Some solar + battery math
This morning, I saw some Very Intelligent Man in the YouTube comments sneer “A Tesla Megapack is only 4MWh and costs a million dollars. Do the arithmetic!” I think this is a great idea. Let’s do it!
His numbers are round but close enough. Tesla Megapacks come with a standard 15 year warranty, which can be extended to 20 years, so I think it’s safe to say they can run for 20 years. A 4MWh battery pack that gets 100% utilization (4MWh/day), 365 days a year for 20 years, can time-shift 29,200MWh (29.2GWh) of clean solar power.
On average, solar panels have about a 20% capacity factor - how much power they really produce relative to their theoretical output, over long time spans (like a year). So a 1kW solar panel will produce something like 4-5kWh/day, on average. So 1MW of solar panels would be enough to charge the battery, on those “average” days. But not all days are average. On a cloudy winter day, capacity factor could be down to 5%. So let’s say 5MW of solar panels, to be sure we can charge the Megapack fully even on bad days.
Solar panels are currently about $0.11/watt wholesale. Throw in retail and installation markups, and let’s say $200k/MW, or a million dollars to build the 5MW needed to insure we can always charge our battery. So now we have $2M in up-front costs, to provide 29GWh of firm power over 20 years. That six or seven cents per kWh. Contrast with the ten or eleven cents per kWh we currently pay in the US (varying widely, depending on which state).
Of course, battery demand will vary by weather and season; not all batteries will need 100% utilization. Batteries that get 50% utilization would cost twice as much, 12-14c. Batteries only getting 10% utilization during the dark days of winter would cost a lot more. That’s the bad news. But now, the good news!
First, remember we’re overbuilding our solar power supply by 5x to guarantee we can keep those batteries charged daily, even on cloudy winter days. But on average, we would be producing 5x more solar power than needed! That an enormous amount of power that could be used for things other than charging batteries - if they don’t require year-round continuous power. It’s reasonable to estimate 100GWh of bonus power from this solar, over 20 years! What could we do with that much power? We could synthesize fuel, for one. If we could get even 10% end-to-end efficiency, making synthetic fuel and burning it to produce electricity, we’d get another 10GWh of clean, firm energy. (Yes, it’d probably mean another million dollars in fuel synthesis hardware to get that 10GWh. Still a good deal!) Surplus power could also be used for charging heat batteries for industrial use (shut down the factories on the worst days), or do CO2 capture.
And perhaps 5x is overkill. In some regions, lower latitudes or with consistently sunny weather, there will be less seasonal capacity variation, and we’ll only need 2x or 3x. But regardless, batteries mean surplus.
Second, lithium batteries aren’t the only game in town for storage. Grid batteries have different requirements than mobile batteries used in portable devices or vehicles. They don’t need minimum mass, or maximum charge/discharge rates. This allows for much cheaper materials to be used - iron/air, zinc, liquid flow batteries, mechanical storage, thermal storage, etc. The lower price makes multi-day battery storage more approachable as a problem.
And finally, keep in mind that once 24/7 renewable+battery is achieved, additional batteries have diminished cost effectiveness. A battery that is only used 50% of the time costs “more” than a battery used 100% of the time. A battery only used 10% of the time is even more expensive, and so on. At a certain point, using fuel is cheaper than adding more batteries. And, in the short run, we can just turn fossil fuel plants back on, using coal or methane to keep the grid going during extended periods. In the longer run, we can synthesize clean fuel to run in those plants.
But the fact that solar+batteries won’t totally eliminate fossil fuel plants in the short run does not mean we shouldn’t bother. Every watt of grid power generated by renewables is a watt not generated by fossil fuel, CO2 not released into the atmosphere. And batteries are incredibly effective at leveling out the “duck curve” and solving the diurnal problem of solar (the sun doesn’t shine at night). It doesn’t take a whole lot of battery to make a huge difference, either! Even the handful of gigawatts of battery that California has deployed recently reduced their fossil methane consumption by over 40% in 2024 alone! Just enough battery to run the grid 24/7 on nice days in the summer is progress - days or weeks at a time with no fossil fuel on the grid. More battery for the shoulder seasons isn’t a big stretch.
So don’t fall for the wildly exaggerated claims that we need weeks or months of batteries we can’t afford to handle ridiculous nightmare scenarios where the sun stops shining and the wind stops blowing for weeks or months at a time. Don’t listen to “What if there’s a huge volcano?” or “What if there’s an asteroid?” or “What if aliens build a shell around the Earth?” or any other such nonsense. Because right now, we’re not even able to do 24/7 renewables at a country level yet, and at a certain point, we’re just adding nines to our reliability. I’m not going to panic over only getting rid of 98% of the fossil fuel in the next 10-20 years, rather than 100%. You shouldn’t, either.