98 comments

[ 0.16 ms ] story [ 194 ms ] thread
Hmm, what about lithium air batteries, and fuel cells, and all the other things that we hear about having 10x the energy density?
Haven't heard about "lithium air batteries" but I'd say fuel cells aren't batteries since they basically just oxidize methanol (which is what an engine does with gasoline).

But the lines might be more blurred than my thinking indicates.

> I'd say fuel cells aren't batteries since they basically just oxidize

Many batteries also use oxidation reactions. That's not how you define a battery. I think the main difference is batteries release all the energy as electricity without associated heat, but fuel cells are partially electricity, partially heat.

Fuel cells are just "open" batteries, where the reactants are continuously fed into the battery and the products are removed (whereas in a battery the reactants and products occupy the same physical volume).
I'm pretty sure you're mistaken. The battery you describe violates the Second Law of Thermodynamics (no system can ever be 100% efficient). Both batteries and fuel cells (and any other open physical system) will lose some energy as waste heat in doing their work (producing electricity).
By the way, are super capacitors still a promising technology? What's the hold up on getting those commercialized?
(1) Supercapacitors are commercialized, see: http://www.mouser.com/Power/Supercapacitors/_/N-6uivw

(2) Today's supercapacitors have very poor energy density even compared to batteries. What they do have is good power density, meaning they can deliver their limited energy very fast.

(3) Boosting the energy density of supercapacitors will happen but requires many years of great strides in nanotech. Research like this, involving 3D nanostructures with enormous surface area: http://phys.org/news/2012-10-sponge-like-graphene-supercapac...

Hmm, I'm trying to compare these to batteries. Does anyone know how to convert farads to watt hours?
Energy stored in a capacitor is equal to:

1/2 * c * v^2

Where c is the capacitance in farads, and v is the voltage. That will produce a value of joules. Note that the capacitor is rated for a given voltage, so that goes in as v.

Converting watt hours to joules goes like this: A watt is one joule per second, so a watt hour is 3600 joules.

By that formula, I only get ~12 joules for a 1 farad capacitor at 5 volts. That seems low. Are you sure it's correct?
that's correct. welcome to capacitors.
Here is my favorite movie about ultracapacitors: http://www.youtube.com/watch?v=EoWMF3VkI6U

Very good explanation on the general math and physics of capacitors, current, heat and energy storage, plus awesome demonstrations at the same time :)

Considering that this post was written in 2008... What changed in these 5 years in the "battery" industry?
What's preventing us from putting fuel cells in laptops and tablets soon? Is it just that today's fuel cell designs are physically bulky? Are we near theoretical limits on fuel cell miniaturization or is there still lots of headroom? What's the safest way to store the fuel? Are there too many downsides to making the fuel cell air-breathing (as opposed to the danger of storing oxygen)?

http://en.wikipedia.org/wiki/Fuel_cell

Wikipedia says that fuel cells are typically "stacked" (placed in series) since each one only makes 0.7v, but if a new tablet is designed to save space by using only a single 0.7v cell and upregulating the voltage with a boost regulator, won't the total efficiency even with regulator losses still be far above li-ion?

There are a lot of fuel cells for sale in Japan right now for recharging a laptop, ipad or cell phone, but they charge lots for refills, so it's not an economical option.
Interesting in the 5 years since this was written, what the friend claimed would happen has come to pass. Car2go runs several all electric car sharing fleets in various cities. I've used the one in San Diego a number of times.
Reminds me of the time when it was declared that there was a physical barrier to sending more than 44kbps through a phone line.
Wasn't there? How much faster did modems over POTS really get? IIRC, even 56kbps only worked downstream, and only if you had a modern digital connection back to your telephone system.
Yes, they were completely right, and improvements only happened by changing the game that was being played.
ADSL connections over the same physical wire as a phone line can often get several Mbps.
But ADSL only goes as far as the exchange before being routed onto an IP backbone network. With modem signals the same signal goes over a voice circuit (albeit a virtual one) all the way to the receiving modem. They're not at all the same thing.
That's the point. People make sweeping assumptions about the limitations of something, and then someone else changes the ground underneath those assumptions and makes the cconclusions invalid. DSL, for example, is not much more difficult in terms of capital equipment cost compared to dialup and it doesn't require running a new line to the home, but the bandwidth available is a thousand times greater than the previous "hard" limit of phone land line based networking.
I'm not a technician, but I know that I now get 12MBps (via DSL) and they haven't exchanged the phone lines. So they must have come up with something.
The physical copper wire is still the same, but you're no longer talking to the phone system on the other end of it, as I understand it. DSL is no longer using the "phone line". It's just using the phone cable.

