Thousands of megawatts of cheap, clean hydroelectricity from Canada are continuously rushing into the New York Power Authority’s sprawling substation in Marcy, NY-enough juice to light up 40 World Trade Centers. For almost a half-century, the Marcy facility, located just a few miles from the remote Adirondack National Park in upstate New York, has transformed this torrent of electricity from a blistering 765,000 volts to the slightly more manageable 345,000 volts used by the overhead transmission cables that feed power-hungry Manhattan 300 kilometers to the southeast.
But the real action at Marcy these days takes place in a nondescript metal building, easily overlooked amidst the 40-meter-high towers supporting the mass of transmission cables. Here, European engineering giant Siemens has just installed the world’s most sophisticated high-power switch. If things get really hot this summer, the ability of the specialized chips inside the device to route electricity exactly where it’s needed just might save New York City’s cool.
The novel Siemens switch that holds these electric-power processors stands four meters high, and its silicon valves juggle electrons at power levels that would blow your cell phone or PC to bits. But just as silicon chips in your cell phone process electromagnetic signals to transmit information, Siemens’s brawny electrical-power processor can filter and manipulate the alternating current flowing through the Marcy station. The immediate goal is to stabilize central New York’s stressed electric grid, making it safe to transmit more energy through the lines. Then next summer, with a few more patch cables and a hefty new fuse added to the system, the power switch should be ready for an even more sophisticated trick: nimbly swapping electricity between high-power transmission cables-a feat never before attempted.
If it succeeds, the electricity swap will be like a coronary bypass for a critical artery in the increasingly sclerotic national power grid. In many ways, the energy “crisis” that is gripping California and threatening the rest of North America is as much about getting electricity to flow where you need it-when you need it-as it is about a lack of energy. The problem is that the existing network of high-power transmission lines, the interconnected web of electricity that keeps the continent charged (power grids in northern Mexico and in Canada are closely intertwined with those in the United States), was built in the middle of the last century and was never meant to handle the complexity and congestion of today’s ever growing energy demands and changing markets.
In light of this situation, the North American Electric Reliability Council, the industry’s voluntary watchdog group, is telling just about anyone who will listen that the system is at risk. “The question is not whether, but when, the next major failure of the grid will occur,” wrote the group’s general counsel David Cook in a recent entreaty to the U.S. Department of Energy in Washington, DC.
The U.S. power grid has been called the largest machine ever built by man. In fact it’s really three loosely interconnected grids: one in Texas, and two more splitting the bulk of the country roughly along the Continental divide ( see “The Electricity Lifeline,” below ). These systems are far from orderly; each grid is composed of a tangle of transmission lines operated by a hodgepodge of owners, from sprawling federal power authorities to regulated utilities to market-savvy conglomerates. An equally variable set of state, regional and federal regulators governs aspects of this mosaic, deciding how much power can enter the grids and flow over each set of lines.
The U.S. power network is built in three large grids: the Eastern, Texas and Western systems. While connected in a handful of spots by direct-current lines, the grids largely operate independently. This map shows the major lines.
Despite its structural and regulatory complexity, though, the power grid operates on a startlingly simple basis: electricity flows from where it’s produced to its destination through the path of least resistance. That worked fine in the days when monolithic electrical monopolies strategically sited their power plants on the grid, with the path of least resistance leading straight to their own customers and no one else’s. But those days are long gone. Deregulation of the electrical industry in the 1990s opened the grid to anyone and everyone who had electricity to sell. Dozens of brokers building new power plants and old utility giants with a fresh entrepreneurial bent now want to supply whoever offers the highest price for their power, wherever he or she may be. And that’s where the physics of the existing grid comes up dangerously short.
The changing nature of the electrical industry dictates complex crisscrossing flows of electricity and the need to send more and more power over long distances. “We’re trying to use [the electric grid] for a lot of longer-distance power transfers, and it’s just stretching to the limit,” warns Thomas Overbye, a power systems expert at the University of Illinois at Urbana-Champaign. Indeed, there already have been signs of troubles. In the blistering summer of 1996, the western U.S. electric grid snapped twice as swollen lines feeding hydroelectric power from the Pacific Northwest to California overloaded and shorted out. The result? Blackouts in 11 western states, Alberta, British Columbia, and Baja California. To avoid a repeat of that crisis, grid operators in California must restrict flows to the state, a fact that is greatly exacerbating its ongoing power crunch.
