In 1996, a sagging power line in Oregon brushed against a tree, and within minutes 12 million customers in eight states lost power. Such is the vulnerability of today's power grid.

To address this weakness, Berkeley Lab scientists are helping to develop a new approach to power generation in which a cluster of small, on-site generators serves office buildings, industrial parks, and homes. Called a microgrid, the system could help shoulder the nation's growing thirst for electricity — estimated to jump by almost 400 gigawatts by 2025 — without overburdening aging transmission lines or building the 1,000 new power plants required to meet this demand. And it may make statewide blackouts a thing of the past, or at least ensure that service to critical equipment is maintained.

"Catastrophic loss of power to all systems like the 1996 blackout should be impossible," says Chris Marnay, an energy scientist in Berkeley Lab's Environmental Energy Technologies Division. "If we sat down today to devise a power system from scratch, our design wouldn't resemble the one we have."

Chris Marnay (left) and Ryan Firestone of Berkeley Lab's Environmental Energy Technologies Division stand in front of generators that power and heat a laboratory at BD Biosciences in San Diego, Calif. The Consortium for Electric Reliability Technology Solutions (CERTS) is studying this project as an early example of on-site power generation.

Instead of relying solely on large power plants, a portion of the nation's electricity needs could be met by small generators such as ordinary reciprocating engines, microturbines, fuel cells, and photovoltaic systems. A small network of these generators, each of which typically produce no more than 500 kilowatts, would provide reliable power to anything from a postal sorting facility to a neighborhood.

This microgrid appears to the larger grid as if it's any other customer. And it can quickly switch between operating on and off the grid: when the grid offers cheap electricity, the microgrid can purchase it, but if prices rise or there's a power failure, the microgrid can isolate itself. It can also temporarily shed unimportant equipment such as refrigerators during power shortages. This ensures uninterrupted power to the critical computers, communications infrastructure, and control systems that drive today's economy.

"Everything is interdependent. For example, if vital communications go down, other sectors falter," Marnay says. "But if sensitive equipment is powered locally, our vulnerable, centralized power system becomes much less critical, and is a less attractive terrorist target."

Microgrids boast other advantages, but it's no coincidence reliability is high on the list. The concept is being pioneered by the Consortium for Electric Reliability Technology Solutions, a national lab, university, and industry group convened by the Department of Energy in 1999 to explore ways to improve power reliability. The consortium, which is supported by the California Energy Commission and centered at Berkeley Lab, is developing several innovative strategies in addition to microgrids, including managing power grids in real time and determining how an emerging open electricity market affects reliability.

The group will also conduct the first microgrid bench test in early 2004, in which three microturbines and several end loads will be linked together at a utility-grade testing facility. This will be followed in late 2004 by the first microgrid field test, with Berkeley Lab researchers playing a key role in selecting the best site.

The who, where, and why of microgrids

Chris Marnay and colleagues are also developing a computer model that predicts who is most likely to adopt a microgrid, and why. Their work underscores the fact that air quality regulatory restrictions, building code constraints, and site limitations mean some microgrids will be able to use only clean, quiet generators such as fuel cells, instead of more common, gas-fueled reciprocating engines.

Despite such hurdles, a microgrid's many advantages will likely win fans. First, between 60 and 80 percent of the energy consumed by power plants isn't converted to electricity. It escapes as heat, which, unlike electricity, is neither transportable nor easy to use locally. But with a microgrid, waste heat could feed a small, adjacent heat load such as a water heater.

"We'd place power generation where heat is needed, rather than where we can conveniently discard it," Marnay says.

A biotechnology laboratory at BD Biosciences receives most of its electricity from two reciprocating engines. The lab is also heated by the reciprocating engines' waste heat.

Recovered waste heat could also cool and dehumidify buildings, using thermally activated processes. This is doubly advantageous. Cooling buildings places tremendous strain on the power grid, and if a microgrid shares some of this load, it will help both the microgrid customer and everyone using the larger grid.

This leads to another selling point. Microgrids could become model citizens — a term the consortium's scientists apply to loads that help the power system rather than simply take from it. Specifically, microgrids can inject power into the larger grid, which lessens stress on the overall system and helps maintain local service quality.

Microgrids are also becoming competitive. The latest distributed generators can produce electricity almost as cheaply as huge power plants, especially if benefits such as heat recovery are considered. In addition, recent advances in power electronics, such as inverters that connect small solar generators to the grid, make control of small-scale systems economically feasible for the first time.

And generators are getting smaller, such as a closet-sized one-kilowatt solid oxide fuel cell being tested in France. There's also the tantalizing possibility of incorporating fuel cell-powered cars into the microgrid. Simply park your car in the garage, plug it in, and supply power to a few homes. Or plug the car into your office and help power the building.

Like other modern generators, this refrigerator-sized microturbine at the University of California at Irvine's testing facility is fitted with a heat recovery package.

"What better way to avoid load on the grid than to have everyone drive his or her own power plant around," Marnay says.

This transformation will not happen overnight, but it demonstrates how microgrids — along with increased end-user efficiency, improved energy transmission, and renewable resources — can help shepherd the nation from decades of centralized power generation to a new era of decentralized, flexible, and environmentally friendly power generation.

And ultimately, microgrids could change the way people think about electricity. As Marnay explains, today's electricity market is driven by wholesale competition between power companies. Retail competition, in which customers choose electricity based on price and service quality, is virtually nonexistent. Microgrids, however, will offer consumers a choice, and this choice will impose competitive discipline on power companies — yet another way they'll help bridge the gap between today's electricity system and tomorrow's demand.

"Electricity demand continues to grow and people think we can get by with traditional power systems, but we can't," Marnay says. "We need better ways, and microgrids are one part of the answer to this puzzle."

Additional information

• More on the Consortium for Electric Reliability Technology Solutions