Forgot your password?
typodupeerror
IBM Power

Lithium Air Batteries Get Boost From IBM and DOE 240

Posted by samzenpus
from the power-up dept.
coondoggie writes "The Department of Energy and IBM are serious about developing controversial lithium air batteries capable of powering a car for 500 miles on a single charge – a huge increase over current plug-in batteries that have a range of about 40 to 100 miles, the DOE said. The agency said 24 million hours of supercomputing time out of a total of 1.6 billion available hours at Argonne and Oak Ridge National Laboratories will be used by IBM and a team of researchers from those labs and Vanderbilt University to design new materials required for a lithium air battery."
This discussion has been archived. No new comments can be posted.

Lithium Air Batteries Get Boost From IBM and DOE

Comments Filter:
  • by Anonymous Coward on Wednesday January 27, 2010 @08:07PM (#30927682)

    and being metals they will require some thought into the use of water to put the flames out at accidents.

    Why are you concerned, because of the electrical shock risk?

    No, Lithium reacts violently with water, as do all the rare-earth metals in the left column of the periodic table (Sodium, Potassium, Cesium...)

  • Re:Patents? (Score:1, Informative)

    by Anonymous Coward on Wednesday January 27, 2010 @08:08PM (#30927686)
    Wow... I actually R'ed that FA, and there's a prime candidate for tl;dr if I ever saw one.
  • by Rei (128717) on Wednesday January 27, 2010 @08:24PM (#30927844) Homepage

    Well, since the average driver drives about 12,000 miles a year, and the average car is on the road for nearly two decades....

    Sure, an individual owner doesn't keep it that long, but what that means is that your depreciation will be lower, since the vehicle remains cheap to operate. Once the luxury of a luxury vehicle wears off, or the style of a stylish vehicle becomes dated, you don't have much left. But efficiency is always a seller. A Hummer doesn't cost that much more than a Prius, but it depreciates three times as fast.

  • by Anonymous Coward on Wednesday January 27, 2010 @08:25PM (#30927854)

    Rare earth elements consist of the f-block metals. The first column s-block metals Li, Na, K, Rb and Cs are all alkali metals. Lithium is actually the least reactive metal of the column. Potassium catches fire on exposure to water and Caesium essentially explodes on contact.

  • Re:Overstated (Score:5, Informative)

    by dragonsomnolent (978815) on Wednesday January 27, 2010 @08:39PM (#30927984) Homepage
    I had read once that they were using the same technology as laptops so that they could let laptop battery manufacturers do the heavy lifting on battery R&D (a sensible approach I suppose) and after reading thier pdf about the batteries it seems to hint that they use them because they are cheap, standard (same ones used in laptop batteries), and should one fail, it doesn't affect the entire system as much overall (there is no mention of fire damage however). I'm sorry that I can't answer your question regarding increase or decrease of performance as size increases. But it doesn't seem Tesla is using small cells to avoid patent licensing issues (after all, Wikipedia indicates that they license their AC Motors in the Roadster)
  • by Rei (128717) on Wednesday January 27, 2010 @08:52PM (#30928128) Homepage

    Accord [automotive.com]. Prius [automotive.com].

    The Prius depreciates a lot as soon as you drive it off the lot, but less than half as much each year after that -- despite being a more expensive vehicle.

    Efficiency = low depreciation for the long run.

  • Re:Well (Score:3, Informative)

    by John Hasler (414242) on Wednesday January 27, 2010 @08:56PM (#30928188) Homepage

    > Can the US Gov hold patents?

    It can and does.

  • by Rei (128717) on Wednesday January 27, 2010 @08:57PM (#30928198) Homepage

    Well, for the record, GM says the Volt's pack costs under $10k. And that's first-generation. The raw materials in these types of cells are dirt cheap, so there's major potential for prices to drop in volume production.

