How the Lights Stay On · Catching Lightning in a Bottle

Catching Lightning in a Bottle, part 2

If you do the math, that works out to about ten to fifteen cents per kilowatt-hour cycled through it — on top of whatever the electricity cost to begin with. In most of the United States, retail electricity costs twelve to thirty cents per kilowatt-hour. So in many cases, the Powerwall does not pay back simply on the value of storing cheap electricity for later use. The savings from time-shifting the electricity are roughly canceled by the cost of the battery itself.

This sounds bad. It is actually a more nuanced story.

A home battery makes economic sense in most American homes today for two reasons that are not pure arbitrage.

Backup power. When the grid goes down — a storm, a wildfire-related power shutoff in California, a winter outage in Texas — the Powerwall keeps your fridge running, your phone charged, and (depending on size) your air conditioning going for some hours. For people who live in areas with frequent outages, or for people who depend on medical equipment, this is worth more than the pure arbitrage math would suggest.

Time-of-use rates. In some states — especially California — your electric bill is no longer a flat rate per kilowatt-hour. It depends on the time of day. Peak rates between 4 PM and 9 PM can be three to four times higher than off-peak rates. Under those rate structures, charging the battery from cheap midday solar and discharging it during the expensive evening hours can produce real savings — enough to make the Powerwall worthwhile.

And from the previous chapter: whether the Powerwall is even installable in your house depends on your service panel, your wiring, and whether your utility allows it. New construction is easier. Old houses may need a panel upgrade first.

The honest summary, the one you will not always get from a Powerwall salesperson, is this: a home battery is a hedge, not an investment. It pays back if you value resilience and live somewhere with the right rate structure. It does not, in most places, pay back purely on the math of buying cheap and selling dear. The cost-benefit changes every year as battery prices fall and rate structures evolve, so the answer for your house may be different next year than it is today.

Lithium-ion is not the only kind of battery, and batteries are not the only kind of storage. Engineers are exploring a whole zoo of options, each with different trade-offs.

Flywheels. Big spinning wheels suspended in vacuum chambers. You spin them up with surplus electricity; you slow them down to recover the energy. Round-trip efficiency is high. Duration is very short — seconds to minutes. They are used for the fastest grid services, like keeping frequency stable. Remember the inertia problem from Chapter 2? Flywheels can help with that.

Compressed air. Use cheap electricity to compress air into underground caverns, often abandoned salt mines. Release the air later to spin turbines. The technology has been around since the 1970s. Round-trip efficiency is mediocre — maybe 50 to 70 percent. But the caverns are huge, so the storage durations can be very long.

Thermal storage. Convert excess electricity into heat — molten salt at 1,000 degrees, hot rocks, sand, even a giant tank of hot water — and reconvert it to electricity when needed. The efficiency depends a lot on what you are doing. If you can use the heat directly (for industrial processes or for heating buildings) instead of converting back to electricity, efficiency is excellent. If you have to convert it back, it is lower.

Gravity storage. Newer companies are experimenting with raising heavy weights up towers, then lowering them slowly to drive generators. This is the same principle as pumped hydro, but without needing a mountain or water. The technology is still being demonstrated; the economics are unclear.

Hydrogen. Use excess electricity to split water (H₂O) into hydrogen (H₂) and oxygen. Store the hydrogen. Burn it later in a gas turbine, or run it through a fuel cell. Hydrogen can be stored almost indefinitely, which is wonderful. But the round-trip efficiency is poor — only 30 to 40 percent of the original electricity comes back. Most of the energy is lost in the conversion both ways.

Iron-air batteries. A new kind of battery, very different from lithium. Iron-air cells use iron and oxygen — basically, rusting and unrusting iron on purpose — to store energy. They are slow. They are cheap. They have terrible energy density (you would never put one in a car). But they may be able to store energy for 100 hours at a tenth of the cost of lithium. The company Form Energy is building the first commercial iron-air battery installation in Minnesota.

