Solar and wind energy are great when the sun is shining and wind is blowing but what happens at night or during calm periods? Many experts see the solution in the combination of two approaches. The first is to build a SuperSmartgrid so that the renewable sources compliment one another. For example, if the wind dies down on the great plains, it may be picking up in the Northeast or West coast. The second approach is energy storage. In a recent NREL study of 80% penetration of renewables by 2050 they used pumped storage hydro and compressed air storage in their simulations. However many trend watchers and experts have looked to EVs as the ultimate storage building block for the new SuperSmart Grid.
EVs offer a way to build astonishing amounts of distributed grid support with no capital cost to the grid operators. The capital outlay is provided by consumers for their cars with grid support a useful byproduct. The SuperSmartGrid offers those consumers an opportunity to derive supplementary income from their EVs through demand support and grid support. There are 2 paths to using EVs for grid support and reducing the impact of renewables intermittency.
The first path to is to time charging so it is done at off peak times. Utilities have often offered customers discounts in exchange for operating large loads off peak. So the infrastructure already exists for this type of demand management. It is a matter of applying it to EV charge stations and providing EV specific modifications such as over rides in the event that you have to charge at peak in some type of emergency.
Even more intriguing is the idea that an EV offers substantial energy storage. The concept that is being actively investigated is to charge off peak but to arbitrage some of the energy back into the grid at times of high demand. This is called vehicle to grid or V2G. Battery prices are plummeting and cycling endurance is improving. As a result, what was science fiction only a very few years ago is now in sight as a very real possibility. People are studying controls by which the vehicle can sell back to the grid when demand is high and the grid operators are willing to purchase the power at a premium price. This will work much as the NASDAQ market works today. There is a slightly less sophisticated version of this already being implemented to do demand management as part of a collaboration between Tesla and Solar City.
Car batteries are expensive and they have a limited number of cycles. However, from 2008 to 1012 the cost of Lithium Ion batteries dropped from about $1000 / kwhr to under $500 / kwhr see pg 6 here EVI_GEO_2013_FullReport. As demand for EVs rises there will be increasing pressure to bring the price of batteries down and to increase the number of cycles. Research in this area is vibrant and getting good results. Page 14 of this publication shows the rapid advance in energy density of Li-Ion batteries. It is reasonable to assume that by 2035 EV battery technology will have advanced dramatically and by 2050 it is hard to imagine that problems of range, cost and cycles will not have been solved.
A Quick ‘Back of the Envelope’ Calculation Shows There Will Be Enough Storage For V2G to Work
Lets do a quick look at the impact that EVs feeding the grid could have. Consider 20% penetration of EVs and 200 kwhr batteries. Note that Tesla S Series cars have 85 kwhr batteries today so this seems a very reasonable and safe assumption that it will increase substantially over the next 1 – 2 decades. There are about 250 million cars in the US today so we are looking at 50 million EVs at 200kwhr in each battery pack. If all were fully charged there would be about 10 Terrawatt hours stored. The current NREL simulation shows electricity use peaking at about 0.8 Terrawatts = 800 GW. But since we are assuming a large EV fleet lets make a guess and say that it increases by about 200 GW to enable charging of the EVs. We get this number by assuming that on average the batteries are discharged by half each day. So that takes us to 1 TW. If we assume that the average car owner is going to permit 10% discharge of their car battery by the grid, then that comes out to 1 TWhr of storage available to the utility for managing grid fluctuations. In other words, if the entire grid were wind and solar powered with turbines scattered around the country, and the sun stopped shining and salt in the molten salt storage of CSP plants cooled, and the wind stopped blowing everywhere from the pacific coast to the atlantic off shore turbines, the EVs would have enough storage to provide the grid its power for a full hour. But that worst case will never happen. For example, if demand is high, price paid for electricity will skyrocket and the majority of EVs will be selling to the grid, not buying. Further, because the price will be high they will willingly discharge by more than 10%. Also, there is no precedent for wind to stop everywhere. The storage in molten salt will ride through the peak so it will contribute as well and there will be other sites with storage and production such as hydro. So actually, without rigorous study it is pretty safe to assume that the EV fleet at 20% penetration could ride through for hours of highly unusual drop-outs in available renewable energy.
Even before turning the analysts loose on this, it is pretty clear we are in the right ball park. By 2035, it is hard for me to imagine that EVs will only make up 20% of the car fleet. The trends in cost reduction are powerful and people will begin to realize that EVs, that take only minutes to plug in and unplug, are far more convenient than autos that require trips to the filling station in all kinds of weather. Also, adding to the downward cost of ownership, there will be a vibrant after market for EV batteries. They will be transitioned to grid support, and the grid will increasingly have demand management. Long story short, from a technical stand point it is pretty clear that V2G passes the technology litmus test and one can see that it is reasonable to consider it as a likely option enabling deep penetration of renewables on the grid.
It is also worth noting that the EVs present a great opportunity for demand management offering further grid stabilization. So reasonable assumptions suggest that the emerging EV fleet will be fully capable of providing regulation on the grid to manage fluctuations in renewable energy. This, of course, assumes that the power transmission grid is upgraded to allow wheeling of power throughout the continent.