Professor calculates optimal sizes for risky systems
April 21, 2014
907-474-7541
04/18/2014
Large, complex systems like electrical power grids can only grow so big before the risks and costs of failure outweigh the benefits of their size, according to David Newman, physics professor at the University of Alaska Fairbanks.
Newman and two coauthors make this claim in the paper “Does size matter?” published in the American Institute of Physics' journal “Chaos.”
Newman, who teaches in UAF's College of Natural Science and Mathematics, said it’s possible to calculate a system’s optimal size, which can help engineers reduce the chances of system failure and its associated human and economic costs.
Many people currently design large networks with the idea that bigger is better, Newman said. This includes electrical transmission grids, the Internet and many other infrastructures that society depends on.
Newman said that this mentality can lead to failures like the 2003 Northeast Blackout, during which an estimated 55 million people lost power in the U.S. and Canada. It started from one failure cascading into others that spread throughout the system like wildfire. The magnitude of the failures was a function of the system’s size, Newman said.
In the paper, he and his coauthors, physicist Ben Carreras and electrical engineer Ian Dobson, investigated models of different electrical transmission grids and determined that each one had an optimal size. Anything smaller is inefficient. Anything larger is susceptible to cascading failures with scale-tipping costs. The optimal size is calculable using six different factors from their models.
Newman said they looked at the models through the branch of physics that studies complex systems and how its many parts connect and behave as a dynamic whole. Biological organisms, nuclear fusion and the Internet are all complex systems and share similar behaviors such as cascading failures.
Newman said that he and his team will continue testing their theories on different models of varying sizes and hope to apply what they learn to other infrastructures like the Internet.
They also want to look at whether complex systems beyond an optimal size can be divided into sections that behave independently and can keep cascading failures from spreading.
The team needs a larger research budget before they can tackle all these goals, Newman said. There are some times, he admits, when bigger is better.
ADDITIONAL CONTACTS: David Newman, UAF physics professor in the College of Natural Science and Mathematics, 907-474-7858, denewman@alaska.edu. Jason Socrates Bardi, 240-535-4954, jbardi@aip.org.
ON THE WEB: The article, "Does size matter?" by B. A. Carreras, D. E. Newman, Ian Dobson appears in Chaos: An Interdisciplinary Journal of Nonlinear Science (DOI: 10.1063/1.4868393). It will be published online on April 8, 2014. After that date, it may be accessed at: http://bit.ly/1syi7LO
Link to Fairbanks Daily News-Miner article: http://bit.ly/1raLJNv
Link to the Marketplace interview: http://bit.ly/1nDlWLZ