July 08, 2007
NaS Batteries Used For Large Scale Storage

Sodium-sulfur (NaS) batteries have begun to enter service for large scale stationary electric power storage.

An NaS battery, by contrast, uses a far more durable porcelain-like material to bridge the electrodes, giving it a life span of about 15 years, Mears says. It also takes up about a fifth of the space. Ford Motor pioneered the battery in the 1960s to power early-model electric cars; NGK and Tokyo Electric refined it for the power grid.

Since the 1990s, Japanese businesses have installed enough NaS batteries to light the equivalent of about 155,000 homes, says Brad Roberts, head of the Electricity Storage Association. In the USA, AEP is using the 30-foot-wide by 15-foot-high battery to supply 10% of the electricity needs of 2,600 customers in north Charleston, says Ali Nourai, AEP manager of distributed energy. The battery, which cost about $2.5 million, is charged by generators from the grid at night, when demand and prices are low, and discharged during the day when power usage peaks.

That $2.5 million cost seems high for 10% of the power for 2,600 customers. That's about $960 per customer for something that lasts 15 years. Plus, there's the cost of the electricity lost when it is stored since no battery stores and retrieves electricity with 100% efficiency..

The biggest drawback is price. The battery costs about $2,500 per kilowatt, about 10% more than a new coal-fired plant. That discourages independent wind farm developers from embracing the battery on fears it will drive the wholesale electricity prices they charge utilities above competing rates, says Christine Real de Azua, spokeswoman for the American Wind Energy Association.

It is worth noting that sodium and sulfur are very cheap with sulfur in the tens of dollars per long ton (which is 1016 kg). Sulfur prices dropped a lot in the 1990s and most marketed sulfur come from removal of sulfur from oil and other fossil fuels in refineries. Due to US regulations which went into effect in 2006 to lower sulfur content of diesel fuel refinery sulfur production is up. In a nutshell, there's plenty of cheap sulfur available for making NaS batteries.

Can the prices for large industrial NaS batteries fall? Does anyone understand the processes involved in making NaS batteries and where the big costs come from?

If ways can be found to make NaS batteries cheaply then that would tend to help nuclear, solar, and wind power. Cheap ways to store nuclear electricity would allow nuclear power generated at night to supply peak power needs during the day. This could greatly reduce the demand for peak power generated from dwindling supplies of natural gas. Batteries would also enable solar and wind power to provide electricity when the sun does not shine and the wind does not blow.

Our main problem with limited remaining supplies of fossil fuels is not a simple energy shortage. Rather, our biggest problem is an energy storage shortage. That distinction is of enormous importance.

Oil produces gasoline and diesel fuels which store compactly in cars. Natural gas can get stored in tanks for use by electric utilities to generate electricity at times of peak electricity demand. Coal can get cheaply stored in big containers and carried around in train box cars for use when heat power is needed by steel mills, electric generator plants, and industrial heaters.

We have affordable alternative sources of energy but they do not store well. While photovoltaics are still too expensive nuclear and wind power are affordable without huge changes in lifestyles. Photovoltaics will eventually become affordable as well.

Solar photovoltaics, wind, and nuclear power all produce electricity that is not easily stored for use where and when it is needed. We can switch away from dwindling and increasingly expensive fossil fuels only when we can find ways to store nuclear, wind, and solar power. Therefore the development of better, cheaper, and longer lasting batteries is essential for the migration away from fossil fuels.

Share |      Randall Parker, 2007 July 08 11:39 AM  Energy Batteries

Fat Man said at July 8, 2007 2:20 PM:

If my calculations are correct, the battery alone would be about 5.9 cents per kilowatt hour.

Jerry Martinson said at July 8, 2007 10:42 PM:

Interesting article Randall. It's surprising but I had never even heard of these kinds of batteries before. They definitely seem rather expensive for bulk grid storage but they're a lot closer than what I thought the state of the art was for electrical storage. They seem like their just too complicated for small scale use. However, the can-do nerd in me suggests that it might be easier to develop this technology closer to the small scale use than wring the cost out of utility-scale storage. If it could be scaled down, there's a lot of places where this would make a lot of sense. There's an awful lot of people living in the boonies that don't want any grid tie and are living off of photovoltaics and diesel generators. They'd pay quite a bit for electrical storage.

Much of the diurnal "peak" of energy consumption can be handled much more cost-effectively on the consumption end rather than the generation and storage end. Most of the peak of consumption is from air conditioning. It is possible to not only shift the peak in time but it is also possible to drastically reduce the peak altogether in certain markets. Better insulation is a major component and a revamped performance-based energy building code would make a big start.

