Electric cars and the grid

This article was original published with Plug-in-America, the voice of plug-in vehicle drivers across the country.

One of the things that we hear as owners of electric vehicles (EVs) is that we’re just moving pollution from the tailpipe to the smokestack of a coal-fired power plant. Is this a fair complaint today and does it have to be in the future? What if there are 100 million EVs driving around the US? Can we charge them all using renewable power?

In this post I’ll show that not only can renewable power like wind and solar provide the energy we need, but EVs actually can increase grid stability and ease our transition to a carbon-free electrical grid.

How much power will be needed to charge all of the EVs on the road?

Tesla Model 3

Chevy, Tesla, Nissan and others are soon planning to offer EVs with a range of at least 200 miles costing less than $40,000. Bloomberg did a study of the growth of EV sales, predicting an explosion in growth and that by 2030 there will be 100 million EVs on the road. When Tesla began accepting pre-orders for their Model 3, over 350,000 people made a down payment of $1,000, suggesting that Bloomberg’s prediction are reasonable.


How much electricity will it take to charge 100 million EVs?

The average American car is driven about 12,000 miles per year (33 miles/day). An electric car today gets about 3 miles/kwhr. That works out to 4,000 kWh/year (for comparison the average electricity consumption per household is 11,000 kWh/year). If we look at the year 2030, when there are projected to be 100 million EVs on the road, this works out to 400 billion kwh per year, which is about 10% of the 4 trillion kwh that the US currently consumes. Meeting this demand over the next 15 years while reducing climate pollution will not happen without preparation, but it is not an insurmountable task.

Wind Power

wind power

If utilities were to meet the needed 400 billion kWh per year using just wind power, they would need to install 138,000 MW of wind production. If we just keep building wind power plants at the same rate as we did in 2015 (8,600 MW of new wind capacity), the utilities will hit that target in 16 years. If we instead build wind farms at the record rate of 13,000 MW (2012), then it would take only 10 years to meet the projected demand from EVs.

Another way to look at it, if we captured all of the commercially viable wind power in North Dakota, that would generate twice the amount of energy needed to power 100 million EVs. That’s assuming no improvements in wind turbines over the next 15 years.

Solar Power

Currently there are over 22 GW of solar capacity installed in the US. Assuming that the panels produce power 25% of the time (i.e. factoring in night and weather), these panels produce 12% of the power that will be needed to charge 100 million EVs. Deloitte studied a variety of growth models for solar power, which resulted in between 100 and 600 billion kWh per year of PV power in 2030. The upper limit just assumes that the growth rate we’re currently seeing continues. The demand created by EVs falls within the uncertainty of these predications.


Managing Intermittent Power

Electric utility operators have a difficult job. At every moment they must precisely balance the energy coming into their grid from generators with the energy leaving the grid to power air conditioners, computers and everything else that’s plugged in. If you turn on a light switch and nothing else changes, a little more power must be put into the grid.

Power plants can be broken into three categories:

  • Baseload (e.g. coal and nuclear) that run at a constant output
  • Dispatchable (e.g. hydro and natural gas peakers) that can easily be turned up and down as demand requires
  • Intermittent (e.g. PV solar and wind) where the operators have no control

As our grid becomes more dependent on renewable/intermittent energy sources, managing these intermittent power sources may be increasing difficult. Currently if the wind isn’t blowing and the sun isn’t shining utilities respond by turning on a dispatchable power source, which is probably a natural gas plant that emits CO2. But there are other options.


The most obvious one is to store electricity:

  • pump water up a mountain when you have excess power, run it down the same mountain through a generator when you need power
  • charge a chemical battery like the Tesla Powerwall
  • spin up a flywheel
  • make ice that will be used later to cool a building.

These storage options are all fairly expensive and would add significantly to the cost of renewables, but the cost is decreasing and storage will be an increasing part of the tool set that utility operators use to manage the grid.

Demand Response

This is where EVs come in. Rather than respond to the imbalance between supply and demand by changing supply, you can change the demand. Currently utilities have almost no way to influence demand, but the Smart Grid is all about giving the utility operators a mechanism to do this. Imagine a future where the utility can send out a signal that the price of power is high (demand outstrips supply). Some appliances (e.g. deep freeze, electric water heaters) read that signal and go into a low power mode. When supply outstrips demand, the price of electricity goes down and your appliance goes into a high power mode.

Some appliances, like your computer or TV, aren’t well suited to this, but charging your EV is. Imagine you have a 200-mile range EV and a level 2 charger at home. You arrive home at 6 pm after driving 80 miles and plug in your car. Since demand is high at that time, your car (which still has 120-mile range) doesn’t start charging. Even if you go out at 7 pm, you still have plenty of range for your evening activities. Around midnight when demand has dropped so low that the utility has a challenge finding customers for its baseload and wind power your EV starts charging. By 4 am your car is fully charged. This encourages the development of more wind power, since the operators know there will be customers for their power day and night. By helping the utility integrate more intermittent sources of energy, your EV has helped the environment twice, by reducing fossil fuel consumption at the pump and at the power plant.

Time-of-Use electricity rates

Time-of-Use electricity rates get most of the advantages of Demand Response with a much easier implementation. For instance, PG&E (a large California based Investor-Owned-Utility), has a time-of-use plan with three rates (off-peak, partial-peak, peak). In the summertime electricity is cheapest from 9:30 pm to 8:30 am, when demand is lowest. The price peaks from noon to 6 pm. This encourages users to move consumption to when the there is excess generation capacity.

We may get enough of the advantages of demand response just by implementing time-of-use plans. It’s not as precise (there’s no way to factor in whether or not the wind is blowing at any moment). The advantage is there is no requirement for communication between the utility and the consumer and it’s completely predictable. It does require an upgrade to the electric meter to one that knows what time it it, but that’s a relatively minor change than what’s needed for true demand response.


The concern that EVs just move the pollution from the tailpipe to the power plant does not have to be true. Today’s technologies deployed at the rate that we’re currently deploying them could easily power 100 million EVs in the USA by the year 2030. And if these EVs are connected to the Smart Grid or time-of-use metering, then they actually make the transition to a low-carbon grid easier by allowing the utilities to power them when electricity is cleanest and cheapest.


Leave a Reply

Your email address will not be published. Required fields are marked *