Battery Storage For Grid Backup: Better Keep Working On It

Advocates of generating electricity mostly with intermittent wind and sun, when challenged on how they would deal with a calm night, are always ready with the obvious answer: energy storage. Just get some batteries, store up excess power from the windy mid-days, discharge as needed, and everything will work out.

Unfortunately, the advocates never acknowledge that the problem of making an electrical grid work 24/7/365 with mostly wind and solar generation is much more difficult than just storing power from the day to discharge that night. Both wind and sun are subject to regular “droughts,” just like rain. There can be many consecutive days, or even weeks, of combined low wind and sun; let alone the entire winter has a lack of sun, and both summer and winter have less wind than spring and fall. Calculating how much energy storage will suffice to get through even a year of average wind/sun variability is a straightforward exercise, yielding an answer of as much as 1000 hours of average consumption. Meanwhile, naive politicians (those in New York being Exhibit A) regularly get duped into buying a few hours or tens of hours worth of batteries for grid backup, spending billions of dollars on amounts of storage that will be almost useless for backing up a primarily wind/sun grid.

I first wrote about this subject way back in 2018, and have had many follow-up pieces since. The conclusion of my pieces has been that to obtain sufficient battery storage to back up a primarily wind/sun grid using current lithium-ion battery technology, and even assuming best case future cost reductions and economies of scale, would cost the full GDP and more of any jurisdiction that makes the effort.

Well, who says you can only use lithium-ion technology? The massive Biden-era green energy handout statutes (e.g., the “Inflation Reduction Act”), together with green energy enthusiasm generally, have brought forth a gusher of entrepreneurialism looking for new, better and cheaper energy storage systems. Recent comments on some of my posts, as well as those of daughter Jane over at @janementonnyc on Instagram, have advocated for two new technologies of energy storage as the solution to the intermittency problem. Those two are flywheel batteries, and iron-air batteries. Could either of those really work?

As background to discussion of those two specifically, I suggest thinking about the model of the storage of drinking water. New York City stores water in a network of reservoirs located in upstate New York. The City’s water consumption is approximately 1 billion gallons per day. The storage capacity of the reservoirs is approximately 550 billion gallons, that is, approximately 550 days, or more than a year and a half of consumption. The amount stored in the reservoirs fluctuates over the course of a year, and generally drops over the summer and into the fall, but it rarely gets below about 70% of capacity, or about 380 billion gallons. And in years with serious droughts, the storage can fall below 50% of capacity, and even down to 40% of capacity. A storage level of under 40% of capacity has only happened once in my lifetime, which was about 60 years ago. In other words, most of the storage capacity is there to guard against a worst-case drought, and much of the water remains in storage for decades on end to guard against that event. Fortunately, a simple reservoir has the capability to do that.

An electrical grid without full dispatchable backup needs the same kind of storage capability. The fact that a group of generators can produce the same number of MWhs of energy in a year as the average amount demanded means little unless the energy can be matched minute by minute to the demand. To meet that criterion, a storage system must be able to store the energy from summer to winter, or from spring one year all the way to spring the next year. Preferably, there should be an ample balance stored for the long term to guard against a worst-case wind/sun drought that may occur only once a decade.

By the way, it is by no means clear that lithium-ion batteries have this level of capability. But for today, let’s consider the technologies advocated by our commenters, flywheel and iron-air.

Flywheel batteries. Flywheel batteries have lots of advantages. For example, they can discharge a very high percentage of the energy originally stored in them (90-95%), and can be charged and discharged potentially thousands or even tens of thousands of times without seriously degrading. Moreover, they can ramp up and down quickly to replace generation from intermittent sources; and they have spinning inertia, which wind and solar generators do not, and which is badly needed for grid stability.

But unfortunately flywheel batteries have very high rates of what is called “self-discharge,” that is, dissipation of the stored power over time. According to this source (something called Permanent Energy), many flywheel batteries lose as much as 12.5% per hour, and even the best ones lose about 5% per day. Other sources give me similar answers. In other words, energy stored in a flywheel battery will be long gone a month later, even if never called on. Flywheel batteries may have many uses, but for purposes of backing up the grid against any serious wind/sun drought, they are worthless. Oh, and they are expensive — currently costing in the range of $400/kWh, which is more even than lithium-ion batteries and translates to many trillions of dollars to buy amounts useful for full grid backup.

Iron-air batteries. This is a type of battery that uses only the simplest and most common of materials — iron and air. The process of storing and discharging energy takes place by repeatedly rusting and unrusting the iron. It turns out that you can store a lot of energy that way. It’s not subject to exploding or catching fire. And it’s cheap: proponents claim that they will be able to achieve a price of $20/kWh, which is a small fraction of the current price of lithium-ion batteries (~$300/kWh). So what could possibly be the problem?

Again, self-discharge is a killer. Current iron-air batteries lose about 2-5% of their stored charge per day. At that rate, all the stored charge will be gone in two months, if not one. Maybe the rate could be improved, but it would need to improve by multiple orders of magnitude to make these batteries even a little useful for large-scale grid backup against worst-case droughts.

And then there are a few other problems. Iron-air batteries can only return about 50% of the energy stored; the rest is lost. And they can only discharge about 1% of the stored energy per hour. That means that they are incapable of ramping up and down quickly to back up the intermittent wind and sun.

Now I’m not at all saying that these two sorts of batteries cannot be useful in certain applications. For example, flywheel batteries appear to be very useful to supplement diesel engines to operate heavy cranes. The cranes go for long periods at low energy, and then have a big surge in power demand when they suddenly lift a heavy load. Pairing a flywheel battery with the diesel engine can cut the size of the needed engine by as much as half.

But why anybody, let alone our commenters, thinks that these sorts of batteries are the answer to grip backup for wind/sun generation, I do not know. Maybe some day. Meanwhile, keep working on it. As I have said before, if someone figures out a battery technology that has the needed capability and is also cost-effective to make a grid work with wind/solar generation, I will be the first to applaud.