Batteries Too Costly for Grid Energy Storage: Analysts


Despite all the attention on battery technology for energy storage for the power grid in the latest round of smart grid stimulus funds, current battery technology is just too expensive to be mass deployed right now, according to analysts at Black and Veatch. “Unless battery technology makes dramatic advances soon, B&V doesn’t see them being much of a factor in helping states like California meet their RPS goals looking out to 2020,” the firm told us in an email. Here’s a handy-dandy chart showing just how much cheaper compressed air energy storage (CAES) technology, flywheels, and pumped hydro are (further descriptions of these technologies here) compared to battery technology (NAS stands for a sodium sulfur battery, which are some of the cheaper forms of batteries already in use for the power grid):



First, concerning a critique above about Li-ion batteries, lithium is a rare earth metal. But the definition of “rare” in this context means “not as common”. There’s a lot of lithium. Like other minerals there is an incremental cost of production of lesser grade ores. More importantly, there is so much of it already found to be economical or nearly economical at or near today’s market price there has been very little exploration for the material. That is of course changing and likely new, high grade (read: low cost) deposits will be found.

Second, regarding the B&V report, I never quite understood how compressed air worked. At best, air compressors are, what, 85% efficient? Then turbines recovering the compressed air are, what, 85% efficient as well? So even before you begin to worry about finding relatively leak free underground caverns that are not heavily contaminated with methane to pump air in and out of you are starting with over 30% power loss just in the compression cycle.

Black and Veatch should be a little more concerned about their image as a legitimate engineering firm.

Bob Wallace

4 Comments + 3 Linkbacks /= 7 Comments….

Batteries might drop significantly in price. Sodium ion batteries with water-based electrolytes.

There are vast numbers of underground resources for installing CAES. From salt domes to limestone caverns, to played out oil wells, to empty/partly empty aquifers. There are few places in North America that do not have potential CAES sites.

Pumped hydro. There’s a company called Riverbank which is in the process of building pump-up storage by creating underground caverns (looks like they intend to use a tunnel boring machine). That makes pump-up usable where there is some water available and it could be located close to available transmission lines.

Bob Wallace

In 2007 the US Departments of Interior, Army and Energy released a review of 871 dams on federal land which are not currently used to produce electricity. Of the 871 they found that 195 (22%) had a adequate elevation (head) and were reasonably close to existing power transmission lines. Of that 195 a total of 71 (8% of the original 871) were located outside of protected areas (national parks, wilderness areas, etc.) and had adequate generating capacity (enough year around flow) based on actual hydrological records to make them good candidates for power generation. The 14% which had enough head but not adequate year round flow should be good candidates for pump-up storage.

In the US we have approximately 80,000 dams with about 2,400 (3%) used for power generation. We get 20% of all our electricity from those 2,400 dams.

If we use the federal findings as an indicator we might suspect that over 6,300 dams might be available to supply power to the grid and another 10,800 might serve pump-up storage purposes.

Furthermore, by using a technique called hydro uprating we can create 100% efficient “storage”. By taking a dam with adequate flow to supply one turbine and installing multiple turbines we can let water accumulate behind the dam when demand is low (the wind is blowing, the sun shining) and then when demand exceeds supply all the turbines can be put in operation to bring a lot of power on line quickly. This creates pump-up storage with out the energy loss from pumping by letting flow of the river recharge the energy supply.

At least two of the dams identified in the 2007 study are in the process of being converted to power producers and work underway on at least four others.

Charles R. Toca


“The 14% which had enough head but not adequate year round flow should be good candidates for pump-up storage.”

Do you know how many of those are in California?

Do you have any opinion of benefits of distributed energy storage by batteries v. centralized storage – hydro, etc.. One study I saw by AEP indicated a 10% savings in line losses by using distributed battery storage to store power at night when congestion on transmission wires is less.

I believe we’ll be more successful installing many megawatts of battery storage – flow batteries – near load then we will be trying to permit and build more central plant pumped hydro facilities.

