Nearly two decades ago, the discovery of a new class of superconductors was announced to the sort of fanfare rarely bestowed on scientific breakthroughs. At a famous March 1987 meeting dubbed "the Woodstock of physics," expectations were high that these materials, which could conduct electricity with no resistance at temperatures attainable with liquid nitrogen cooling, would revolutionize everything from tiny sensors to magnetic-levitation trains.

Because superconductors carry much more current for a given volume, transmission cables, motors, generators, and transformers would be not only more efficient but also much more compact. Magazines like Time and Newsweek hailed a coming technological utopia, with illustrations showing details of the good life that, they suggested, was less than a decade away.

Now, almost 19 years after that meeting of the American Physical Society in New York City, many of the anticipated applications haven't made it past the pilot-project stage. To be sure, hopes for the superconductors remain high, and there have been several encouraging developments: motors using the high-temperature superconductors have been successfully tested for the U.S. Navy, and some highly specialized instruments are on the market. But to date there has been no commercial installation of cable, generators, or transformers.

Instead, it appears that the first superconducting machine to be commercially used in a power system will be one nobody had thought of previously: a superconducting synchronous condenser. Not exactly a household item, a synchronous condenser is basically an electric generator that is optimized and operated to act like a capacitor or inductor, providing the grid with reactive power, an abstruse but absolutely vital element in any power system relying on alternating current.

In any modern-day power grid, the power delivered can do useful work only when current and voltage are perfectly in phase with each other, so that reactive power is neither added nor subtracted. But rotating equipment like motors and generators, in particular, tends to cause current to lag behind voltage, as energy transfers to the magnetic fields of their coils. Left unchecked, this lagging can cause voltage on a network to plunge and ultimately bring down an entire system—the way one did on 14 August 2003, when a huge chunk of the U.S.-Canadian grid collapsed.

Reactive power is what's required to counteract lagging currents and sagging voltage, and if it isn't supplied quickly or efficiently enough, networks crash or equipment suffers. Indeed, reactive-power-supply problems are among the chief culprits in an overall power-anomaly and -disturbance problem that costs the United States alone between US $119 billion and $188 billion a year in lost economic activity, according to a 2001 report by the Electric Power Research Institute, in Palo Alto, Calif. Such staggering losses add up to 1.2 percent to 1.9 percent a year of the country's gross domestic product, the report noted.

In simplest terms, there are two ways of providing reactive voltage support. Banks of capacitors, outfitted with sensors and the latest in semiconductor-controlled switches, can be used to make current lead voltage, opposing inductive effects. Ten years ago, it was widely expected among power system technologists that the latest such devices would enable grid controllers not only to sustain voltage but to steer current through the grid along optimal paths by regulating reactive inductance and capacitance. A big demonstration device, a static compensator, was installed for testing at a Tennessee Valley Authority (TVA) substation in Knoxville, Tenn., in 1995.

The compensator performed pretty well, but grid specialists began to grumble that it was too expensive and cumbersome to perform the role envisioned by proponents of futuristic transmission systems. Perhaps an alternative way of making reactive power was better, after all. This is the synchronous condenser, in essence a generator rigged and operated to act like a capacitor. Like any electric motor or generator, it has stator coils, which don't move, and rotor coils, which move when their magnetic fields interact with those of the stator coils.

The synchronous condenser has one more key component—an exciter—which controls the amount of current feeding the rotor coils. When the exciter stimulates the rotor, it induces reactive power in the stator coils to match the reactive power being consumed on the grid.

Consider the machine under normal, routine conditions. The condenser's inductive stator coils, connected to the grid, consume reactive power. To keep the machine in phase with the grid, the exciter pushes just enough current into the rotor coils to induce compensating reactive power in the stator—that is, the electromotive force induced in the stator coils is equivalent to the reactive power being consumed by the same coils. In effect, the condenser is idling, staying in step with the grid.

But now suppose that it is late morning on a hot day, and thousands of people are turning on their air conditioners. As all those compressor motors get connected to the grid, the voltage sags. To keep up, the current flowing to the rotor coils via the exciter can be increased until the condenser is overexcited, to use the technical term, to provide exactly compensating reactive capacitance.

Because of this innate flexibility and high sensitivity to external conditions, the synchronous condenser is "the way God meant for the job to be done," says Mike Ingram, a TVA research engineer, who asks that his "somewhat southerly way of speaking" be indulged [see photograph, "Nature's Way"].

But even this setup has a serious drawback: resistance in the condenser's rotor coils slows and damps its responsiveness, and the loading and unloading that occur during voltage sags put a lot of thermal stress on its components. Enter SuperVAR, a synchronous condenser in which the rotor coils are made from the high-temperature superconductors that first set imaginations aflame after their discovery in the 1980s. (VAR stands for voltamperes- reactive, the measure of reactive power.)

As Ingram sees it, SuperVAR is nature's absolute best way of creating reactive power. SuperVAR responds to a voltage sag almost instantaneously and, with very little excitation current, can induce much more current than its conventional counterparts. Thermal stresses are minimal or nonexistent.

Conceived by TVA's operations chief, Terry Boston, and manufactured by American Superconductor Corp., in Westborough, Mass., the first SuperVAR machine was installed for testing at a Tennessee steel mill at the beginning of last year. By all accounts, its performance has been outstanding, and American Superconductor expects TVA to soon place an order for two follow-on units. These two units will let engineers see what SuperVAR can do in larger, more complex transmission systems.