If today’s situation sounds to you like a recipe for even worse power meltdowns, get your candles ready-because while hundreds of planned new power plants around the country will increase the amount of available electricity, utilities are investing next to nothing in additional transmission lines to get the juice to where it’s needed. It used to be that the big utilities owned and maintained their share of the grid. But deregulation has orphaned the transmission business, uncoupling the lines that deliver electricity from revenue-producing power plants. And owning transmission lines is a business few want any part of. If you think building new power plants is unpopular, try running high-power transmission lines through someone’s backyard. (Do electromagnetic radiation and contentious town-hall meetings come to mind?) Just 13,500 kilometers of high-voltage transmission additions are planned throughout North America over the next decade-a 4.2 percent increase-of which only a fraction are likely to get built. Meanwhile, the U.S. Department of Energy estimates that generating capacity in the United States alone will grow more than 20 percent over that period.
Enter power electronics-like the ones being installed at Marcy. If you can’t build enough new transmission lines to keep pace with the growing power demand, it becomes imperative to build a more efficient way to direct electricity over long distances. In the same way that telecommunications companies have created a complex yet seamless network controlled by automated electronic switches that zap phone calls and data around the world, the engineering giants that build transmission systems are attempting to reenergize the grid electronically.
That transformation has already begun, as a handful of groups, like the Tennessee Valley Authority, install electronic power systems to prop up the far edges of their distribution networks, which are especially vulnerable to energy fluctuations. But even more sophisticated systems, like the one being installed at Marcy, could take electronic control to the heart of the grid. Not only could these power processors make the network more efficient, they could enable a new level of control over the transmission grid, allowing power cables to operate like toll roads and providing revenue sources that could attract the private capital badly needed to upgrade and maintain the systems.
For those wanting to make that scenario happen, the Marcy substation is a critical experiment, a high-stakes test of the technology on a system that tens of millions of people depend on for electricity. Marcy may be 300 kilometers from the bright lights of Broadway, but if power electronics can make it there, it can make it anywhere.
Electronic control over the grid has been a long time coming. The core problem has always been to switch high levels of electricity-and to do it fast enough. The potential solution dates back to the late 1950s, when General Electric pioneered the thyristor, a cousin of the transistor.
Like transistors, thyristors turn the flow of electrons through an integrated circuit on or off. Thyristors are more efficient for handling big power loads because, unlike transistors, once turned on they stay on-allowing energy to flow with little resistance. But the solid-state “latch” that’s the key to this ability also renders the thyristor thousands of times slower to switch than the transistor, limiting its ability to regulate the high-speed action of the electric grid. That began to change in the 1980s with the arrival of a speedy hybrid device called the gate turn-off thyristor, which employs its own dedicated circuit of transistors to electronically open and close the thyristor’s latch. These and other advances have enabled power electronics to be used increasingly in consumer applications such as smoothing out the flow of juice from small-scale power generators (” Power to the People ,” TR May 2001 ).
Now, thanks to the Electric Power Research Institute, a utility-funded R&D consortium in Palo Alto, CA, and demonstrations with utilities and large engineering firms such as Siemens, ABB and Mitsubishi Electric, power electronics is finally ready for heavy lifting in the transmission grid. The first payoff of these systems will be to make the grid less vulnerable to voltage sags and surges, as well as to noise in the power signal.
While the electric grid was originally interconnected to increase reliability and reduce cost, that’s turned out to be a mixed blessing. Interconnection means the most expensive generators can be kept off if others-even several hundred or several thousand kilometers away-can fill the need more cost-effectively. The bad news is that the grid can also transmit disturbances, making the whole system harder to control. Fluctuations can work their way around the grid like the wave among fans at a football stadium. And just as the wave works better at a crowded arena, an electric disturbance becomes more pronounced at higher power levels and with increased power transfers. “Unfortunately, by the nature of the physics involved, the higher the power flows the more dynamically unstable you become,” says Karl Stahlkopf, vice president of power delivery at the Electric Power Research Institute.
