producing those remarkable strands, and it did wind the superconducting rotor of the 5-MW unit, American Superconductor isn't in the business of making complete motors. So it turned to Alstom Power Conversion in Rugby, UK, to produce an oil-cooled stator using copper wire and to assemble all the pieces into a complete motor.

After those jobs were done, last summer, Alstom performed brief tests in which technicians hooked the motor up to another one, which provided a test load [see photo, " Passing Grade"]. The point was to make sure the motor could produce the expected level of torque--a whopping 208 000 newton-meters. "I saw the end of it," ONR's Littlefield says of the Rugby test. "It was the first time I saw the whole motor put together," he says of the momentous moment and motor his office funded. It's an ONR project, so the Navy felt "it was a high-risk technology at this stage," Littlefield recalls. "There was at least a possibility it would fail. But it didn't. Seeing it test without any significant trouble was gratifying. I've seen--and heard--motors. This one was smaller and quieter than the conventional motor supplying partial load for the factory testing."

Since then, the motor and its electronic controller from Alstom have been shipped to a laboratory at Florida State University for load tests on a dynamometer to confirm full torque at full speed, and at variable torques and speeds to simulate realistic loads. A dynamometer is a test setup based on a second electric motor that opposes the spinning of a motor being tested, as in the brief trial at Alstom. In this case, two 2.5-MW load motors will operate in tandem.

"We don't want to damage the machine in testing," Littlefield continues. "But we'll need to test full RPMs and full torque at load--plus do a backing-down maneuver, where the ship suddenly slows and feeds power back up the shaft, turning the motor into a generator." The fun will start this month at the Florida laboratory called the Center for Advanced Power Systems (CAPS) and is funded by the Office of Naval Research.

CAPS was built specifically to test motors up to 5 MW running at up to 450 RPM and with the flexibility to accommodate a variety of equipment for testing. The three-story, basic red-brick CAPS building is so new that doctoral candidates mark their cubicles with paper signs, and Florida's ubiquitous ants haven't had time to set up shop yet in the newly laid grass framing the parking lot.

In November, CAPS engineers took a reporter on a tour of the facility, starting with the high-ceilinged test bay where the superconductor motor sits, ready to go, waiting for the two load motors that make up the dynamometer to arrive. The brain of the test facility is actually upstairs, in a spacious second-floor corner room. It's a mostly serene and airy place--the home of the real-time digital simulator. The simulator's disks hum faintly, sunlight streams in from huge windows overlooking the gravel service yard where transformers supply the high voltages needed by all the test equipment, and occasionally, there's a thunk as a bird swoops down into the big glass windows.

It's also the domain of Stephen Woodruff, a soft-spoken aeronautical engineer at CAPS. For the past six months, Woodruff has been painstakingly creating a software model of the motor in the simulator. "We'll evaluate the model by running it against the live tests," he says.

Woodruff will run simulations of his model undergoing tests even as those tests are run on the real thing downstairs. The simulator will control the dynamometer, which will provide the superconducting motor with the conditions it would experience at sea.

As the motor runs through various speed scenarios, the simulator reads the real-time test-bed data, such as torque, speed, phase currents and voltages, and stator and rotor temperatures. Then, also in real time, it computes such dynamic effects as that of motor drive currents on the rest of the ship's power system and the actual load torque. Finally, the simulator tells the motor drive converters how to produce the desired torque.

"That's how the motor is able to think it's in that simulated environment," says Tom Baldwin, an IEEE Senior Member and assistant professor of electrical engineering at Florida State. The dynamometers will mimic mechanical loading in real time, so the simulator can be programmed to control the dynamometer and use it to reproduce ship movements in the water or in ice or when the propeller or even the ship's stern hits a rock.

The brain behind the testing is a programmable-logic control (PLC) unit, from Rockwell Automation Inc., in Milwaukee, Wis., which sits in a first-floor room. From the two touchscreens on it, testing operators will be able to monitor and control the test setup, including the superconductor motor and the dynamometer. The PLC is always in control of testing except when it hands control over to the real-time simulator, which can act in its stead in some situations.

