Ten years after a major blackout affected at least 50 million people in the northeastern US and some parts of Canada, physicists in Israel and the United States are calling attention to the fact that it could happen again, and that its affects next time could be even worse.
The studies have revolved around spatial networks and their inherent flaw of relying too heavily on central points, or nodes. On the surface, creating a power grid based on geography makes complete sense - at least if you believe neatness and order is the best way to do things. Here, though, spatial networks rely on a number of critical nodes, and if one of those happens to fail, it'll lead to a blackout. According to these researchers, it could also lead to a cascading effect which ends with a complete collapse.
Flickr: Michael Kappel
This is essentially what occurred with the blackout ten years ago. A number of transmission lines went down over the course of a single hour, and thanks to some critical error, it eventually lead to rolling blackouts which headed towards the east coast.
With the problems that spatial networks can cause, at least one researcher believes that randomly-structured networks is the best implementation for something like a power grid, where nodes that go down will result in a degradation instead of a complete blackout. In this sort of event, workers might have an opportunity to remedy the situation before it gets worse, rather than have to deal with a catastrophic failure under intense pressure.
It should be stressed that this is all based on theory, and while it sounds reasonable enough, other researchers have a hard time buying into it. And if randomness for a power grid is in fact the smarter route to take, it's hard to imagine that anything will be done to the current power grid structure given the immense cost it'd undoubtedly require. And that's a bit scary.