What makes a power grid smart, and can it be hacked?

The ordinary grid was built to move power one way, from big power stations to customers. Information moved mostly the other way, and not much of it. A smart grid is an enhancement of that 20th-century grid that uses two-way communications and distributed intelligent devices, so that both electricity and information can flow in both directions.

Why bother? Because one-way systems are wasteful. The utility has to guess how much power people will use, and keep spare plants waiting: the last 10 percent of generating capacity may be required as little as 1 percent of the time. The first AC grid, built in Great Barrington, Massachusetts, in 1886, was a centralized one-way system, and the main price signal was a dual tariff that charged night-time power at a lower rate than daytime power. Smart meters add continuous communications, so utilities can see use in real time, and demand response lets generators and loads interact automatically to flatten spikes. Some customers go all the way: Trimet Aluminium uses its smelter as a short-term mega-battery.

It also helps the clean transition. The improved flexibility of the smart grid permits greater use of variable renewable sources such as solar and wind, even without adding energy storage, and it can handle two-way flows from rooftop panels and electric car batteries.

The price is a larger attack surface. A smart grid, being online, can be vulnerable to cyberattacks, and with so many access points like smart meters, defending every weak point is difficult. One incapacitated substation could take from nine days to over a year to repair.