  • by Rei (128717) on Wednesday January 27, 2010 @09:05PM (#30928266) Homepage

    Oh, and also, to help you "extrapolate" properly in the future:

      * EV drivetrains are currently handmade in small volumes, so they're very expensive. Even a low-end AC drivetrain will cost you about $10k (say, a DMOC445, AC24LS, and a Manzanita Micro PFC charger). A good one like the AC-150 that the Roadster's drivetrain was originally based on will run you more like $25k.
      * The Tesla Roadster's pack is very, very different from the Volt's, so it's not a good idea to compare the two. The Roadster's is a high capacity based on cobalt cells with a massive cooling system and a high DoD. The Volt's is low capacity based on manganese cells with a smaller cooling system and a low DoD.
      * The Tesla Roadster is a luxury carbon fiber sports car that does 0-60 in under 4 seconds. You get what you pay for.

  • Re:Recharge time? (Score:5, Informative)

    by Rei (128717) on Wednesday January 27, 2010 @09:22PM (#30928404) Homepage

    8 hour charge for how many miles? I don't know about you, but my daily commute isn't 600 miles.

    It's level 1 or level 2 charging at home, and level 3 or higher for long trips. And that's what it's going to be for probably the next century. It doesn't make sense to do it any other way. You only need fast charges when you're taking long trips, so you need fast charging stations available on the road. Around home, you want slow charging, which is gentler on the batteries (and, not to mention, the grid), as well as being more efficient.

    By the way, for those who are curious:

    Level 1: ~110V, 20A or less. US standard: SAE J1772 or the ever-common NEMA 5-15 plug.
    Level 2: ~220V, 80A or less. US standard: SAE J1772. European standard: Mennekes, based on IEC 60309.
    Level 3: ~440V, up to "hundreds" of amps. No official standard, but the TESCO connector seems to be becoming dominant.

    The most powerful EV charger I'm aware of is an 800kW charger created by Aerovironment for TARDEC. That's ~800V and ~1000A, if I recall correctly. It's about the size of four vending machines pushed together.

  • Re:Overstated (Score:5, Informative)

    by Rei (128717) on Wednesday January 27, 2010 @09:30PM (#30928448) Homepage

    The person who responded to you first is indeed correct. It's not about patents; you're mixing up this with the old EV1 debacle. The Roadster uses 18650-format cobalt/graphite li-ion cells, which are already in mass production. They did this for obvious reasons; when they started out, the phosphates and spinels that everyone else is now using weren't really available.

    As for fire, which the previous person commented on, each cell is contained within its own can that's designed to isolate failures to just that cell. It's a pretty complex pack indeed. Future EVs won't have such a complex pack. It's doubtful that even the Model S will, even though it's still going to be based on cobalt tech (that's what Tesla has experience with, after all -- and despite all its downsides, it is quite energy dense)

    If you're curious as to how the pack is structured, there are 11 "sheets", each one made of 9 "bricks", and each of those made of 69 cells. Each of the cells in a brick are wired in parallel. The failure of one, therefore, has relatively little impact on the performance of the brick. The bricks and sheets are wired in series. Each sheet monitors the performance of all of its bricks and does load balancing on them, as well as logging failures. It's a pretty impressive piece of engineering.

  • by iamhassi (659463) on Wednesday January 27, 2010 @09:31PM (#30928462) Journal
    " And after the 1.6 billion hours, does the computer self destruct? Just curious."

    Same. 24 million hours? There's only 8,765 hours in a year [google.com], so what is that, about 2,500 years?

    So I googled it. Apparently supercomputer hours aren't people hours, they're processor-hours, so 1 processor working for 1 hour is 1 processor hour. [rochester.edu] 24 million hours means (# of processors) * (# of hours) = 24 million. For example, (24,000 processors) * (1,000 hours) = 24 million. So it could be done in 41 days, not 2,500 years, if they have 24,000 processors working on it.

    Not sure if I like this method of measuring processor usage since a project that took a million hours in 2001 wouldn't take a million hours in 2010 but that's what's in the article.

    Oh and to answer your question: no, it probably doesn't self-destruct but it'd probably be replaced since I'd imagine if 1.5% is anywhere near my hypothetical 41 days then that'd put 1.6 billion at about 7.4 yrs.
  • Re:DOE is serious? (Score:5, Informative)

    by iamhassi (659463) on Wednesday January 27, 2010 @09:40PM (#30928522) Journal
    "And after the 1.6 billion hours, does the computer self destruct? Just curious"

    Sorry I'm back and I have answers.

