Years ago my home state of Colorado passed laws to provide electric customers with what is called “net metering.” That meant that if we installed solar panels, the electric utility would be required to credit us with any power we produced. And that credit would be given at the same rate per kilowatt-hour that we paid when we bought a kilowatt-hour from the utility! That was too good a deal to pass up. I dutifully installed a large bank of solar panels. And even though I will likely never get my investment back, it felt as if I was doing my part to help the environment. And, frankly, I still get a kick out of occasionally seeing $0.00 due on my summer electric bills.

Although I didn’t realize it at the time, if I think about it now, I am taking advantage of the system and maybe not even helping the environment that much. My summer electric bills are often zero because on a sunny day I generate much more power than I use. Happily for me, the utility takes my excess power during the day and gives me back that power at night. Of course, on cloudy days, or when the sun is lower in winter, my panels do not generate nearly enough power to offset what I use. But no worries; my lights stay on because the utility has other means to supply power.

This system is great for me but not so good for the utility or for their other customers. The utility must build and maintain a great deal of infrastructure—which other costumers pay for—so they can take my excess power and feed it back to me later. They are, in effect, acting as a giant free battery. And, obviously, this system would not work if too many of their customers started doing the same thing. That is true because the utility actually has no such giant battery and no way to store my excess power for use when it is needed. So my nighttime, cloudy day, and winter power comes from their coal and gas fired plants, which, therefore, they cannot shut down.

Wouldn’t it be wonderful if we could build enough battery storage to allow my utility to actually store my excess power so they could close down those dirty coal and gas plants?

It would appear that Americans are trying. In the last few years, there has been a massive buildout of battery storage by both utilities and homeowners. In fact, by the end of this year,Wood Mackenzie and the Energy Storage Association expect that homeowners and utilities in the United States will have almost 3,000 megawatts (MW) of installed battery capacity. That’s huge! For comparison, a Tesla PowerWall battery is rated at only 0.007MW (13.5KWh).

All that new battery capacity is remarkable progress to be sure. Now what we’d really like to know is how many existing power plants would all that battery storage—if used as the reliable backup for wind and solar—allow us to close? To answer that question, it’s helpful to look at a very specific example.

A couple of years ago, activists tried to close Arizona’s  Palo Verde nuclear power station. The idea was to replace it with renewables, primarily solar, to take advantage of Arizona’s incredible sunshine. The idea died, in part, because at the time, on a hot, windless night, there would have been no way to keep air conditioners, factories, and cities running.

But what if there were batteries to store excess power during the day and give it back at night? Let’s take a realistic look at how much battery storage would have been required, along with the size of the solar farms necessary to fully charge the batteries and provide 24/7 power to all their customers.

The Palo Verde plant generates about 4,000MW of 24/7 power. To replace Palo Verde’s power with solar would take at least a 16,000 MW solar farm, possibly more (because even that 16,000MW is only produced around noon on a sunny summer day). When it’s completed, the largest utility-scale solar farm in the country will be the Mount Signal farm currently being built in southern California. It will have a peak capacity of 800MW, but because of nights, clouds, and winter’s weak sun, when averaged over a full year, Mount Signal is expected to generate something less than about 150MWac. So to replace Palo Verde, simple math tells us that it would take at least 25 Mount Signal solar farms. But, because even Arizona has several cloudy days in a row, the battery backup would have to be large enough to supply full power for not just one night but for several full days or longer. To supply that power and ensure no blackouts, how much battery storage would be required?

To work out the answer, suppose we could hook together all the 3,000MW of battery storage expected to be in place in the United States at the end of 2019.  Then suppose that gigantic battery were charged to its full 3,000MW capacity.  When the sun went down for Palo Verde’s former customers, how long could that massive battery keep on supplying them with the same power the plant had been supplying?  The answer is 45 minutes! In other words, all that amazing battery capacity could backup just that one power plant for not even one hour. And, by the way: Have you heard about the largest battery in the world, that giant $66 million Tesla 129MW battery in Australia? It would last all of two minutes!

But if we consider what’s required to power the whole state of Arizona, it gets even more challenging. Palo Verde supplies only about a third of Arizona’s power. To keep the air conditioners, factories, cities, and towns running all over Arizona 24/7, it would take solar farms covering as much land as the entire state of Rhode Island—land which, once covered with solar panels, could be used for nothing else. And if Tesla’s giant Australian battery were used as the backup, it would take many thousands of them. Moreover, because of limits on useful life, solar panels have to be replaced in less than 25 years, and the batteries much more often. Worse, both the decommissioning and the mining and refining of the specialized minerals necessary to build just one replacement cycle generates millions of tons of waste that remains lethally toxic for millions of years (currently a daunting problem for China).

And that’s the problem. On a small scale, like rooftops and vacant land, solar, wind, and PowerWalls can be quite attractive. But the success of those small installations fools us into thinking that scaling up is just a simple multiplication problem. It is not. When it comes to providing 24/7 power for industry and cities, the scale of the problem overwhelms any possibility of meeting it with batteries. So even if we had the huge quantity of available land that we were willing to industrialize with solar and wind farms, they would still need a backup system capable of taking the full load. And because batteries will never have enough capacity for that job, the backup power will always have to come from conventional power plants. In other words, and in utility-speak, batteries are extremely valuable and profitable for peaking and load balancing, but they are completely unsuitable for providing long-duration baseload power.

If we are to truly get serious about decarbonization and closing America’s fossil fuel plants, we must acknowledge that solar and wind can get us only part way there. They must be supplemented, backed up, by some other technology capable of generating tremendous quantities of power. In short, every megawatt of solar and wind power needs a megawatt of backup power.

Climate scientists are telling us that time is of the essence. If we are to have any hope of meeting our decarbonization goals, we have no choice but to go with proven technologies that we know will work.

Frankly, safe nuclear plants like Palo Verde are the only carbon-free technology we know of that can provide reliable 24/7 backup power at scale. And even if a week is windless and clouds cover the sun, a nuclear generator doesn’t care—it just keeps on sending us the power we need, day after day, week after week, all year long.