Keeping the Juice Flowing

It turns out America has power to spare. But much of it leaks away because of outdated power lines

No element of infrastructure is more important to the health of the U.S. economy than the electric grid, which feeds electricity to every street light, desktop computer, dishwasher, and aluminum smelter in the country. Last year, during California's energy crisis, a task force headed by Vice-President Dick Cheney garnered headlines by predicting that the U.S. would have to build more than one power plant a week through 2020 to avoid a capacity shortfall.

Generating enough juice is turning out to be a short-term problem, though. Investors have proven quick to put up new plants where profit is available and approval is forthcoming. What's more, the U.S. as a whole had enough generating capacity last summer. It's just that far-off power plants couldn't ship enough electrons to power-hungry California because the wires couldn't handle the load. Over the long term, the national transmission grid--the network of high-voltage lines that shuttles power from where it's produced to substations for local distribution--is the biggest problem of the electricity infrastructure.

The transmission grid is old, huge, complex, and delicate--and it has no backup. Control engineers must precisely match the supply of power to fluctuating demand to keep its voltage steady and its alternating current strumming at precisely 60 cycles per second. But deregulation has stressed the system by unleashing huge volumes of long-distance power sales. Since electricity always follows the path of least resistance, a sale of power from Wisconsin to Florida can inadvertently add to congestion everywhere from Ohio to South Dakota. Meanwhile, uncertainty about how transmission owners will be compensated under deregulation has discouraged spending on new power lines. And rival companies that own pieces of the grid have become less interested in cooperating.

The upshot is that there are too few power lines, and they're being managed poorly. To protect themselves from sags, spikes, and outages, industrial customers are forced to install costly backup batteries known as uninterruptible power supplies. A year ago, the North American Electric Reliability Council, an industry group, wrote to the Energy Dept.: "The question is not whether, but when, the next major failure of the grid will occur."

Fortunately, new technology can improve matters by enabling system operators to make optimal use of the grid--and to plan its expansion in the most cost-effective way. Technology isn't the only answer, of course. It's even more important to get the regulatory structure right so market players have incentives to strengthen the grid. But rule makers such as the Federal Energy Regulatory Commission (FERC) can tap into recent advances, particularly in software, to create an industry structure that encourages private investment while restoring the grid's reliability.

When the grid fails, economic activity grinds to a halt. Yet annual inflation-adjusted spending on transmission nearly fell by half during the 1990s as peak demand rose 15%, according to industry data.

And outages aren't the only problem of an unhealthy power grid. Bottlenecks in the system also raise the daily cost of electricity by forcing a more costly mix of generation to be used. Low-cost plants can't produce up to their capacity when overloaded power lines can't handle their additional juice. Meanwhile, major cities such as New York and San Francisco must use expensive locally generated power because they can't import enough cheaper power from the outside. The operator of New England's grid, ISO New England Inc., for example, projected last year that transmission bottlenecks would increase the cost of electricity in the six-state region by as much as $600 million a year through 2006. Nationwide, the annual cost stretches into the billions.

How can technology make a difference? Start with reliability. FERC is pushing utilities to join Regional Transmission Organizations (RTOs) that will be responsible for operating the high-voltage lines across large swaths of the country. Those groups--some of which are already in operation--need advanced software to keep order as hundreds of generators sell power to millions of customers across vast distances.

PJM Interconnection LLC in Valley Forge, Pa., is leading the way. The budding RTO includes Pennsylvania, New Jersey, and Maryland and describes itself as the world's largest competitive wholesale electric market. Instead of fighting the profit motive, PJM harnesses it. Prices for power don't get averaged out. Instead, they're specific to the actual cost of supplying the juice to a particular point. So customers who live in places that don't have enough power lines or power plants pay more. There's even an active market in "financial transmission rights" that guarantee capacity on specified routes. This system directs investment where it's most needed. So if a certain area has chronically high electricity prices, investors know they can profit by building a new plant in the area or extending a line into it. By doing so, they help solve the power-supply problem.

Elegant as PJM's solution is in concept, it depends on having accurate data about the actual physical state of the network. PJM uses software from Germany's Siemens (SI ) for real-time grid management, France's Alsthom (ALALF ) for market operations, and U.S. companies General Electric Co. (GE ) and Power Technologies Inc. for planning. But lots of other companies offer solutions, particularly for planning. Power Computer Applications Corp. in Mesa, Ariz., is one well-established player. Another, PowerWorld Corp. of Champaign, Ill., has visualization tools that help explain congestion problems to nonengineers. The Tennessee Valley Authority says software from PowerWorld and other changes are saving $30 million a year on its generating expenses while reducing outages.

But the most intriguing approach may be that of Optimal Technologies Inc. of Benicia, Calif. It claims to have perfected an optimization algorithm that better understands the nonlinear nature of power flows--the way small perturbations can have catastrophic results. In a demonstration last year for the California Energy Commission, Optimal did a postmortem analysis of a rolling blackout in the San Francisco Bay area on June 14, 2000, when the temperature hit 103 degrees. It concluded that if the California Independent System Operator and Pacific Gas & Co. had done things differently that day--by changing the settings of the remote controls on substation equipment in just the right way--it could have kept certain circuits from overloading, avoided the emergency blackout, and even sold excess power. CEO Roland Schoettle says the software is fast enough that it can be used for operations, not just planning. Optimal is completing patent applications and expects to go commercial later this year. "They've got a terrific tool," says Don Kondoleon, manager of the California commission's transmission-education program.

Power lines and power plants are only two tools for stabilizing the grid. A third is to curtail peak demand by paying customers to turn off equipment when soaring demand threatens to cause instabilities--an old idea that has gotten more sophisticated. Atlanta-based RETX Inc., for instance, sells software that lets system operators instantaneously order up "negawatts"--reductions in demand--from places where power flows are causing congestion.

Trouble is, even terrific tools won't work if the regulatory structure doesn't give companies the incentive to use them. Today, for instance, most system operators focus primarily on maintaining transmission reliability. So optimization programs that show them how to cut generating costs are of little interest. Yet under deregulation, system operators aren't doing much long-term planning, either. No wonder, since it's not even clear what their objectives should be. For instance, entrepreneurs would like to build generating plants in Nevada and ship their output to California. Should Nevada be able to say no? Should it be paid to say yes? Software-optimization tools can figure out the best way to balance competing interests--but only after the political system has decided how to rank priorities.

Beefing up and securing the U.S. power-transmission grid is crucial to the country's economic growth. Tough decisions about who should run the grid--and how--still have to be made. The good news is that the technology to help solve the grid's problems already exists.

By Peter Coy in New York

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