To Keep Electric Cars Powered, the Grid Needs To Be 'Omniscient'

Consider an average weekday afternoon in the electric car-dominated future of some typical California suburb: many thousands of plug-in cars returning from typically brutal commutes with batteries sliding toward empty. Some of them will plug in at home, using higher voltage clothes dryer outlets or even 120 V regular wall sockets (for those that don’t mind waiting actual days for a charge), while others will stop off at public charging stations with “level 3” charging ability, e.g. the ability to juice your Leaf up to 80 percent in less than an hour. Whatever the exact charging source, the basic result is a whole lot of high voltage outlets delivering at maximum capacity in a short amount of time.

If you really wanted to piss off an electric grid that's just what you'd do, plug in a bunch of current-sucking devices at once. Look at heatwave brownouts, for example, as highly concentrated groups of people all crank their air conditioners in unison and keep them cranked. This results in equipment failures within overtaxed infrastructure—tons of bonus current overheating and blowing up transformers, for starters—and as a result, you find power companies dropping the voltage available to the grid to protect the stuff that makes said grid work. That's a brownout.

The grid is more or less premised on the notion that your home can survive well enough on 120 V, and almost all of your outlets deliver this voltage. The house itself receives around 220 V and it’s common for homes to have one single special outlet that delivers all of those volts. This is typically with a clothes dryer in mind, but it’ll get more and more common for electric car owners to add another one of these higher voltage outlets somewhere for their vehicles. So, imagine entire cities all running their clothes dryers at once. It’d put a New York heatwave to shame. The electric car future is a looming problem that makes utility companies anxious, and will demand a solution more clever than upgrading entire physical grid infrastructures.

A team of researchers at the University of Vermont just unveiled one possible solution that, naturally, depends on the future’s grid also being “smart.” The idea, reported in the engineering journal IEEE Transactions on Smart Grid, is based somewhat on how information gets around the internet: packet switching. Packet switching is basically the packaging up of streams of information into discrete units that can be sent around a network with maximum efficiency. Instead of flushing a bunch of bits of data from many different streams at once through a network, with packets it becomes possible to manage that data as you would cars or some other physical thing that can be “hopped” place to place within a network via switches (or traffic signals).

The smart-grid offers an opportunity to do that with electricity, packaging it into discrete time periods of unrestricted current flow. These packets of flow can be swapped around the grid in ways that maximize the amount of voltage being delivered to a household while ensuring that equipment isn’t overtaxed. Essentially, the smart-grid version of your home’s electrical meter fires out its power demands into the grid, which obliges, but only in 10 minute or so fragments. After each burst of charging, your outlet in the grid gets “back in line.” And this repeats until overall demand drops, and everyone can again have unfettered access to the 220 V that powers their cars and dryers.

That might not sound like a great deal as a consumer, but the UVM proponents note that most commuters park in the evening and don’t leave again until the morning. So having your car fully charged at, say, 1 a.m. vs 5 a.m. is more a function of having a blind grid than needing a charged car in the middle of the night. “The vehicle doesn't care. And, most of the time, as long as people get charged by morning, they won’t care either,” notes UVM’s Paul Hines, a co-author of the report. “By charging cars in this way, it's really easy to let everybody share the capacity that is available on the grid.”

Of course, in such a system a driver could always pay a premium to cut to the front of the line (or just visit a charging station). “We assumed that drivers can decide to choose between urgent and non-urgent charging modes,” the researchers note in a press release. Packetized charging would function as sort of a discounted service compared to the “as much as possible as soon as possible” most of us take for granted now. (Though adding another tier to the already highly tiered world of car technology is … uncomfortable.)

The UVM team even has a way around the privacy concerns of packet power. Maybe you aren’t very into the idea of alerting your local electric utility to your daily driving habits. Such a privacy reveal might not float well with the (screwy, possibly backwards) ethos of driving as a unique source of autonomy. So, instead of the packet system operating house by house, human to human, it would be determined by an algorithm known as an “automaton,” which, instead of delivering power according to what you yourself ask for, determines/predicts your needs via finely tuned probabilities. The result: “omniscient centralized optimization,” a grid that both knows more and less about you at the same time.

The concept only seems realistic because something is going to have to happen for electric cars to scale up. The alternative is brownouts and/or a new civil army of linemen on call to replace grid equipment, both of which mean bad things for rates in general and for the electric car future. “And the problem of peaks and valleys is becoming more pronounced as we get more intermittent power—wind and solar—in the system,” says Hines. “There is a growing need to smooth out supply and demand.”