President Obama in this year's State of the Union address talked about the future of energy and mentioned "self-healing power grids" -- a grid that is able to keep itself stable during normal conditions and also to self-recover in the event of a disturbance caused, for example, by severe weather.

But as the national power-grid network becomes larger and more complex achieving reliability across the network is increasingly difficult. Now Northwestern University scientists have identified conditions and properties that power companies can consider using to keep power generators in the desired synchronized state and help make a self-healing power grid a reality.

The Northwestern team's design for a better power grid could help reduce both the frequency of blackouts and the cost of electricity as well as offer an improved plan for handling the intermittent power sources of renewable energy, such as solar and wind power, which can destabilize the network.

"We will be looking at a completely different power grid in the future," said Adilson E. Motter, who led the research.

"The use of renewable energy is growing. More people will be driving electric cars, and the power grid will be delivering this energy, not gas stations. We need a power grid that is more capable and more reliable. This requires a better understanding of the current power grid as well as new ways to stabilize it."

Motter is the Harold H. and Virginia Anderson Professor of Physics and Astronomy at Northwestern's Weinberg College of Arts and Sciences.

The crux of the challenge is that for the U.S. power grid to function the power generators in each of its three interconnections (Eastern, Western and Texas) must be synchronized, all operating at the frequency of 60 hertz. Out-of-synch power generators can lead to blackouts that affect millions of people and cost billions of dollars -- losses similar to those of the Northeast blackout of 2003.

Having a network that can synchronize spontaneously and recover from failures in real time -- in other words, a self-healing power grid -- could prevent such blackouts. To help achieve this, power companies could apply the Northwestern guidelines as they add power generators to the network or tweak existing generators.

A paper describing the researchers' mathematical model, titled "Spontaneous synchrony in power-grid networks," is published in the March 2013 issue of the journal Nature Physics.

When a problem develops in the power-grid network, control devices are used to return power generators to a synchronized state. Motter likens this to using medicine to treat someone who is ill. He and his colleagues are suggesting conditions to keep synchronicity in good shape so interventions are kept to a minimum.

"Our approach is preventive care -- preventing failures instead of mitigating them," said Motter, an author of the paper and an executive committee member of the Northwestern Institute on Complex Systems (NICO). "The guidelines we offer could be very useful as the power grid expands."

The researchers derived a condition under which the desired synchronous state of a power grid is stable. They then used this condition to identify tunable parameters of the power generators that result in spontaneous synchronization. This synchronization can be autonomous, not guided by control devices.

"The blackout at this year's Super Bowl was caused by a device that was installed specifically to prevent blackouts," said Takashi Nishikawa, an author of the paper and a research associate professor of physics and astronomy at Northwestern. "A large fraction of blackouts have human and equipment errors among the causes.

"Reduced dependence on conventional control devices can improve the reliability of the grid," he said. "Our analysis also suggests ways to design control strategies that potentially can improve the existing ones."

Power generators are very different from each other; some are large and others small. Motter and his colleagues identified a "body mass index" for power generators, which they suggest should be kept approximately the same (making, in essence, all generators look the same to the network) in order to strengthen spontaneous synchronicity in the system. If the body mass indices change, they should be changed in a coordinated way.

The researchers demonstrated their model using real power grids of hundreds of power generators, similar to the size of the Texas portion of the U.S. power grid.

In addition to Motter and Nishikawa, other authors of the paper are Seth A. Myers of Stanford University and Marian Anghel of Los Alamos National Laboratory. The Nature Physics article is available here.