Big Data + Big Batteries = Smart Grid?

Peter Judge

The smart grid can’t do much unless it has analytics and batteries, says Peter Judge

Can a smart grid really save significant amounts of electricity? South Korea seems to think so, as it was recently reported that smart meters there could cut demand by enough to avoid the need to build a nuclear power plant.

I have my doubts about the story, which comes from Bloomberg. For one thing, smart meters are good at removing fluctuating demand, and nuclear plants are good at providing a steady supply.

Smart grid cuts load

When a lot of electrical equipment is switched on, say when a the nation makes a cup of tea during a TV commercial break, that can increase demand and force the utilities suppliers to switch on generators – and the power plants that can be that responsive are the ones fired by oil or gas.

Smart meters monitor devices that are switched on, and can be used to smooth out demand by turning off non-essential kit. So if South Korea really does change the pattern of its demand using smart meters, it is far more likely to avoid the need for extra fossil fuel-fired  capacity.

Which is fine of course, because fossil fuel is the climate change baddie and nuclear doesn’t produce greenhouse gases. I think the Korean government mentioned nuclear here because its public image is somewhat tarnished by the fate of the Fukushima reactors in Japan.

But seriously, the story makes me wonder just what potential there is for smart meters to cut energy use. I’ve had some limited experiments with devices like the Owl and the more sophisticated network connected kit from AlertMe,  which you strap onto your power supply at home, and I’ve found a few things.

These devices are a hassle to use. For some types of people, it’s fun to watch your power usage, and see which bits can be switched off, but it’s not an exciting hobby.

Even when you have optimised your electricity, it’s only going to take a few percent off your normal usage, and the figure will creep back up again if you don’t keep on top of it.

It’s not for everyone, and the effects are limited. Add in the fact that proper smart meters will cost around £500 per household, and you have to wonder if it is going to be worth it.

A giant virtual battery…

“The kind of people who get smart meters  first, are the kind of people that want them,” agreed Mark Perrett, a consultant in analytics HP, when I met him a couple of weeks ago.

He thinks the benefits will come when two other things happen. Firstly, smart meter data is made available and shared, so that users can pay a third party to do the optimisation for them – and the utilities can apply centralised management and Big Data analytics to reduce overall demands.

Suppose the utilities made their tariffs vary infinitely according to the level of demand. Then higher prices might put off consumption and avoid the high peak demands that rqeuire more power plants to be switched on. Smart meters could automatically reduce demand when the grid tells them those peak loads are occurring.

But how much leeway would the meters have to reduce demand? Remember, in my own experiments I found at a given time there was relatively little difference I could make.

Well, suppose the house had its own energy store. When demand is high and the energy price goes up, then the house could stop using grid electricity and draw form its own local battery, until the price goes down.

Will there be a lot of these local batteries around though? In fact, there may well be. If plans for electric cars come to fruition, then a lot of houses will have cars with large batteries that spend a lot of time sitting by the house, plugged in and charging. Reverse the flow to the battery and they could power the house for short periods to reduce the load on the grid.

Utilities have long felt the need for large amounts of storage that could offset the need for peak generating capacity. The giant Dinorwig hydroelectric power station in Wales is actually a pumped-storage system, in which water is pumped up to the reservoir using off-peak electricity, so the station can generate more electricity for peak demands.

Combined together, and using Big Data analytics, the nation’s electric car batteries could add up to a virtual Dinorwig electricity storage system. In a big enough scheme, the economies of scale could make the electric cars affordable to the masses.

That’s the concept, but the technical and financial details still need working out. Most electric cars might not be fully charged at peak times when they are needed – but if it would save them money, people might be encouraged to shift work times so they can get their cars parked and charged up in time to act as batteries for the evening demand surge.

If it all gets put together right though, it looks as if smart meters really can make a difference.

Are you the type of person to monitor a meter? Take our Green Tech Quiz!