The battery analogy would be creating a fuel cell in a AA (or whatever) form factor. It circumvents the limits on battery technology by not being a battery, in spite of plugging into the same physical slot.

(NB: I am not a DSL technician, I could be wrong).

Maybe, but the point is the constraint was never "use the phone system", only to use the cable.
That's not the takeaway I got. What I got is that the (communication endpoint,fundamental energy storage) technology has to change before there can be large improvements in (network speed, energy storage capacity).
Sure, but that article seems to make some assumptions about batteries that might be irrelevant. The requirement is not "a battery has to power the car", it is just "a clean energy source has to power an electric motor".
Yes, 56k is the limit to "phone lines" meaning the system between you and another phone subscriber, given the limitations of how voice is transmitted.

ADSL works using the wire from your house to the phone company equipment. A much shorter path and with less interference (still, the wiring may be bad, which limits transmission rates)

There is a big difference in moving some bits of data and moving a few thousand pounds, trying to draw a correlation as such is disingenuous.
I am not saying the same thing will happen with batteries, just that I wouldn't give up just yet.
That's a very pessimistic view. There is plenty of on-going research in improving graphene batteries, molecular energy storage, and technology that is not that far from coming to market like Li-Si, Air-Lithium or solid hydrogen storage.

Tesla now makes cars with a very useful range using standard Li-ion batteries and good thermal management, I wonder the author thinks of the Model S.

http://gigaom.com/2013/01/14/13-battery-startups-to-watch-in...

At this point, the Tesla's problem isn't range, it's charging convenience. 1/2 hour's manageable, but still a bit long.
It's more than enough for urban use where the car will sit still for most of the night and have plenty of breaks during the day.
That's the happy path. What if things are hectic and you forgot to plug in last night, then you get an urgent page the next morning. That 1/2 hour won't seem so casual right then. However, I suspect you are taking my comment opposite to its intent. The only thing standing between Tesla and a 5 minute charge is enough current and economies of scale to allow cars with massively parallel charging hardware. Basically, if they can survive, they've won!
Still faster than walking to the gas station and back :) but yes, you're right. I hope they are not put down by a media scandal or legislation before that happens.
> Still faster than walking to the gas station and back

Dude, the amount of current you're going to need for a 5 minute charge is going to put such capabilities only in the hands of businesses for decades at least.

> Still faster than walking to the gas station and back

Dude, the amount of current you're going to need for a 5 minute charge is going to put such capabilities only in the hands of businesses for decades at least.

Basically, it's like running a fleet of 30 cars all at once for five minutes.

(comment deleted)
But "molecular energy storage" and hydrogen (solid or otherwise) are not battery technologies. They are better classified as fuels which store (chemical) energy in chemical bonds, which is distinct from batteries which store (electrical) energy by charge separation.

I agree with the article author. I was convinced after reading chemistry professor Daniel Nocera a few years ago, after he became famous for discovering a cobalt-phosphate water-splitting catalyst, for the first time enabling hydrogen generation under normal conditions (neutral pH). He had been looking for it for some time (his "holy grail" of renewable energy).

He's a proponent of solar and hydrogen (collecting energy from solar and storing it in hydrogen) and a detractor of batteries. He (convincingly) argues that battery tech is fundamentally limited by the physics of charge separation. The only way to beat that limit is to store the energy in chemical bonds - as fuel - like nature does.