    The Oak Ridge "Jaguar" Supercomputer is the World's Fastest, with 37,376 six-core AMD processors [sciencedaily.com]. That puts it at 224,256 processors, so those 24 million hours should be done in 107 hours, or a little more than 4 days.

    The 1.6 billion hours comes from the here: [sciencedaily.com] "....computing facilities at Oak Ridge and Argonne national laboratories will employ a competitive peer review process to allocate researchers 1.6 billion processor hours in 2010." That works out to be about 297 days.
  • by Anonymous Coward on Wednesday January 27, 2010 @09:40PM (#30928526)

    Ummm... No it's not. It may be one of the best energy densities that we've found yet for relatively-safe, convenient, and stable common compounds, but it's nowhere near "the most dense". Most rocket fuels, explosives, and the class of more-efficient combustion products (such as hydrogen) are all better. They just have this annoying tendency to release their energy unexpectedly, all at once, or both.

  • by Cyberax (705495) on Wednesday January 27, 2010 @09:41PM (#30928532)

    Nope.

    Lithium is plentiful, you can mine it from seawater indefinitely for about $60 per kg. It's just that some countries can supply lithium at smaller prices.

  • by Rei (128717) on Wednesday January 27, 2010 @09:41PM (#30928534) Homepage

    The annual energy usage of automobiles is more than the current electricity usage in the US.

    True but grossly misleading. :) The average car has a tank-to-wheel average efficiency in normal combined city/highway driving of about 20%. Your average li-ion electric vehicle has a plug-to-wheel average efficiency under the same conditions of about 85%.

    The reality is that almost no new generating capacity is needed [pnl.gov].

  • by Rei (128717) on Wednesday January 27, 2010 @09:48PM (#30928574) Homepage

    Gasoline at 50MJ/kg is pretty much the most dense energy storage possible in this universe excluding nuclear energy.

    Not even close. For example, beryllium blows it away in both volumetric and gravimetric energy density (and hydrogen blows beryllium out of the water in gravimetric comparisons, but sucks at volumetric). And comparing any of them to nuclear energy is laughable.

    This is kind of a fundamental limit as to how much energy can be stored in *any* system using potential energy of the electric field of matter.

    No, it isn't. Nor is beryllium. Energy doesn't even have to be stored in chemical bonds (see, for example, digital quantum batteries).

    You may get 2x better efficiency in an electric motor,

    Try 4x in typical driving conditions.

    but I can not see how a battery can approach this value.

    It doesn't need to. A motor the size of a watermelon propels the Tesla Roadster from 0-60 in under 4 seconds. In gasoline cars, the fuel is light and the engine is heavy. In EVs, the motor is light and the "fuel" (the battery pack) is heavy. It's a reversed paradigm. You have to compare the mass and volume of the engine + fuel to the mass and volume of motor + fuel. And with current battery tech, you'll find that EVs are about 1/4 to 1/3 of the way to matching gasoline cars. But batteries have increased nearly 5-fold in energy density the past 21 years, and show no signs of stopping.

  • by John Hasler (414242) on Wednesday January 27, 2010 @09:52PM (#30928606) Homepage

    > Last I heard was lithium was a precious metal...

    You last heard wrong. It goes for around $100/kg, less than 1/4 the price of silver.

    > ...50% of the world's sources were in one country (So Am).

    Chile seems to currently have the largest proven reserves, but lithium is not very rare (similar in concentration in the Earth's crust to nickel and lead) and is widely distributed.

  • by shermo (1284310) on Wednesday January 27, 2010 @10:05PM (#30928712)

    Coal plants can and do back off their generation at times of low demand. Typically they can go down to about 50-60% capacity without problems. You're correct in the general concept that they don't switch off and on quickly, but they certainly don't generate at max capacity 24/7.

  • Re:Recharge time? (Score:3, Informative)

    by iamhassi (659463) on Wednesday January 27, 2010 @10:12PM (#30928756) Journal
    "Many residential homes in the US have 100 amp service.

