Grid, Ready to Talk

September 05, 2008

Perhaps the most potent piece of the developing energy order will have been the smart grid. The momentum for renewables gathers like the winds of war. Big automakers like Toyota are pushing up their plans to deliver the plug-in. Little automaker wannabes are everywhere. But the clincher, the deal closer, the thing that draws it all up in a bow, may well be the smart grid.

Rocky Mountain Institute is helping to shape the paradigm shift with their Smart Garage effort, the cruxes of which are:

• the storage devices of electrical vehicles could potentially harbor more power than all powerplants connected to the grid
• electric vehicles could do much to stabilize the grid: eliminating blackouts and counterbalancing peak demand (feeding their stores to the grid during peak)
• smart vehicles on a smart grid could recharge when it's smart to do so (night charging)
• "wind turbines built to serve the night car-charging market would still spin whenever the wind blew and may eventually be able to meet a considerable portion of America’s electricity demand"
• the technology that was, even a decade ago, non-viable is emergent: electric vehicles and battery storage, building integrated renewables, digital wireless internet stuff

Talk to me, Grid.

Photo: Carla Dal Mas on flickr

Comments

[stomv]

I love smartgrid.

But, there's different levels of smart. The grid can accomplish a whole lot by simply communicating variable pricing in real time, allowing the users on the grid to be smart.

For example, consider the PHEV. The user programs the car to make sure that if it's after 10pm, it must be charged by 6am. The vehicle only needs 4 hours to charge, but it can choose from all 8. Which does it choose? The 4 cheapest. How does it know which 4 hours will be the cheapest? It doesn't *know*, but it uses historical information which it's learned from the grid to make a good guess. The price of electricity is cheapest when supply is cheapest, due to low demand. This allows the users to shift their demand based on price, which will over time level the price overnight. If the wind is blowing, more electricity has a marginal cost of $0, so the price goes down -- and more chargers kick in to counterbalance.

To prevent surges, each user's device is programmed to have random time intervals thrown in so some cars start charging 1-20 minutes early, others 1-20 minutes late. Otherwise, all the Prius PHEVs might kick in at exactly the same time, and that'd bring the grid to it's knees.

Other than using plug in cars to balance demand, what else can be balanced? Dishwashers, which need to run overnight but don't have to start just before bedtime. To a lesser extent, air conditioners and heat pumps could work outside the thermostat range by a degree or two if the price was attractive enough.

As mentioned above, this whole process can also work in reverse in the case of PHEVs. Price of electricity is higher in the day, and often peaks at 3pm. If the PHEV only needs 20% to make it home, there's no reason why it can't sell 50% of it's charge during the day to the grid at high prices.

This is trickier though. For one thing, daytime prices are far more variable than overnight prices, so timing when to sell is trickier. If all the cars sell earlier in the day, there'll be no help at the high part of the peak. If all cars wait for the peak, they might shave so much of it that some could have made more money selling earlier. This isn't a showstopper, but will require some skill. The trickier problem is when I'm not going straight home from work. In my example, I kept 30% "of a tank" in reserves. What happens when I forget to change my charge schedule because I'm driving out-of-state after work? I get in my car and it's only got 50% charge because it sold electricity late in the day, and now I've only got half a tank. That's going to burn when it happens. By giving the charger's unique IDs and names we can avoid the weekends/holidays/vacations problem because the car can know to only sell when plugged into workplace chargers.

The good news is that the early adopters of this program will drive for free. After all, since nighttime electricity is so much cheaper than daytime, they'll be buying low and selling high every single day. They're not the only ones who benefit of course, because all ratepayers will see a lower price during the day when they're likely to be using more electricity and a higher price overnight when they use less. Of course, the more devices "playing the market" the flatter the S-D curve will be, resulting in smaller and smaller advantages, and PHEV owners will find that charging overnight and selling during the day provides only a few cents of profit instead of the dollars it used to. That's OK -- at that point we'll have a grid that can handle large amounts of variable green energy [wind, solar], has far more predictable supply-demand curves, and has stable real prices.

As an added bonus, since real-time pricing will make electricity far more expensive during the day initially, there'll be a huge incentive for homeowners to install programmable thermostats so that they're not using so much electricity for climate control during the day when (a) they're not home anyway and (b) it's so expensive. These will be pure negawatts -- pricing signals leading to voluntary conservation.

Well have reduced non-auto demand, and we'll be using auto demand to shape the S-D curves, allowing for far more wind and solar installations. As this progresses, it will be easier to dismantle oil and coal burning power plants. We could easily get to the point where we're using less gasoline to drive and less coal to make electricity... that'll be a big step in managing climate change and foreign trade policy.

[Chris]

stomv, absolutely, the different levels of smart that come with the smart grid is what makes it exciting. too many levels to capture in a bite sized post. cars. dishwashers. people with cars arbitraging their power sales and purchases (maybe, I don't pretend to understand how this would really be implemented). power companies shutting down non-essentials at peak or to avoid brown-outs. all this and more makes it seem that the smart grid could be the heavy hitter out of all that is bubbling up out there: there's so much it could do, and it seems it could be implemented easily and inexpensively (relative to other initiatives anyway).

sort by Grid or Energy efficiency categories to see our recent posts on the smart grid. let us know if there are other twistss on the smart grid worth discussing.

stomv,as always, thanks for your insightful comments.

[stomv]

For me, the key is this:

The power grid doesn't control what individual users do. The power grid merely provides the real time pricing, and individual users make (automated) choices which best benefit the power company because it smooths the S-D curve, and best benefits the individuals because these voluntary measures save/make the individuals money.

The key is that it's all free choice -- neither the power company nor the government is forcing anyone to do anything.

[Chris]

A challenge will be to make pricing fair and transparent as we depart from single rate systems, but there should be rich reward for all participants (individual users, utilities, and the country at large) when we do.

You can imagine explosive, viral adoption if the system is set up to be transparent, based on free choice, and offers appropriate incentives to both end users and utilities.

[Michael Jung]

Dynamic pricing, demand response, PHEVs...it's very exciting to consider the wealth of revolutionary energy services and applications that would be possible with a smart grid. However, a key point that is too often overlooked is the importance of building robust, interoperable, and cost-effective networks as the foundation of the smart grid. We must use open standards and recognize that latency matters more than bandwidth when it comes to the smart grid. End-to-end Internet Protocol (IP) will enable universal interoperability for all the applications we can imagine today, plus leave the door open for the killer apps that we haven't even invented, much less named!