Because high-power transmission is so unstable, operators must often limit a line’s load to as little as 60 percent of its ultimate thermal capacity (the point at which the wire overheats, sags into trees or onto the ground, and shorts out). Power electronics is beginning to reclaim this lost capacity using programmable processors that can patch over a surge or sag within a small fraction of a second. That’s a big advance over conventional grid controls, which can be as slow as manually adjusting a transformer or as unsophisticated as automatic breakers that sense a disturbance and “trip” a transmission cable off line, sending a tsunami of power surging through neighboring circuits.
The first of these power processors were installed in 1995 by the Electric Power Research Institute, Siemens and the Tennessee Valley Authority in northeast Tennessee. By smoothing out the flow, electronics eliminated the need for a new power line, saving at least $14 million. It was a modest but encouraging first step, and Siemens and ABB have recently sold three more commercial power processors to Texas utility Central and South West. Meanwhile, Mitsubishi Electric recently installed one in Vermont and is engineering the first of four units destined for San Diego Gas and Electric. By making it safe to draw hundreds of extra megawatts from distant sources over existing transmission lines, San Diego’s devices will help overcome the shortage of local power generation that has left power companies throughout California vulnerable to outages and price swings.
Impressive, maybe, but these systems are still more or less Band-Aids. The installation at Marcy represents the first attempt at major surgery for the grid. The diseased arteries are the two 345,000-volt transmission lines that run south from Marcy to New York City: one skirts Albany and then chases the Hudson; the other traverses the Catskill Mountains to the west before it heads for Manhattan. With a push of a button, a series of breakers at Marcy will reconfigure the station’s circuits to pump electricity from one line to the other as needed ( see “The Electronics Solution at Marcy,” below ).”We will have the ability to actually alter the flow of energy-take it off of one line, put it on another line, particularly if that line is starting to get into an overload condition,” says Gerald LaRose, who runs the Power Authority’s mission control center at Marcy.
LaRose is particularly eager to relieve the line connecting Marcy to New York City via Albany. The operators know it as Marcy-South, and it is easily the state’s most congested high-power transmission line, stuffed to capacity 25 percent of the time and skating within 100 megawatts of overload for much of the rest. The line is obsessively monitored, and each time it approaches critical, the New York Independent System Operator, the agency that manages the state’s grid, must cease adding power to it-indeed, to all of the highly connected circuits running throughout central New York. Even those lines with spare capacity must be squelched, since some fraction of any additional power could reach the stressed Marcy-South line and push it over the edge.
On a hot day when power demand is peaking, squelching electricity flowing from upstate could spell trouble in New York City. At best, the city must fire up expensive and polluting gas- or oil-fired power plants to make up for the constrained flow of hydropower from Quebec. At worst, neighborhoods could be plunged into darkness. While this worst-case scenario has yet to happen, experts agree that New York City is becoming ever more vulnerable.
Then there is the matter of getting cheap electricity to Long Island, one of the country’s fastest growing areas. There, too, power electronics could change everything. The problem is that New York City literally stands between Long Island and cheap power. Squeezing more electricity past this massive bottleneck is nearly impossible. The transmission lines running into New York City are simply too full to carry additional power onward to Long Island. So while Connecticut residents just 40 kilometers away across the Long Island Sound gorge on vast amounts of cheap juice flowing down from Canada’s hydroelectric plants, the eccentricities of the existing grid mean electrically isolated Long Islanders must fend for themselves, relying on local generators to supply a hefty 93 percent of peak power demand. This is one reason that Long Island endures some of the highest electricity rates in the country.