    Most have 200. 400 is usually available at extra cost.*"


    *citation need

    Got it right here, says [answers.com] you're [yahoo.com] wrong [answerbag.com]

    New construction homes get 200 amp, but even as recent as 2006 builders were providing 100 amp and 200 amp as an upgrade. [allexperts.com] This electrician in Wisconsin recommends 100 amps for house under 2,000 sq/ft [romanelectric.com]. I don't exactly know date when 200 amp became the standard for new construction but it's clear 100 amp is the norm for your average pre-owned home. 400 amp service for a residence basically doesn't exist [jlconline.com] unless you have extreme circumstances, like you were dumb enough to buy a 15kW tankless electric water heater [globalindustrial.com] (idiot should have bought gas) that's sucking down 130 amps when in use.
  • by scdeimos (632778) on Wednesday January 27, 2010 @10:15PM (#30928792)

    More stuff on Prius battery ranges here. [caradvice.com.au]

    The only two recorded Prius battery changes in Australia (at the time of the article) where at 350,000km (220,000mi) and 500,000km (310,000mi). That's pretty good mileage and they're thrashing these things about in Taxis clocking up around 200,000km (125,000mi) per annum.

  • Re:Recharge time? (Score:5, Informative)

    by scdeimos (632778) on Wednesday January 27, 2010 @10:30PM (#30928916)
    Didn't anyone ever tell you that the fuse is *meant* to be the weakest link? Now, with your 200 amp fuse/breaker in place you'll burn out the house wiring instead.
  • by fractoid (1076465) on Wednesday January 27, 2010 @10:36PM (#30928964) Homepage
    NiMH batteries are terrible compared to LiPoly or nanophosphate lithium batteries. The only reason we're still stuck with them is manufacturers trying to gouge back the R&D costs that they sunk before lithium batteries appeared and sunk their NiMH market.

    Metal-air battery chemistries have been used before in EVs - specifically zinc-air batteries [wikipedia.org] - but they are generally primary cells and need to be mechanically recharged. TFA mentions charging so possibly the lithium-air cells are proper secondary cells. Also, the specific power of air-based batteries is historically very low, and I note that the only mention of power in TFA is where they say:

    The most important [scientific challenges] are to realize a high percentage of the theoretical energy density, to improve electrical efficiency of recharging, to increase the number of times the battery can be cycled, to limit the negative effects of moisture in the air, and to improve the power density.

    Of course you could always do a hybrid battery pack using Li-Air for bulk storage and nanophosphate lithium or even ultracaps for load levelling.

  • Re:Recharge time? (Score:4, Informative)

    by scdeimos (632778) on Wednesday January 27, 2010 @10:38PM (#30928982)

    It all depends on the discharge/charge ratings for the cells. We regularly punish Li cells in hotliner electric gliders.

    For example, a 1,000mAH Li-Ion cell with a 5C charge rating can be safely charged at 5,000mA from near flat in 10 to 12 minutes. The charge ratings tend to go down as cell sizes increase, though, due to ventilation issues - you just can't dissipate the heat from the battery packs quickly enough unless you involve forced-flow systems, and if it gets too hot you'll get a runaway situation and BOOM.

  • by RobVB (1566105) on Thursday January 28, 2010 @12:09AM (#30929458)

    Electric energy can propel your car for $0.03 per mile. If gas taxes were taken out (I used my states gas tax, yours could be several cents different either direction), you are paying roughly $2.30 per gallon and if you car gets 35mph per gallon you are paying $0.06 cents per mile, that's HALF the cost.

    $2.30 per gallon is dirt cheap, compared to prices here (Belgium). Gasoline here is about €1.40 per liter, that's almost $2 per liter or roughly $7.50 per gallon. The difference in electricity prices is much smaller: with a separate installation that works only during the night you can charge your car at €0.09 per kWh. Regular daytime prices are around €0.18. Judging from this list [doe.gov], that's not too different from prices in the U.S. ($0.0764 in North Dakota, $0.2028 in Connecticut, $0.2379 in Hawaii)

  • by Rei (128717) on Thursday January 28, 2010 @12:23AM (#30929546) Homepage

    I did respond to it. What he wrote was complete pseudoscientific nonsense.

  • by Rei (128717) on Thursday January 28, 2010 @12:29AM (#30929572) Homepage

    The *average* age of a car on the road today is 9.4 years and rising [consumerreports.org].

