The big challenge is that the most obvious renewable resources -- wind and solar -- have massive scalability issues because they are not dispatchable power resources. They don't generate power on an as-needed basis, they generate when the wind is blowing or the sun is shining. And because electricity supply must meet demand on an instantaneous basis, the inability to control when wind and solar resources generate power is a huge problem that makes it virtually impossible for them to meet 100% of power load requirements without massive overbuilding.
What we need, then, is a dispatchable resource that can lower our carbon emissions. Natural gas is a possibility -- kind of. It is much cleaner than coal, but still emits carbon. Roberts estimates that switching primarily to natural gas could get us to roughly 60% decarbonization. That sounds pretty good, even if the target we need to hit is actually 80% - 100%. But there's a big problem:
Natural gas is cleaner than coal (by roughly half, depending on how you measure methane leakage), but it’s still a fossil fuel. At least without CCS [Carbon Capture Sequestration], it is incompatible with decarbonization beyond 60 percent or so.
If you build out a bunch of natural gas plants to get to 60 percent, then you’re stuck shutting them down to get past 60 percent.
It would be very difficult to strand all those assets. There would be a lot of resistance. It’s just one example of path dependence in energy — choices, once made, tend to perpetuate themselves through inertia. Leaning too heavily into natural gas in the next 20 years will make it more difficult to pull away in the subsequent 20.Enter nuclear power, the new darling of (some) environmentalists. Nuclear power has a high capacity (it can generate a lot of power), zero-emission (no carbon), and dispatchable -- a holy trinity if your only goal is to decarbonize. It isn't renewable (though I don't think there's any immediate risk to our nuclear fuel reserves), and of course nuclear power has other risks and associations which make it politically controversial. But it strikes me as the most straight-forward, feasible, and immediately accessible method for taking big chunks out of our carbon footprint right now.
What are the alternatives? The best one is high-capacity energy storage (which can convert a variable resource like wind into a dispatchable one like nuclear). But the technology to have such storage on the scale and flexibility necessary is just not there yet, and while it's more than an eye-twinkle, it's also not particularly close at hand. After that, we could simply engage in massive, massive overbuilding of wind and solar. But even then we'd need to also basically globalize our transmission network (and massively upgrade that too), so that we could be confident that the wind is blowing/sun is shining somewhere.
Roberts indicates that there is a debate between those who think we can go 100% renewable (no nuclear, no CCS) versus those who want the latter on the table. Count me decisively in the latter camp. It might theoretically be possible to design a grid system available today that is zero-emission and entirely renewable. But the political, economic, and technological obstacles to putting it together are more than formidable, they are towering. Nuclear power is a technology we have now, that checks all of the key decarbonization boxes.