His paper "Chemistry of Personalized Solar Energy" is a good place to start.[1] Just the introduction and conclusion - you can skip the middle (its dense orbital theory).

1. http://www.tarleton.edu/Faculty/alow/solar2.pdf

Battery technology might never hit the density of fuels, but that doesn't mean it won't be cost-effective. The 10x increase in density that a slew of new technologies promise is more than enough for making electric cars mainstream. seeing we already have viable ones with today's battery tech.
No. Just no. I don't know how freely I can talk about my research, but there's plenty going on with batteries.
Just a note that this is from 2008. At least in the short term, he was right. Not much has happened in battery technology available to the public in the last 5 years.
I saw that it's from 2008. I'm not denying it's a hard problem, but to claim that improvement has stalled is short-sighted.
Without any evidence -- even theoretical -- you are just posturing. The fundamentals of the post are not "short sighted", they describe a deep fundamental flaw that is (has been) historically perisistent. There is a difference between being "short sighted" and a lack of transparency. Until the "magic" ships.
> Until the "magic" ships.

So, there is a better battery. And it's inevitable.

I'm not sure this is posturing. Or at least, it is certainly not "just" posturing.

I'm not in research like this, so I just hear things fed off of Y(HN). I've heard of some interesting devices like the carbon nanotube based supercapacitors (was that what it was?). Who knows?

In research, maybe.

But what we care about is when that research will result in better batteries _that_ we can buy.

The premise of the piece was that there's no new battery technology in the pipeline. Not on a particular timescale (he talks about lifetimes). The author believes batteries as a technology have reached peak energy density. If there's a research that suggests otherwise, that invalidates his whole point.
I think the author is explicitly NOT stating that battery technology has reached it's peak, but rather the safe bet isn't on huge changes in a decade time frame.

He's not commenting on the tech so much as the wisdom of ASSUMING it will be better in a certain time frame.

He's saying, it's not wise to build a business on top of the idea that battery tech will progress predictably or at all.

I disagree, because any business built on that will be the first to exploit it properly when it comes out, as a competitive advantage. But if they're wrong, they'll deal with the downside risk of, you know, being wrong.

Do you have any links to related papers that were published?
[disclaimer, I work at one or more of the places commercializing these systems]

http://en.wikipedia.org/wiki/Flow_battery http://www.ted.com/talks/donald_sadoway_the_missing_link_to_... http://en.wikipedia.org/wiki/Sodium-ion_battery http://en.wikipedia.org/wiki/Sodium-sulfur_battery

Seeing how much progress is being made extremely rapidly on getting these types of technologies into the marketplace, the idea that the last five years have been stagnant is difficult to believe.

Unless, of course, all battery research was stagnant under Bush? I suppose that's possible, and what we've seen is the standard 5-10 year research backlog starting to catch back up to a lack of funding for oil alternatives.

> None of this has done much to improve the fundamental energy storage densities of the underlying chemistries.

The LENR (Low Energy Nuclear Reactions) field is really hot right now.

COPs of 6-20x (and sometimes more) are being reported.

Several labs have been burned to the ground who experimented with it and had a run-away reaction that was not expected.

Whether it's Fusion or Transmutation or something else, the results are real and we are seeing heat anomalies that can't be explained, in table-top devices that use Ni nano-powder, some catalysts, a loading current and heat, and nothing particularly radioactive or difficult to procure.

Also findings of direct EMF generation were reported recently, which could potentially bypass the heat-to-electricity step and the wasted energy in the Carnot cycle altogether.