Bob Wallace

First, I made a BIG mistake in my previous post. We don’t get 20% of our electricity from hydro, it’s more in the 6-7% range. I was looking at the unused dam and a lot of ‘run of the river’ potential identified in a 2006 DOE study and projecting that we might get as much as 20% of our energy from hydro. And then didn’t do an adequate job of proofing my post. Sorry.

Now, I don’t know how many of the 871 dams on federal land are in CA, but all the sites are listed in the linked article. Nor do I know how many of the non-used 80,000 dams are in CA. But I do know that we have dams that don’t run full speed year round due to lack of incoming water. We certainly could create some ‘dry season’ pump up storage in many of our dams and the dry season is when demand is highest. Night wind could be stored to power peak hour air conditioners.

We shouldn’t limit ourselves to thinking about CA as an isolated place. We’ve got a HVDC transmission line (the Pacific Intertie) running from Southern CA to the Northwest hydro facilities. And we’ve got the Intermountain Intertie running from SoCA to Utah. And plans underway to extend from Utah to bring abundant Wyoming wind to the Pacific Coast.

We’ve got some pump up storage being installed in the greater area. (I think there’s a SoCA project underway, but I’m not sure about that.)

Right now, I suspect pump up is less expensive for long term power shifting (night to day) than are batteries. But that could quickly change. There’s a storage battery company converting a large ex-Ford plant into a battery manufacturing plant. They must have a product that they think is going to be competitive.

10% loss savings seems high to me. But if it’s true then I’d think there’s money to be made by creating storage. Storage is going to likely be helpful both closer to point of generation and point of use. It means that transmission flows can be more moderate, making for smaller transmission lines.

Charles R. Toca

There must be more to the B&V report, otherwise this is a silly conclusion.

I agree with the first two posters – flywheels for energy storage? Beacon flywheels are being built for $3,000 per kW, and they are $12,000 on a per kWh basis – and you only get 15 minutes of energy.

CAES is enhanced natural gas generation – assuming one can find a spot to build. What is the life-cycle cost, including fuel? And what’s the point in using natural gas generation to “store” renewable energy?

Pumped hydro is the best storage – but not practical given environmental and siting considerations.

NAS and flow batteries are the best solution for grid connected storage, and flow batteries are best for multiple cycle requirements of renewables.


I think you may have cropped the graph before posting this image – there’s no legend describing what the different colored sections of each bar are.

Any direct link to the Black and Veatch report or to an un-cropped version of the graphs?


I’d like to see an analysis of energy storage that presents the associated costs in a manner consistent with the way utilities and IPP’s analyze the generation systems that the storage will either augment or supplant. Namely to look not just at the capital costs but at the lifetime delivered cost of electricity on a $/kWh basis.

As the previous poster mentions, the B&V analysis isn’t worth the paper it’s printed on. Comparing Flywheel and NaS on a $/kW basis is uninformative at best and deceptive at worst. NaS is a seven hour energy system whereas the flywheel has extremely long “cycle life” but no energy density.


analysts reports really aren’t worth the paper they’re written on! Who are the looking to ‘short’ with this report?
The whole idea is that a gradual roll out of battery mass storage along side BEV volume growth actually accelerates the downward curve on battery prices by helping to raise production volume early. To help BEVs reduce in price utilities can not only buy new battery storage but later in the cycle they can also buy the used 10 year old battery packs that are down to 80% DOC. That will also cushion the price of new battery packs.
Finally, if anyone tries to argue that this adds a huge cost burden to utility bills, simply remember that mass storage isn’t just an ornamental green gesture, it can save utilities from not only having to fire up axillary power generators to serve peak demand but from the need to build new peaking facilities both of which reduces the cost of power generation.


The economics of batteries don’t follow the normal model with economies of scale anyway. Batteries are made of metals – generally rare earth metals like cadmium and lithium. As demand goes up so does the price of the metals and the price of the battery. Unless you can make a battery out of sand and water, relying on them for storage or mass production is a waste of time and money.

Comments are closed.