Why so many approaches? Because no single kind of storage solves every problem. The grid needs storage at every time scale, from seconds (flywheels) to hours (lithium) to days (pumped hydro) to seasons (hydrogen, iron-air, maybe). The future grid will probably use a portfolio of these technologies, picking the right tool for each job.

If you remember one thing about storage, remember this:

Lithium batteries are great for four hours.

The standard size for a utility-scale lithium battery is something like 100 megawatts of power for four hours, which means 400 megawatt-hours of energy. After four hours, the battery is empty, and it takes a few hours to recharge.

Four hours is enough to handle a daily duck-curve cycle — fill the battery at noon, drain it during the evening peak. This is what California’s 13 gigawatts of batteries are doing. They are flattening the duck.

But what happens when the wind does not blow for a week? What happens when a winter storm covers your solar panels in snow for days? What happens during a cloudy stretch in February with no wind?

You cannot solve a week-long shortage with a four-hour battery. Even if you had a thousand of them, you would drain them all on the first afternoon.

To run a grid on mostly solar and wind, you need something that can store energy for days or weeks, not just hours. This is the great unsolved problem of storage.

The candidates for long-duration storage are exactly the technologies described in the previous section that no one has yet deployed at large scale: hydrogen, iron-air, compressed air, thermal. Some of them will work. Some will turn out to be more expensive or more limited than their promoters claim. We do not yet know which will dominate.

If you are a teenager reading this book and thinking about a career, this is one of the genuinely exciting open problems in the energy world. Whoever solves long-duration storage cheaply will probably make the energy transition possible, and might also become extraordinarily wealthy along the way.

A book about batteries has to be honest about one more thing: where the materials come from.

Lithium-ion batteries require lithium (mostly mined in Australia, Chile, and Argentina, with processing concentrated in China), cobalt (mostly mined in the Democratic Republic of the Congo, where some mines have been associated with serious labor and human rights abuses), nickel (Indonesia and the Philippines), and graphite (mostly China).

The supply chain is concentrated. China processes the majority of the world’s battery materials, even when those materials come from somewhere else. The mining itself uses water, energy, and land. The labor practices in some mines have been documented as abusive, particularly in cobalt mining.

Newer battery chemistries are reducing or eliminating some of these problems. Lithium-iron-phosphate (LFP) batteries, which dominate the Chinese market and are spreading globally, do not use cobalt or nickel. Iron-air batteries, if they scale up, will use one of the most common materials on Earth. Sodium-ion batteries — which trade some performance for using one of the most abundant elements there is — are starting to appear in commercial products.

But it is fair to say that today’s battery boom is, in part, built on mining. There is no version of “clean energy” that involves zero environmental impact. The honest framing is this: the climate damage from mining lithium is much smaller than the climate damage from burning coal, and the human-rights problems can be reduced through standards and oversight — but the damage is not zero. It is something we are choosing to do because the alternative is worse.

This is a real trade-off, and you deserve to know about it. The world is not all good guys and bad guys. The world is people making choices about which problems to live with.

For a hundred years, the grid had almost no memory. Whatever was generated had to be used in the same instant. The grid was a system without a clock — it could only ever live in the present.

Storage gives the grid a clock. It gives the grid a kind of memory.

The noon sun in California can now be stored and released at dinner time. The midnight wind in Texas can be stored and released at noon. The duck is starting to flatten. The synchronized machine from Chapter 2 is learning to wait.

We are not there yet. Lithium gives us four hours. Pumped hydro gives us a day, but only in mountains. We do not yet have an economical way to store electricity for a week. The grids of 2050 will need that capability, and the engineers and entrepreneurs and policy people of the next twenty years are the ones who will have to invent it.

If you find yourself looking up at a Virginia mountain hillside with a quiet reservoir near the top, or driving past a fenced-off field of gray Tesla Megapacks humming in the desert heat, look at it for a moment. These are the experiments that may turn out to have built the world.

The next time the sun goes down behind your house, the grid still has to come up with the next few hours of electricity from somewhere. Increasingly, that somewhere is a place where, hours ago, the same sun put energy in.

The grid is learning to remember.

All 18 chapters of How the Lights Stay On · Library · All courses