LENR is a fundamental shift in the energy storage density the author is talking about.

http://www.slideshare.net/fullscreen/tylervan/lenr/2

LENR is the new baggage-free term for Cold Fusion. Extraordinary claims like yours need extraordinary proof. Please provide some.
(comment deleted)
http://psiram.com/en/index.php/Cold_Fusion

LENR is a fundamental shift from science to pseudo-science

I have searched for LENR and what I get is this wikipedia entry:

http://en.wikipedia.org/wiki/Cold_fusion

"By late 1989, most scientists considered cold fusion claims dead,[6][7] and cold fusion subsequently gained a reputation as pathological science.[8]"

"A small community of researchers continues to investigate cold fusion,[6][11] now often preferring the designation low-energy nuclear reactions (LENR).[12][13] Some have reported that, "under certain extreme conditions", they observe excess heat effects by interaction of hydrogen or deuterium with palladium, nickel or platinum.[14] Since cold fusion articles are rarely published in refereed scientific journals, the results do not receive as much scrutiny as more mainstream topics.[15]"

Batteries have improved significantly since 2008, and are still improving.

The most important shifts have been in cost ($/kWh), but there have also been improvements in energy density (Wh/kg), power density (W/kg), and cycle life.

Some of this is due to electronics improvements, but a lot is due to better cell design and manufacturing, funded in part by higher demand.

The advance from NiCd to NiMh chemistry gave approximately a doubling of watts per kg. So did NiMh to LiPo. It's hardly Moore's Law, but it's a significant rate of improvement. I can't imagine what smartphones and tablets would look like without the development of LiPo. We've seen substantial improvements in those battery technologies, perhaps most usefully Sanyo's development of low-self-discharge NiMh cells, which vastly broadened the useful market for rechargeable batteries.

The rate of change seems glacial compared to the rest of the computer industry, but it's unfair to imagine that no change is occurring. Properly sexy stuff is happening in battery technology and we'd be extremely foolish to imagine that batteries are at a dead-end.

The author addresses this in the article. The claim is that the battery chemistry itself hasn't improved, just our ability to keep it stable has which has allowed us to explore ideas that were previously not feasible due to a limitation with charging/discharging electronics. The assumption that there is little to no further progress to be made in this area (charging or chemistry). So, while battery technology may be improving, it will continue to improve slowly or not improve at all as there's a limit to how much more improvement we'll get from better charging/discharging circuits. Whether that's a valid claim is debatable, I guess, but I share the same opinion personally.

Now, we could see a super capacitor or something like that one day which may change things, but so far those have yet to materialize.

Claiming that battery chemistry itself hasn't improved is like claiming that the physics constants haven't changed... scratch that, it is indirectly claiming that physics haven't changed. This seems like it's a few steps from a tautology.

If you look at this graph ( http://en.wikipedia.org/wiki/File:Energy_density.svg ), which is from a wikipedia section linked in the article, there are materials that hold more energy both per unit volume and unit mass than gasoline, we just haven't been able to exploit them yet. Isn't that the whole of materials engineering progress?

I like "Fat Metabolism" sitting up there near diesel and gasoline. Imagine a fat-powered fuel cell that can draw from our own bodies....
So long as it doesn't smell like ketonic B.O.
> a doubling of watts per kg

Do you mean watts (power or rate of conversion) or watt hours (total energy stored)? I'm not positive but I believe you said it correctly, which is to say that the NiMh batteries don't store more energy than their NiCd predecessors. They can only discharge it faster (can push a higher wattage but still has the same total energy capacity).

I'll second this; it's the first thought I had when I started reading TFA. One thing I'll add, though, is that Sanyo might not have been the first (or at least the only one) to start pushing low self-discharge batteries. I've got a whole bunch of Powerex Imedions (not linking because I'm not a shill, just a satisfied customer; google it) that are awesome. Keeping 85% of their charge over a year of storage has made me happy. Being able to swap to rechargeables in everything and getting rid of the alkalines that _always_ end up leaking (yes, even the name brands; especially the name brands) has made me very happy.
The book "Powering the Future" also argues fuel will always outperform batteries, pointing out that both fuel and batteries store chemical energy, and that chemical energy density is fundamentally contrained by physics:

Since all atoms are approximately the same size there is a limit to how many bonds/volume you can have, and since all chemical bonds are roughly the same amount of energy there's a limit on how much energy/volume you can store with chemical energy.