Michael Coleman has been programming the PLC for about a month. He's designed the interface for the PLC touchscreens. From here he can control the cooling for the stator, the dynamometer, and lots of other variables. And he can hand off control of the testing to Woodruff and the simulator or just enable the simulator to make the motor feel like it's riding ocean waves to see how it reacts.

The Navy wants a motor that'll run for 30­40 years, but no one has that kind of time to run tests. So CAPS will run the motor around the clock for a month and see how it does. After that, the Navy will ship the motor to Philadelphia to another lab for tests with a real propeller.

Perhaps the biggest boon of the simulation, though, is that it will let next-generation machines be built on a solid foundation of validated engineering data. In this case, that next generation is the full-size, 36.5-MW motor. "You save R and D money with validation," notes Steinar Dale, an IEEE Fellow and the CAPS laboratory director.

Trends in marine propulsion suggest that all the simulation and validation going into the 36.5-MW motor is time well spent. Not only are the drive systems of big commercial ships going diesel-electric, they are also increasingly being built in a configuration known as "podded" [see diagram, " Powerful Pod"].

In-hull propulsion is what most people think of as ship propulsion. Propeller blades are at the end of a long drive shaft sticking out of the back end of the ship's hull. In turn, the drive shaft is connected to a turbine or engine either directly or through a gearbox. In a diesel-electric system, the diesels drive a generator that powers a motor that connects to a shortened drive shaft.

In podded propulsion, the motors are installed in pods that hang from the stern of the boat below the waterline. The generators and turbines or engines can be several decks above the lowest, from which the pod hangs. The pod, including the propeller blades, can be swung in a wide arc, directing thrust as needed to maneuver. So the pod itself serves as the ship's rudder. Instead of sticking out the back of the pod as an in-hull propeller does from the hull, the propeller sticks out the front of the pod. It's attached to the shaft of the electric motor sitting in the pod.

Pods maximize maneuverability and hydrodynamic efficiency, and free up space for paying passengers, cargo, or ammunition. For example, the Queen Mary 2, nearing completion in France for Cunard Line Ltd., in Southampton, UK, has four propulsion pods and room for 15 or so cabins where the original Queen Mary had propeller shafts. Shipbuilders estimate that a container ship using podded superconductor propulsors could carry anywhere from 2.5 to 4 percent more containers, about 40 for every 454 tons saved in ship's weight. The two 44-MW conventional motors aboard the Queen Elizabeth 2 weigh about 400 tons each; the 36.5-MW superconductor motor is expected to come in at 75 tons.

The U.S. Navy is interested in pods, but power density is the issue, according to ONR's Littlefield. The power-density requirements on a destroyer require a motor too big to fit in a pod. But superconductors offer a way to shrink the motor, gain power density, and "at least consider pods," Littlefield says.

Industry experts put the global market for electric motors and generators for electric-drive ship systems at $400 million in 2002 and upwards of $2 billion a decade later. Commercial cruise ships and some cargo vessels have spent the last decade switching to electric propulsion: low-speed, high-torque electric motors that directly turn the propellers that drive the ships. The low-speed (230 RPM), high-torque (208 000 N*m), 5-MW motor sitting in the high bay at CAPS is just the power rating required today to move passenger ferries, research ships, cable-laying vessels, and tanker and oil-rig supply ships.

With all those ships poised to go electric-drive, the U.S. Navy intends to follow suit and take it a step further, eventually converting virtually all shipboard systems to electric power. That will happen as soon as American Superconductor and its partners--ONR, Northrop Grumman Ship Systems, Northrop Grumman Marine Systems, Ideal Electric, Syntek Technologies, CAPS, and maritime consultancy MSCL--deliver the 120 RPM, 36.5-MW motor for testing and everything turns out shipshape.

This article is for IEEE members only. Join IEEE to access our full archive.

Join the world’s largest professional organization devoted to engineering and applied sciences and get access to all of Spectrum’s articles, podcasts, and special reports. Learn more →

If you're already an IEEE member, please sign in to continue reading.

Membership includes:

  • Get unlimited access to IEEE Spectrum content
  • Follow your favorite topics to create a personalized feed of IEEE Spectrum content
  • Save Spectrum articles to read later
  • Network with other technology professionals
  • Establish a professional profile
  • Create a group to share and collaborate on projects
  • Discover IEEE events and activities
  • Join and participate in discussions