  • by Rei (128717) on Thursday January 28, 2010 @01:29AM (#30929930) Homepage

    Not even close. For example, beryllium blows it away in both volumetric and gravimetric energy density (and hydrogen blows beryllium out of the water in gravimetric comparisons, but sucks at volumetric).

    Hydrogen was included in TFA comparison.

    Nice try at changing the subject away from the fact that you're quite simply wrong about gasoline being the most energy-dense or nearly most energy dense substance in the universe. It's not even close. If you really want to find the most energy dense chemicals, you need to look at metastable solids. Cubane and nitrogen rings, for example. And there are some theoretical ones that may be even higher, such as triplet helium. These things way, way outclass gasoline in terms of energy density.

    Energy is still stored in the electrical field in matter. A quantum battery needs a lot of infrastructure to handle the forces, so at least 50% of the weight will be wasted. (compare to the weight of a clamp holding a spring.)

    1) "Still"? Chemical batteries don't store energies in electrical fields.
    2) You're trying to bond energy released in a chemical reaction with tensile strength. Tensile strength != energy. And no, they're not related. A beryllium cord has a *lot* less tensile strength than a carbon nanotube cord (orders of magnitude), but releases significantly more energy when it burns.

    No, A small VW diesel has up to 40% efficiency.

    "Up to" != "Average usage". Duh. Diesel cars average about 25% efficiency in typical mixed usage. Engines only get their peak efficiency within a narrow power band.

    An elelctric car may have 90%, but you can only use 60% of the battery without damaging it in a few cycles, so overall, 2x is conservative.

    Wrong on so many different levels.

    1) Efficiency has nothing to do with pack capacity. You're equating the two. 90% *efficiency*. Versus 20% *efficiency*.
    2) The Tesla Roadster uses over 90% of its pack's capacity. Most li-ion BEVs are in the 75-90% DoD range. Not 60%. The Volt uses 50%, but only because A) they're taking an extremely conservative approach, and B) it's a small-pack PHEV.

    You are partially correct. A brushless electric motor can have very high intermittent power density. maybe 10x of a gas engine. It is only limited by cooling. For continous power its power density is the same as a gas engine.

    First off, you're confusing DC and AC motors. All AC motors are brushless. Brushless is a category of DC motors. Secondly, no. The Tesla Roadster can do anything but track duty without a liquid cooling system. With a liquid system it could easily due track duty. And even with just air cooling, it beats the hell out of non-sports cars in sustained power output, despite having an engine much smaller than even non-sports-cars that run on gasoline. And furthermore, how important is track duty to the average person?

    It is actually quite complicated to cool an electric motor.

    No. You can buy motors with the cooling already in place.

    Think 100kW power, and 10kW heat.

    First off, 100kW power is something you'll only ever get during very high acceleration or extremely high speeds. Cruising power is more like 10kW, meaning 1kW heat. Secondly, since gasoline cars average about 20% net efficiency, 100kW of gasoline power output equals *80* kW of heat that you need to get rid of. It's much, much easier for the EV.

    Find an electric motor that had higher energy density than a gas engine for continous output, and I will stand corrected, and learn something new.

    The very one we're talking about. The Roadster's motor can do 2/3rds of its peak output as sustained. And peak output does 0-60 in under 4 seconds.

    Note that the Roadster's motor is hardly the most power dense electric motor out there. Look at the PML Flightlink in-wheel motors used in the Lightning GT, for example. Each in-wheel motor is rated for 120kW peak and are.. well, the size of a wheel.

  • by asc99c (938635) on Thursday January 28, 2010 @09:17AM (#30932198) Homepage

    LiPoly doesn't currently have the power-to-weight ratio of some battery technologies, which is a big factor for car batteries.

    I was looking for a definite reference for this, but I can't see many LiPoly specific references. The Wikipedia page says 7.1kW/kg, which seems stupendously high and, I suspect, completely wrong. The Wikipedia entry for Li-Ion says 250-340 W/kg, which is more reasonable, while NiMH shows as 25-1000 W/kg. Both of these ranges are easily found elsewhere. LiPoly runs similar chemistry so the results should be similar.

    Taking middle of the range figures for each battery type - say 300 W/kg and 600 W/kg - your NiMh car will have twice the bhp of the LiPoly car.

I bet the human brain is a kludge. -- Marvin Minsky

Working...