Thursday, July 21, 2011

The Billion-Dollar Branch: From flashover to blackout

Power blackouts begin with a single triggering event, like relays failing, or a heat wave that causes a transformer to explode, or a key power plant going down. These are single-point failures and usually don't lead to monster blackouts ... unless they're helped along by more failures, errors, and missed signals.

Trees that have grown too close to high-voltage lines (called "grow-ins" by the industry) are a common contributing cause, because these can lead to massive short circuits. The risk escalates during extreme weather events. Prevention of grow-ins costs billions of dollars every year.
Flashovers from powerlines to trees were a contributing cause of the August 14, 2003, Northeast Blackout. It started about noon, with a monitoring tool called a state estimator being accidentally left off line. This, along with a previously unknown software glitch in an alarm system (called a "race condition"), combined to reduce the operators' awareness of what was about to happen. At 1:30 pm, FirstEnergy's Eastlake, OH, generating station went off line. This redirected the flow of power around the region and beginning at 2 pm, caused high-voltage lines in three areas to overheat, soften, and sag toward tree branches below. Flashovers followed when the electricity arced to the tree, then to the ground. The huge flow of current caused protective relays to trip and each line went out of service in turn. These events set up a positive feedback loop (a bad one) that progressively put more strain on the system. Other problems followed and two hours after the first tree flashover, the blackout was fully underway, flicking off the lights for 55 million people. 

Why not trim every branch as soon as it poses a hazard? There's always room for improvement, and more assiduous branch-lopping was a key recommendation from the 2003 blackout lessons-learned reports. But nationwide there's almost 180,000 miles of high-voltage right-of-way, and it's not possible to know and respond to every hazard. That's because the problem is dynamic. Lines sway; trees sway and fall over. Branches grow toward the right of way, since there's more sunlight there.

Winter poses its own problems; here's my last post on the southwestern US freeze-out in early February, which set off a wave of generator outages and rolling blackouts across Texas.



3 comments:

  1. Tree trimming near power lines also has risks. Here's what happened when tree trimmers accidentally had branches break lines, that then fell on lower voltage lines. And, the reclosers, a form of circuit breakers, had been removed for testing. Quite a light show:

    http://youtu.be/lEAtWiUGeiA

    On the local news:

    http://youtu.be/dRnwSmDWG0E

    Blackout AND damage.

    ReplyDelete
  2. Thanks - good addition! I've seen the effect of reclosers, but haven't looked at one up-close.

    ReplyDelete
  3. Some history of tree arcing causing major blackouts:

    July 2, 1996: A wire sagged from heating due to heavy load, causing it to arc to a tree. This started a chain of events leading to widespread Western US blackouts.

    July 3, 1996: The same thing started all over! On the same power lines, again, they sagged into a tree, even though the operating conditions of the western power grid were different. This time, operators stepped in, and did manual power load shedding, preventing another widespread blackout.

    August 7, 1996: Power system breakdowns caused a chain reaction of power failures in Indiana, made worse when an overloaded power line sagged into a tree, causing that line to shut down.

    August 10, 1996: A 500kV power line in Oregon arc to a tree, causing another series of line shut downs, made worse by 3 other locations having tree to line arcing, from heavy load line sag.

    A quote: "BPA’s right-of-way maintenance was inadequate. Consequently, BPA’s failure to trim trees and remove others identified as a danger to the system caused flashovers from and the opening of several 500 kV transmission lines, the last of which led to overloads and cascading outages throughout the Western Interconnection."

    http://www.nerc.com/files/disturb96.pdf

    I remember experiencing one of the above outages.

    ReplyDelete