Power electronics is providing a solution so compelling that it is driving the construction of the first U.S. high-power line built by transmission entrepreneurs, who see a huge profit in bridging Long Island’s power gap. As in the Marcy system, two power processors will control energy flow on this transmission line, which will stretch underneath the sound. But instead of sitting back-to-back as they do at Marcy, the processors will lie on opposite shores, sucking AC power out of one grid, pumping it as DC power under the sound via a 42-kilometer underwater cable, and regenerating the AC wave on the other side. (The conversion to DC cuts costs, because underwater AC cables are more expensive than their DC counterparts.) “You want 100 megawatts to go in one direction? Just turn the dial. Want it in the other direction? You just turn the dial,” says Jeffrey Donahue, president of Transnergie U.S., a subsidiary of Montreal-based power giant Hydro-Qubec and the builder of the $120 million link.
New York’s power problems rank among the country’s toughest because the state’s grid is particularly complex and congested, but its situation is far from unique. Grid operators up and down the eastern seaboard are straining to meet rapid growth in electricity demand, while persistent regional bottlenecks present a constant cause of worries in the Midwest. (Transnergie is eyeing several projects, similar to the one in Long Island, to span the Great Lakes.)
Fortunately, even as the problems escalate, power electronics may be about to benefit from smaller and cheaper silicon switches-in the same way PCs have benefited from faster and cheaper computer chips. Call it Moore’s Law for power electronics. Quicker switches providing cheaper power processors mean that, for the same buck, grid operators can install larger systems with greater impact on the grid. And that’s good news for grid controllers, for whom power over power is intoxicating. “Now that I understand what it does, I wish it was bigger,” says Len Panzica, one of the controls engineers who put Marcy’s system through its paces this spring.
The grand vision, of course, is to electronically tame the nation’s vast power network. Unlike isolated devices that regulate a few lines each, integrated network controls could synchronously tweak all of a system’s electronics to optimize flow over the entire grid. Stahlkopf of the Electric Power Research Institute estimates that integrated control could boost the overall transmission capacity of existing infrastructure by 30 to 40 percent. Stahlkopf figures this leap forward is at least 10 years off, but he says utilities are already beginning to take an important step-wide-area telemetry providing operators with a real-time picture of how much power is flowing over their lines and from where.
However, with anarchy gripping today’s power grid, integrated controls-even in a decade from now-seem like a bit of a pipe dream. Rapid deregulation has swept away the old rules without offering coherent alternatives for who should run the network and how they will get paid for it-making it an especially tough time to market advances offered by power electronics.
New systems like those at Marcy are even more politically charged because they can spontaneously reconfigure the grid, potentially increasing the strategic value of some power plants and idling others. LaRose says the Marcy project has few enemies because it can only shift a few hundred megawatts in a system that handles over 30,000 megawatts daily. But he says a larger project could find itself facing formid-able opposition. To implement these new technologies, you’ve “got to walk very gingerly through that minefield,” says LaRose.
What’s more, power electronics aren’t cheap. Even with Siemens, the Electric Power Research Institute and 21 utilities keen to demonstrate the technology chipping in $13 million for the Marcy project, the New York Power Authority must come up with $35 million more, which it’s attempting to do by selling bonds to wary financial investors. With the rules for transmission investments in flux, there is a real possibility that the Power Authority will never fully recoup its investment.
So why roll the dice on expensive equipment? The engineering answer is that for the foreseeable future, anyway, power electronics is the best hope for stabilizing the electric grid. The more pragmatic answer is that the New York Power Authority, a state-owned corporation, is directly accountable to politicians who fear the wrath of voters if the rolling blackouts darkening California’s economy roll across the Empire State. Unfortunately, it may take a few more dark days and cold nights without electricity before the rest of the grid’s numerous interested parties begin to see the light.
The 50-year-old problem that eludes theoretical computer science
A solution to P vs NP could unlock countless computational problems—or keep them forever out of reach.
The moon didn’t die as early as we thought
Samples from China’s lunar lander could change everything we know about the moon’s volcanic record.
Forget dating apps: Here’s how the net’s newest matchmakers help you find love
Fed up with apps, people looking for romance are finding inspiration on Twitter, TikTok—and even email newsletters.
Inside the machine that saved Moore’s Law
The Dutch firm ASML spent $9 billion and 17 years developing a way to keep making denser computer chips.
Get the latest updates from
MIT Technology Review
Discover special offers, top stories, upcoming events, and more.