During the early days of the Cold War, “Special Weapons” was the American military’s code name for nuclear bombs. The United States had some 30,000 of them spread around the world, and they needed to be serviced. In 1961, when I was pressed into service, the U.S. Navy decided I should become an expert in “Special Weapons.” In my freshly minted officer’s uniform, I was sent off for months of top-secret training and schooling in New Mexico. I learned a great deal about the science and engineering behind those instruments of mass destruction. But in the process, I developed a deep and visceral fear of all things nuclear. Four years later when I returned to civilian life, I was more than glad to put those fearsome things behind me and out of mind.

A decade later, the unthinkable happened. The Three Mile Island nuclear power station suffered a partial meltdown. It was just 50 miles upwind from my home. The TV and press reports were terrifying. Like many others, I applauded the demonstrations demanding NO NUKES! And I applauded again when popular pressure helped put a stop to plans that would have nuclearized the nation’s power grid. True, nobody liked all the new coal plants, but they were certainly better than scary nuclear.

Fast forward a few more decades.

Something called global warming began to capture more and more attention. Al Gore promoted a book and slide show prophesying environmental disaster if humanity didn’t change course. Altering this warming trajectory, he argued, would require, among other things, changing the world’s entire energy system and making an urgent commitment to stop burning fossil fuels. Concerned people began saying that our cars must burn ethanol from corn, and all our electricity must come from renewables.

To replace our dirty electric power plants with clean, renewably generated electricity sure sounded good. Corn was easy to grow, biomass (wood, predominantly) was abundant, wind and sunshine were free. And even better, many believed that renewable energy would also allow us to close our 100-odd nuclear power plants. So, the nation started down the path of creating clean energy from renewables.

Or what we thought was clean energy.

After converting over 10 million acres to growing corn for ethanol, reality set in. We discovered that repurposing cropland drove up food prices. Studies showed that over its full cycle growing and then burning corn ethanol actually created as much or more atmospheric carbon than gasoline. Corn ethanol could not be the answer.

Next, we tried biomass. But biomass, when looked at full cycle, like corn ethanol, tends not to be carbon neutral. Worse, the carbon released from burning biomass can take 30 to 100 years to be fully re-absorbed by the replacement forest. In the meantime, all that carbon is in the atmosphere heating the planet. As it turns out, burning biomass is almost worse than burning coal.

Hydro is terrific but, unfortunately, most of the U.S. hydro potential is already on the grid. Hydro is also dependent on long-term weather cycles.

That leaves massive deployment of wind and solar as our primary hope for renewable energy to replace fossil fuels. Can wind and solar, by themselves, really do all or most of that job?

Answering that question requires knowing how much electricity needs to be generated. Decarbonizing is not only about replacing existing power plants. Electricity generation represents less than 40 percent of the energy we use. In order to fully decarbonize any economy, it will be necessary to electrify transportation, heating, and more. Even our present electricity demand requires over 8,000 large power plants. Decarbonizing through universal electrification means we’ll need to more than double that generating capacity. Could wind and solar alone get us there?

Sadly, it is becoming clear that although they will certainly make a valuable contribution, to do the whole job, wind and solar will need a lot of help. Here are the dominant reasons:

Intermittency. All of the excitement over the plunging cost of solar panels and wind turbines rarely allows for the much larger system costs necessary to power a society primarily with unpredictable intermittent energy sources. Absent super-cheap and massive battery storage (for which we do not yet have the technology), wind and solar must be backed up by a reliable 24/7 power source capable of taking the entire load. Although it might seem that stretching powerlines across large geographies could relieve much of the problem, the data say that doing so does not help as much as we’d hoped. In Germany, even though on some days last year wind and solar provided substantially more than 100 percent of the country’s power, for the full year they supplied barely 29 percent. For a society that demands light anytime the switch is flipped, intermittent electricity is not an option.

Land. Our current experience with solar is very misleading. In spite of significant solar buildouts, in 2018, all our solar installations supplied barely 1.6 percent of U.S. electricity. That means that minimal land has been required so far and it’s been, therefore, inexpensive. But enough solar panels to power a fully electrified United States would require between 50 and 100 thousand square miles of land (a little perspective: Louisiana is about 50,000 square miles). And that land could be used for nothing else. Wind requires two to six times as much land as solar, although wind land can usually be multi-purposed. In both cases, however, once the best and cheapest sites are gone, additional land becomes progressively more expensive. And that’s without factoring in NIMBY problems or the huge amounts of land necessary to mine the rare and often toxic minerals needed to produce wind and solar equipment.

Wind and solar do not eliminate the need for traditional power plants. This is the most disappointing problem of all. Because of intermittency, a backup system must always be on standby, ready to seamlessly take over the full load the instant the wind stops or the sun sets. The dominant source of that backup in the United States is natural gas (methane)—an improvement over coal but still a carbon-generating fossil fuel. Perhaps that is why so many fossil fuel companies strongly support renewables even as ExxonMobil and other fossil companies are budgeting billions over the next decade to expand their fossil fuel production. They seem to believe that wind and solar are, far from being a threat, a way to ensure a continuing need for their fossil fuels.

Cost. A small amount of wind and solar can be extremely cheap, cheaper even than natural gas. But as soon as wind and solar constitute a significant percentage of generation, the systems cost—compensating for intermittency, running a standby fossil system, land costs—rises dramatically above any other form of generation.

Real-life experience demonstrates the problem. Fifteen years ago, Germany committed to transition to 100 percent renewables, even building enough to replace its nuclear plants. After spending almost $400 billion on renewables, they have seen little commensurate reduction in carbon emissions. And Germany now has the highest electricity costs in Europe. Similarly, as California has expanded renewables and closed its nuclear plants, its emissions have gone up and so have its electricity costs. Indeed, in places where emissions have declined, it has been mostly the result of conversions from coal to gas. Yes, wind and solar generate carbon-free electricity. But because they require backup, they lock infossil fuel power plants as an essential part of the electricity system.

Considering the fleet of new coal and gas plants currently being built around the world, carbon capture and sequestration (CCS) is another very promising way to reduce carbon emissions. Unfortunately, we do not yet have the technology to make CCS even close to being economical. We also haven’t figured out how to handle the massive quantity of CO2 that would be captured. It is an area that deserves much more R&D than it’s getting.

All of this brings us to that very inconvenient truth: As much as I hate to admit it, nuclear power is the only technology we presently know of that can provide all the new electric power we’ll need as fast as we need it. In contrast to Germany, the countries such as France and Sweden that got rid of fossil-fuel electric generation did it with nuclear power. And they did it in less than 15 years. Nuclear can run 24/7 at nearly full power, so it doesn’t need fossil fuel backups. It is also an incredibly concentrated energy source, so generation capacity can be built out fast with much less land required. And despite my visceral fears of all things nuclear in the 1970s, data from decades of operation have shown that nuclear is the safest energy source available, producing less waste even than solar, producing power at a cost competitive or better than anything except some gas, and harming thousands of times fewer people in its worst accidents than coal does in daily operation. Gen IV nuclear promises to be even better. So although renewables will play an increasingly essential and important part on the path to deep decarbonization, without nuclear in the mix we will never be able to get there in time.

As we now understand the urgency of climate change, I and many others have moved from a position of deep distrust of nuclear power to an acknowledgement that in combination with wind, solar, and CCS, nuclear must play a serious role in climate solutions.