It is estimated that we would need 1.1 million square kilometers of solar panels in order to power the world – smaller than the state of Alaska at about 1.7 million square km. Of course, we would want to spread these solar panels out as much as possible. Rooftop solar alone would provide much of this needed area. In the US there is an estimated usable rooftop area for about 1 terrawatt of production capacity, compared to 1.2 terrawatt total current us power production capacity. Solar is also currently the cheapest form of new energy, in fact the cheapest source of electricity in human history.

Now come the well-known caveats – solar panels only produce energy when the sun is shining. So capacity is misleading as solar panels only produce electricity during the day when there isn’t too much cloud cover, and they produce less energy in the winter than the summer. At low solar power penetration this does not matter much because solar can displace more expensive and more polluting sources of energy when the sun is shining. We can also partly shift some of our energy consumption to sunny times – run the dishwasher or laundry during the day.

There are basically three ways to solve this problem. One is to have on-demand and baseload power to cover a good chunk of energy consumption, essentially limiting the percentage of intermittent sources. Another is to install overcapacity (more capacity than is needed at any one time) and share electricity over a broad grid. This works best for wind power, as the wind is always blowing somewhere, but could also help with solar. With a large enough grid solar power generated in Arizona could cover peak demand in New York. The third method is grid storage, storing up energy during producing and then releasing in when needed. We will likely need all three methods to get to net zero carbon energy production as quickly as possible.

But let’s focus on grid storage. One interesting question is how much solar and wind can we have in production before we essentially need grid storage? Actually, the better way to frame this is – how much grid storage do we need to achieve different levels of intermittent power penetration? The more grid storage we have, the higher tolerance the system will have for intermittent sources. This, apparently, is a very complicated calculation and I could not find any simple numbers to tell the tale. But it does seem clear that if we want to get to 50% intermittent energy sources or greater we will need increasing amounts of grid storage.

There are also different kinds of grid storage. There is distributed storage, such as home-owners having battery backup in their home. This can be used not only for emergency power but for peak shaving, storing energy during off-peak hours (or during the day if you have solar) and then using battery power during peak consumption. We can also have centralized grid storage – a massive system operated by the utility company to store intermittent energy production. We can also distinguish grid storage by duration. Having enough battery capacity to store energy for 4-6 hours is enough for peak shaving, and would have a tremendous benefit for balancing the grid. However, if we want a significant amount of solar power, then we also need seasonal storage, especially for countries farther away from the equator. This involves storing energy in the summer for use in the winter.

Right now we are using lithium ion batteries for grid storage, but they are expensive. There are many non-battery types of grid storage, like pumped hydro, melting salts, and spinning up flywheels, but none of these will have the capacity we need to massive grid storage. What we need is a successor to the lithium ion battery that is cheaper, uses Earth-abundant material, is safe (doesn’t catch on fire), can have many charge-discharge cycles, and can hold energy for months without significant loss. It also has to be fast-charging enough to take a lot of energy during high production. Two recent studies both turn to aluminum as a possible solution.

Aluminum is a relatively light metal (although not as light as lithium), and is the most abundant metal on Earth (and the second most produced metal after iron). Aluminum also has very good energy storage capacity. One study, led by scientists at MIT, looks at aluminum-chalcogen batteries, which uses aluminum and sulfur as the electrodes and a low-melting point salt as the electrolyte.  They report:

In their experiments, the team showed that the battery cells could endure hundreds of cycles at exceptionally high charging rates, with a projected cost per cell of about one-sixth that of comparable lithium-ion cells. They showed that the charging rate was highly dependent on the working temperature, with 110 degrees Celsius (230 degrees Fahrenheit) showing 25 times faster rates than 25 °C (77 °F).

Weight is not really an issue as these would be stationary. As a bonus they found that this battery has very little dendrite formation, a type of crystal formation that reduces capacity and can short circuit batteries. Sounds like this approach has potential, but of course until it scales up to production level we never know.

Another study from Swiss scientists just uses pure aluminum in an aluminum redox cycle to store energy. The advantage of this approach (in addition to aluminum being cheap and abundant) is that this system would have 50 times the capacity of lithium ion batteries. Also, they can store this energy for months, and the block of aluminum can be easily shipped around. This approach could therefore work for seasonal storage.

I don’t know if either of these exact approaches will end up being the ones that are used. Perhaps they will in some capacity, with other battery designs being used in different locations and for different purposes. I am always in favor of a system that uses many methods, each optimized for their use, rather than trying to force fit one solution for everything. Beyond these two studies, there is a lot of research into aluminum as a substrate for various battery designs. It would not surprise me if some type of aluminum battery plays a massive role in the future of our energy technology.