Also, here’s an interesting paper that examines the historical rate of battery improvement (~3% energy/$ per year for the past 60 years), and tries to connect it to : On improvement rates for renewable energy technologies: Solar PV, wind turbines, capacitors, and batteries
I plan to write more later, but here some initial plots and thoughts:
+Over the last decade or two, batteries have become ubiquitous in consumer electronics. However, this is mostly from improvements in electronic energy efficiency, not improvements in batteries.
+Since the commercial release of lithium ion batteries in 1991, energy density has only increased by about 1.5x.
+Lithium ion energy density may still double with cleverer engineering, but it can’t go too much further–we’re already in the neighborhood of its theoretical maximum.
+Although energy density progress is slow and limited, there may be more potential along the dimensions of cost and cycle life.
+My (possibly wrong) impression of cost is that the cost is mostly driven by materials cost. This means that there is limited upside to improving assembly efficiency (for example, the aspirational goal of Tesla’s Gigafactory is a mere 30% reduction in cost).
+As for cycle life, I personally see no reason why breakthroughs couldn’t happen. Some promising claims have been made by ‘ultrabattery’ folks who combine lead acid batteries with supercapacitor electrodes. However, it’s not clear that these claims will pan out. Also, I’ve been warned by experts who know more than me that cycle life is actually a very difficult thing to improve because you’re fighting against thermodynamics. Despite this warning, I remain open-minded.
+Even though I may sound pessimistic about improving energy density and reducing cost, I nevertheless believe that small changes could have huge impact. A factor of 2 improvement sounds lame when measured on the scale of Moore’s law, but when you consider that batteries are on the edge of being cost-effective in a number of applications, such as vehicles and grid storage, there is a large sensitivity of demand to improvements in technology.
+Lithium is the smallest non-gas atom we have, so it makes the most sense as a battery ion. (Though there are other creative ideas that have the potential to do better than lithium for various reasons. One such idea is to use magnesium, which can potentially carry twice as much charge as lithium with less than twice the volume.)
+Fortunately, lithium is decently abundant. There is enough economically extractable lithium to build an electric car for everybody someday, though that would severely stretch current estimated reserves. Unfortunately, despite being common, lithium is also dilute, which makes it relatively more expensive to produce.
+Unlike many other metals, which are mined from hard rock in energy-intensive processes, lithium is often captured from brine on salt flats, at least these days. I don’t know what drives cost for brine and how much these costs could change in the future. Brine vs hard rock mining infographic
+I am pessimistic about the market for home battery packs (Tesla’s Power Wall, for example). First, battery packs are still too expensive to be economic (though there are some interesting non-obvious value propositions like being able to turn off the grid during dry storms to prevent fires). Second, even if home battery packs do become economic, I imagine it would make far more sense for a utility or other company to do it at scale than to individually install packs in people’s homes. The benefit to co-locating storage is not large. Third, the only way I do see it making sense is if there is inefficient electricity pricing/regulation that indirectly incentivizes home battery packs. But if that was the case, I’m confident the utility would change its pricing/regulation if home batteries ever started to really take off.