Looking for a Grid's Weak Spots? Scientists May Have the Answer
University researchers think they’ve figured out how to better predict future blackouts on the U.S. power grid.
(Bloomberg) -- University researchers think they’ve figured out how to better predict future blackouts on the U.S. power grid.
Up until now, there’s been little information available on why a local failure can disrupt the wider grid, making it difficult to understand and predict cascading power outages. Previous models either examined only a small section of the grid, or didn’t account for the complexities of power flow.
In an article published in the journal Science, three researchers from Northwestern’s physics and astronomy department say they’ve made a significant step forward in solving that problem. In determining the likelihood of cascading failures, they built a mathematical model around data from the entire national grid that spots vulnerabilities by examining how electricity would be redistributed in the event of a line outage.
Their findings are sure to be of interest to grid operators, who have invested billions in recent years in shoring up an aging power system made up of thousands of miles of transmission lines that crisscross the nation. Its vulnerability was highlighted in 2003 when a high-voltage power line brushed into an overgrown tree in Ohio, causing a staggering cascade of failures that left 50 million people without power, including most of those in New York City.
The researchers factor in the seasons, peak usage times and out-of-service lines. That’s important, they wrote, because previous studies weren’t of the same scale, and didn’t apply the physics of cascading failures or grid operation practices.
The good news is that the study found that the number of vulnerable lines is fairly small: only around 11 percent of linked lines in their tests ever experienced a primary failure that would lead to a cascade. To mitigate the risk, though, grid operators could boost capacity on those.
To contact the reporter on this story: Emma Ockerman in New York at eockerman@bloomberg.net.
To contact the editors responsible for this story: Lynn Doan at ldoan6@bloomberg.net, Stephen Cunningham, Carlos Caminada
©2017 Bloomberg L.P.