You know how technology works: a computer capable of calculating space flight used to fill a building, and now it fits in your pocket. Your “record collection” had its own wall a few decades ago; now you can pass out as party favors storage with more music than you'll hear in the rest of your lifetime.
So, with venture capital and the other usual suspects turning their attention to, among other things, batteries, a peanut-sized power pack will soon run your car and permanently energize your tablet. Right?
Wrong. Badly wrong. While there's a lot of progress to make, the physical limits are clear. That's not, "The limits with existing techniques" or "We forgot the square root that showed up in a Science Fair poster that will save us." The limitations are: “We know how an electron dances with more precision even than the population at large has about the latest 'American Idol' scores, and there's just no room to wiggle.”
Here's why.
Fundamental Physics and Engineering
There are two crucial concepts to understand in considering batteries. One is their fundamental physics, and the limits those fundamentals impose. The second is how well we engineer existing devices. That is: what kind of performance we get from batteries now, and how close that is to the fundamental limits.
Batteries are improving now, a lot; maybe, in extreme cases, by a factor of ten. A factor of ten is valuable, and worthwhile. It is not the factor of a thousand or a million to which computing experience habituates us. Computing improvement isn't done yet; memory will almost certainly improve by another factor of at least a million in the coming years. We measure computing in information units, where the physics tell us what history suggests: We're only getting started.
Batteries, though, are about energy conversion. Our scientific knowledge is deeper about energy, and it is ultimately much more tightly constrained. (The energy, that is. Our knowledge always has room for expansion.)
A battery has lots of requirements: It shouldn't explode; it needs to supply a predictable voltage; it needs to operate in the conditions typical for human life (not too hot, not too cold, in proximity to people, and so on); and we need to be able to manufacture it at a reasonable cost.
All these requirements are secondary to a straightforward measurement, though: How much energy does the battery of a particular weight deliver? If that number is too small, nothing else matters. No one pays for a battery that only keeps an iPod going for ten minutes. If the number is big enough, we find a way to adjust the other constraints. A half-ounce battery that would keep a laptop going for two days would fly off the shelves, even if expensive, difficult to recharge, or otherwise unpleasant.
Here are the measurements and calculations, then, for specific energy:
|
Technology |
Energy density, in kilojoules per gram |
|
Theoretical hydrogen combustion |
123 |
|
Theoretical gasoline combustion |
47 |
|
Theoretical propane combustion |
46 |
|
Theoretical lithium-air battery |
40 |
|
Theoretical aluminum-air battery |
19 |
|
Laboratory lithium-air battery |
9 |
|
Laboratory aluminum-air battery |
5 |
|
Laboratory hydrogen-consuming fuel cell |
1.6
|
|
Laboratory lithium battery |
1.3 |
|
Commercial lithium-ion battery |
0.7 |
|
Theoretical lead-acid battery |
0.6 |
|
Commercial alkaline battery |
0.6 |
|
Laboratory graphene-based supercapacitor |
0.6 |
|
Laboratory flywheel |
0.5 |
|
Commercial lead-acid battery |
0.1 |
|
Commercial flywheel |
0.04 |
The "theoretical" figures are absolute limits. There's no "Moore's Law" for chemistry that results in gasoline burning at a different temperature next year than it has in the past. If "battery" means anything like what it has until now — a package which produces electricity from consumption of electrochemical reactants — we can calculate precisely how strong any battery can be.
Even that "theoretical" limit is simplistic. It doesn't account at all for packaging, electrolyte, or conductors. Lead-acid batteries achieve a sixth of theoretical limits, and at the cost of mass exposure of toxins including lead and sulfuric acid; we won't soon be as lucky in our engineering of higher-performance reactions.
It's possible to fantasize about energy sources that depart from usual chemistry. Low-temperature fusion, for example, or anti-particle bottles, could, in principle, open up whole new landscapes for energy supply. Mastery of such techniques would result in even more momentous changes than a battery pack that survives a trans-oceanic flight, though. We'll leave the science fiction to another day.
Engineering within the Limits
No after-market drop-in will extend the range of our current automobiles and laptops to run for days without recharging. How much better can we take advantage of the chemistry we have, then?
Think of a power circuit in its most abstract this way:
- Source
- Controller
- Sink
For now, our greatest gains are on the "sink" side, where electricity is used. Laptops can produce all the same results they usually do, while using less electricity for their displays, memory, CPUs, and auxiliary processes. With a few more rounds of advancements, one can imagine "sustainable" laptops, where much of the power comes from capture of part of the power in keyboard strokes and ambient light.
Lighting up a conventional display, though, will remain a barrier. Even with improved technologies, active illumination of a surface to render pictures visible to nearby eyeballs will remain an energetically-wasteful way to communicate with human brains. Improvement by a factor of over a hundred will require a switch to a new model.
Some of the most satisfying gains of the last couple of decades home appeared in the least dramatic segment of our model: the middle, controlling one. Naive power converters and conditioners of 50 years ago were frequently over 50% efficient, so the energy improvements are necessarily less than a factor of two. The greater impacts of controller advance have been and will be in other dimensions: reduction of heat loss, weight, and manufacturing cost; and improvement in reliability and safety.
Breathless press releases about battery breakthroughs will continue to appear. At least some of them will lead to worthwhile products. Keep in mind, though, that, at their best, even with miraculously good engineering fortune, batteries can only improve by modest factors in the energy density already commercialized.
See also:
