Researchers collaborate so smart grid communicates

 

Dave Bakken speaks at the Smart Grid World Forum 2011 in Beijing, China, earlier
this year.

 

 

PULLMAN, Wash. – When the lights went out on the East Coast on Aug. 14, 2003, the initial failure was caused by trees in power lines. But as the outage cascaded, data communications among power operators became the key issue, said Dave Bakken, associate professor in the School of Electrical Engineering and Computer Science at Washington State University.
 
“Problems in providing the right information to the right place within the right time’’ is how it was described in a report by Frances Cleveland, an electric power utilities consultant.

 

Ultimately, 55 million people were affected, from Cleveland to New York, Boston and Canada.

 

 

Bakken and his WSU colleagues are leading efforts to improve communication on the power/smart grid. “Smart grid” generally refers to a class of technology used to bring utility electricity delivery systems into the 21st century, using computer-based remote control and automation.

 

Because WSU computer scientists are in the same school as the top-ranked power engineering program, interdisciplinary research is readily leveraged to address one of the smart grid’s greatest challenges.

Earlier this year, Bakken gave a presentation at the Smart Grid World Forum 2011 in Beijing, China. The prestigious gathering aimed to share information about smart grid research and construction. At Bakken’s plenary session, one of the other presenters was the CEO of General Electric, Jeff Immelts.

 

“Professor Bakken and his group are leading experts on wide-area data delivery issues and mechanisms that can make power grids much smarter,’’ said Stefano Galli, a member of the  IEEE Communications Society’s Board of Governors, who invited Bakken to present. “His plenary presentation at this very important smart grid forum remarked the fundamental role that communications play in the smart grid. It was important and very well received.”

 

The communication structure for the power grid came about in the 1970s, Bakken said, in response to a massive East Coast blackout in 1965. The system, called Supervisory Control and Data Acquisition (SCADA), was meant to allow rudimentary data collection within an electric utility.

 

Sensor data is collected every few seconds – an eternity in computing time. If a disruption happens, the power grid may not be able to react fast enough.

Technology for computer networking and wide-area data delivery (software called “middleware”) has undergone radical changes.

 

 

Concerned about costs and constrained by regulators, however, utilities have tended to make upgrades to computer technology on a slow and piecemeal basis. This approach has made it particularly difficult to improve communications between electric utilities.

 

Working closely with leading power researchers at WSU, Bakken is developing a communication infrastructure to be implemented as part of the smart grid.

 
Or, as Bakken irreverently says in his e-mail signature: “Since 1999, cheerfully and audaciously dragging the wide-area data delivery services of the electric power grid – kicking and screaming – into the mid-1990s.”

 

Bakken takes a light-hearted approach, but he is serious about having a better communication system for the power grid widely deployed in the next 5 to 10 years, even in an industry that many call “the most conservative of conservative industries.’’

 

“The continued piecemeal expansion of the power grid’s communication system is unnecessarily expensive and does not even meet today’s requirements, let alone tomorrow’s,’’ he said.

 

 

Bakken, who is a WSU undergraduate alumnus (’85), returned to work at WSU after doing research on wide-area data delivery systems at BBN, a research laboratory that  built the first Internet in 1969.

 

When he arrived at WSU in 1999, Bakken was approached by Anjan Bose, a leading power engineering professor who is a member of the National Academy of Engineering. Bose recognized the communication problems of the power grid, and the two decided to “re-think grid communications from the ground up,’’ Bakken said.

He began a collaboration of more than a decade with Bose and Carl Hauser, an associate professor, WSU undergraduate alumnus (’75) and computer scientist with 21 years of industry research lab experience.

 

Bakken and Hauser are rare among computer scientists in that they spend most of their time working in the power engineering domain. Bakken often visits utilities and, when he goes to other universities to give a talk, he is more likely to talk with power engineering departments than to computer science professors and students. Sometimes he will give different invited presentations to both groups.

 

Such a collaboration between computer scientists and power engineers is “very rare today and unheard of when we started in 1999,’’ said Bakken.
 
The two groups of researchers use different vocabularies and generally travel in different circles.

 

Acquiring research funding can be a challenge. The U.S. Department of Energy doesn’t tend to fund communication research, while the science funding agencies like the National Science Foundation have assumed utilities or industry would pay for such research innovations.

 

In spite of the challenges, the unique team has successfully attracted a large amount of funding support over the past dozen years.

 

“There are great problems and great opportunities,’’ Bakken said.

 

The researchers’ efforts have led to development of a new data delivery system, called GridStat, that is separate and above the electric controls of the traditional power grid.

 
“We’ve shown that it’s possible to achieve and to provide low latencies (delays) over the whole grid at a millisecond or so over the speed of light with high availability at a lower cost for developing and evolving power application programs,’’ Bakken said. “This enables many new kinds of control and protection strategies to be deployed.’’

 

But, he added, it’s a “quantum leap’’ that utilities are going to have to buy into. Utilities have begun such efforts, at least in one case: a national initiative to use sensors that are called synchrophasors.

“It’s a disruptive technology, not evolutionary,’’ Bakken said. “But we’re in this interdisciplinary endeavor for the long haul.”