In most areas, building codes covering energy efficiency are essentially prescriptive, meaning that they say you should stuff the wall with whatever fiberglass fits in a 2x4 cavity, rather than performance based. Part of the problem is that few people understand thermodynamics well enough to understand that that the whole R-value concept for describing a wall is completely flawed. If we had true aggressive performance metrics, we'd all be using structural insulated panels.

While at first glance it seemed optimal, Title 24 in California certainly leaves a lot to be desired. I'm not aware of any think-tank studies on the practical effects of Title 24 but I think there's a lot that could be improved after seeing what behaviors it actually causes in people. Right now it does a lot more to reduce cooling load with respect to glazing but actually increases heating load. Conversely with respect to insulation it cuts heating load but ironically increases cooling load. Although there's a lot of stuff about passive solar design, there's essentially no credit for Victorian-era solar design like simple awnings and dutch gables. Before there was natural gas and air conditioning, people were living comfortably in coastal California with no HVAC at all. I know people living in 100+ year old houses that use a space heater one day a year and nothing else while Title24 houses next to them are having mammoth utility bills due to HVAC costs.

I can't believe how many building professionals just do not understand the thermodynamics and psychrometrics of a building envelope. There's quite a bit of existing housing stock that is improperly insulated that could be improved dramatically. However, there are bureaucratic hurdles to getting these upgraded. Most attic insulation can be improved easily. However, the floor of many houses has been neglected and it is the same amount of surface area. Most houses with crawlspaces either have no crawlspace insulation or are insulated and vented in a scheme that made only partial sense before the invention of the 6 mill plastic vapor barrier in the 1950's and no sense at all since. Yet insulating a crawlspace the right way (insulating the perimeter walls and not the floors), is still essentially illegal because the building codes (or at least the building inspectors) are trapped in the 1930's. I insulated my crawlspace the right way and it only cost about $200. I've done some detailed analysis and believe that it'll pay for itself within 2 years. There are millions of houses like mine, many with air conditioning - 20% of it leaking out uncontrolled through the floor due to someone's well-intentioned but misguided effort to protect the floor joists from rotting.

In arid areas, most houses can be adequately conditioned by using a whole-house fan during much of the cooling load days. In many cases, the attic is a scorching hot 150F because the attic ventilation is very poor during the afternoon and that supposedly optimal R38 attic insulation is actually a giant superheated thermal mass for the evening when people come home that the air conditioner has to fight for the next 8 hours. There are houses all around me that have AC running through much of the night in the summer, even when it is lower than 65F out.

Some of the neatest ideas seem like they just suffer from some economies of scale problems where they exceed what makes sense for a single household but would make sense for about 20 or 30 homes. You could then have evacuated tube solar collectors drive large NH3 or LiBr absorption chillers for district heating and cooling with a neighborhood pool becoming the "cooling pond". Combine it with photovoltaic roofs and these batteries you have and a neighborhood could go "off grid" at a reasonable price.

Bruce said at July 9, 2007 7:07 AM:

The cost of sodium and sulfur is low, but the battery requires specially careful handling of these materials. The NaS cell requires molten components at 300-350 C, which mandates the complete exclusion of water and oxygen from the cell. The sodium and sulfur are separated by a ceramic container, which might be difficult to cast in large industrial sized cells. Some means must be employed to initially melt the stuff, but the 10% electrical losses incurred in charge/discharge cycles generate enough waste heat to maintain temperature. Also, there's the AC-DC-AC conversion problem, which isn't cheap at megawatt power levels.

Randall Parker said at July 9, 2007 6:15 PM:


Thanks a lot for the information.

Do you have any idea of the prospects for lowering the cost of manufacture?

The need to maintain a high temperature suggests it would only work in scenarios where you expect regular charge/discharge cycles. Charging up a cell on a sunny or windy day to use over a period of days might not work because the thing would cool down too much over a period of days?

Randall Parker said at July 9, 2007 7:00 PM:

Jerry Martinson,

Great response. Thanks.

Peak shifting: Dynamic pricing would help. We need higher prices during hot afternoons and lower prices at night. Make the prices follow demand and people will find ways to shift their demand.

Housing design deficiencies: I only wish this topic got half the attention that car fuel efficiency gets. Per dollar spent we could get far greater reductions in energy demand by better housing design and appliance choices than by buying hybrid or diesel versions of cars.

As for low tech old ways to keep houses hot or cold: Yes, whatever happened to awnings?