The book then argues that hydrocarbon fuels come very close to the ideal energy density, and have many other advantages over batteries (weight (they're formed with light elements), extremely safe byproducts CO2 H2O, simple ingredients).

From an energy policy perspective that sounds about right, but from the standpoint of what's theoretically possible that would seem to be ignoring some really promising areas of research.
A 4 years old article on the front page of HN, with an attention-grabbing headline and no sign of serious research into the topic? Come on HN!
He's right. I just got through the battery course at Stanford Online and batteries suck. Downer.
You only need a battery to support you while you are off-grid.

Electric trains don't run on batteries.

Wireless charging is getting better.

Electricity which is "too cheap to meter" would mean we didn't need 'great' battery tech, current tech would be ok.

The other way to solve this problem is to say "burning hydrocarbons have great energy density, and the global transport industry is already tooled up to use them, so lets use those as our 'battery' tech", and just find:

- a good way to turn our global energy output into stored chemical energy in hydrocarbons

- a good way to use hydrocarbons in small-scale non-transport devices (phones?)

Man, every time great new technology comes along you guys have to come in and bust it all up with your logic and reason :)
Hmm, didn't know there is a Palladium acid battery.
Not sure if that's the joke flying over my head, but I believe that's a typo, and should be Pb (i.e. lead) acid.
I realised it was a typo, I was just being facetious.
That graph shows Li-ion as the highest energy density, which is in accord with the article. What am I missing?
(comment deleted)
The trend. The point is that energy densities are improving steadily, contrary to the article.
Which is a bit moot because, going by the trend, it's not until ~ the year 2220 that batteries will equal petrol for energy density by weight.
I agree with OP if your timeline is the next few years, but there is still a ton of headroom between, say, lithium polymer and the theoretical maximum for specific energy (the speed of light squared).

Obviously nuclear batteries are going to be a long time coming (I'm not even aware of any theoretical way to make, say, a fission reaction release electrical rather than thermal energy), but if you expand on the definition of "battery" a bit from something you just plug in to recharge to something that maybe requires being physically swapped out and reprocessed, there are some very promising avenues of research.

The one I'm most excited about is the vanadium boride-air cell, which has a practical energy density on par with diesel fuel. The catch is that is likely impractical to make one that can be recharged by simply reversing the current through it, so it would require a network of reprocessing plants and battery-exchange facilities.

Perhaps the direction of improvement at this point is lifetime rather than energy density. 500 cycles was fine for portable electronics. Not so much for vehicles.

Double the lifetime for a 10% increase in upfront costs would be a huge win. It isn't sexy enough to make it into the news, but there is progress here.

The post was written in 2008. Has there been any changes in the past five years to indicate any fundmental breakthrough in energy density? I know Tesla has come out with practical electric cars, but was that an improvement in energy density or an improvement in energy efficiency and management, as the article argues?
Biological organisms have quite a lot of energy stored in a small amount of space.

A lot of inventions were based on nature, e.g. wings for flying. Who knows, one day maybe a battery invention will also be based on biology.

Biological organisms aren't batteries. We manage to store lots of energy by storing it chemically. Internal combustion engines are already based on this. I burn fats, proteins, and carbohydrates for energy. My car burns gasoline. The issue is that we want to get away from that design, because it's not particularly sustainable.

If we could come up with a reasonably-efficient way to produce gasoline (or an equivalent) in a renewable way, that would be wonderful, but it wouldn't be a battery.

Using something like cytochrome c reductase to turn hydrocarbons (bio- or geo-) into electricity and then into motion would be a big efficiency improvement over our current heat engines.

Still, it's only half the problem. We also want an efficient and compact way to turn electricity into energy-dense chemicals. AFAIK, that doesn't exist in nature; we'd have to start from scratch.

Has anyone looked into using ATP in a battery? It seems to work really well for cells.