Also, should house roofs be black as so many are? It doesn't make sense. In the winter there's not enough sun to heat a house's roof much. In the summer we want the sun to bounce off roofs. The ability to easily change roof and siding colors could save a lot of energy, no?

How to cost effectively update tens of millions of existing houses and apartment buildings for higher energy efficiency? Could software do the job of analyzing existing structures to identify some cost effective ways to improve them?

Fat Man said at July 9, 2007 10:20 PM:

Diurnal fluctuation due to air conditioning could also be cushioned by using tubs of water that the refrigeration unit could chill off-peak. One issue is just economic. Who will make the capital investment? homeowners or power companies?

Randall Parker said at July 10, 2007 6:31 PM:

Fat Man,

I'd like to see cost estimates for the equipment needed to shift heat and cool between day and night. How much of differences in pennies per kwh between peak and late night are needed to justify the shifting of heat or cold between day and night?

TheBaldGuy said at July 11, 2007 2:53 PM:

Regarding the comment "Peak shifting: Dynamic pricing would help. We need higher prices during hot afternoons and lower prices at night. Make the prices follow demand and people will find ways to shift their demand." Absolutely, but there is a problem. Perception of bias in the pricing. As Randall comments, you get into "justification". I've got a possible explanation but figure it's so obvious there has to be a good reason(s) why nobody seems to be doing it.

As the article points out, solar and wind (for example) provide good power during the day, and less to none at night. Sure, in terms of a per-building context this isn't a big help since you still need night power. But what about at the plant level. What is wrong with using the more expensive solar as a peak demand (daytime such as cooling) source. Yes it is more expensive, but that's the obvious justification to the customers.You've got a simple and understandable "justification" for making peak demand rates more expensive without making people think too much about it. If the solar generation does not need to store the energy for later use, but is used to supply extra demand, it seems to me it should be cheaper.

Sure the average Joe *should* be thinking about his or her energy usage patterns, but the world has shown us that they would rather not. Higher prices "to make you use it better" doesn't seem to fly well. But higher prices because you have to meet the demand, that people get and don't feel insulted over.

As for awnings, absolutely! Lately I've been thinking about a small "awning" or roof for my roof on one side of the house. Something along the lines of a sail or tar strung up over that side of the roof to lower the direct effect. Shading your roof. Hey if trees can do it, why not a small barrier over it? This weekend I'll be "shading" about 1/3 my back yard - and part of that will shade the house on the sunny side.

Oh, and Jerry, can you share what you did to your crawl space, and how a radiant flooring install would affect how or what you did? I'm quite interested in insulating my crawlspace.

Fat Man said at July 16, 2007 11:32 AM:


Your wish is my command.

"System Relies on Ice to Chill Buildings" By Colleen Long of the Associated Press on July 16, 2007:

... some office towers and buildings have found a way to stay cool while keeping the AC to a minimum - by using an energy-saving system that relies on blocks of ice to pump chilly air through buildings.


... State officials say there are at least 3,000 ice-cooling systems worldwide.

... Because electricity is needed to make the ice, water is frozen in large silver tanks at night when power demands are low. The cool air emanating from the ice blocks is then piped throughout the building more or less like traditional air conditioning. At night the water is frozen again and the cycle repeats.


In the basement, three main cooling rooms house chilling machines and 64 tanks that hold 800 gallons of water each. Credit Suisse has a traditional air conditioning system, but engineers use the more efficient system first.


Ice storage at Credit Suisse lowers the facility's peak energy use by 900 kilowatts, and reduces overall electric usage by 2.15 million kilowatt-hours annually ...

... costs are considerable: Credit Suisse spent more than $3 million to renovate its cooling system; ...

Randall Parker said at July 16, 2007 10:05 PM:

Fat Man,

Great story. But it is not clear how much money they save.

First off, do big commercial customers pay different day and night time rates even though home owners do not?

Second, how do they save 2.15 million kwh annually? And what is their rate on that? If it is 15 cents/kwh (which is ballpark - perhaps low - for New York) then that's $322,500 per year saved right there. Great return on $3 million invested - over 10%. But are they saving even more with lower priced electricity at night?

Another thought: It is easier to get the big returns on investment from ice cooling in large buildings. The smaller buildings all the way down to houses lose economies of scale.

We need pricing which varies with time of day, day of week, and time of year. We really need dynamic pricing that varies with demand and supply.

odhran said at July 28, 2007 2:20 PM:

does anyone know is this available on a domestic scale?

Post a comment
Name (not anon or anonymous):
Email Address:
Remember info?

Go Read More Posts On FuturePundit
Site Traffic Info
The contents of this site are copyright