Comment by AnthonyMouse | Hacker News Reader

Comment by AnthonyMouse | Hacker News Reader

Parent Full thread

> Furthermore, nuclear plants take 5 to 10 years to even build.

But that's the problem.

The cost of solar and batteries is declining rapidly. For how long? We don't know. Exponential curves eventually flatten out. We had Dennard scaling until we didn't.

So the question is, do they get cheap enough before they hit the wall?

We have two choices. Choice one, build new nuclear plants, and then if alternatives are even cheaper, we pay the current/historical price for electricity for a while instead of some lower price (or the investors in the nuclear plants lose money, take your pick). This also gives us new nuclear plants to use to destroy all the stupid plutonium created by old nuclear plants, even if it costs more, and we really do have to get rid of that stuff.

Choice two, we don't, and bet everything that the alternatives get cheap enough before we hit the wall. In which case we're completely screwed if we bet wrong, because it takes a long time to build new nuclear plants, so if we don't start now by then it'll be too late.

They are already cheap enough. Renewables are much cheaper than nuclear and competitive with fossil fuels. We don't need another factor 10 cost improvement. All we need is political will to implement the change. We could be running 70% renewable in a couple of years if we wanted to. To get to 100% we might need to improve the cost effectiveness of power-to-gas facilities, but there is zero reason today not to replace the majority of our energy consumption with renewables.

They're cheap enough when they're not being used as baseload. To get to 100% renewable you would need to pay for storage and then have enough over-supply to deal with extended periods of overcast. That's a lot more expensive.

This is another reason to do both. Use nuclear for baseload and you remove all the storage you would need for solar to work overnight. Meanwhile demand is higher during the day when solar is generating, so still use that there. It reduces the amount of storage you need from the entire night to just the high demand period between sunset and when people go to bed.

> All we need is political will to implement the change.

You only need political will for subsidies.[1] You only need subsidies if it's not actually cheaper.

[1] Or regulatory approval, but let's not bind everything in red tape all around, shall we?

If it is cheaper then people just do it on their own without any government involvement.

Using nuclear for baseload doesn't work economically. Nuclear plant economics are such that you want to run them 100% all the time. Renewables sometimes produce pretty close to 0 energy. If you don't want to invest into storage you need nuclear capacity to cover close to all demand. If you have that capacity you don't need renewables. I don't think we can build nuclear plants quickly enough to prevent catastrophic warming.

If you instead use renewables, batteries, and power-to-gas you can also reach net-zero emissions without waiting twenty years until you have enough nuclear plants. Renewables and batteries are already cheap enough to compete with fossil fuels, and provide enough stability to reach a high percentage of the energy demand. 50% renewables is doable with hardly any storage at all. Power-to-gas can provide the storage needed for seasonal variations, but there are still cost problems with that.

Imo the lack of political will for the energy transformation mostly takes the form of unpriced externalities for fossil fuels. There still is no carbon price that is even close to the actual damages. Damages from the extraction of fossil fuels are not paid for by the companies either. Coal plants still emit tons of Mercury without paying for it. Aviation fuel is essentially untaxed. The list goes on and on.

> Using nuclear for baseload doesn't work economically. Nuclear plant economics are such that you want to run them 100% all the time.

That's what baseload is.

> Renewables sometimes produce pretty close to 0 energy.

Which is why they need storage, and storage is expensive. If you do half nuclear and half renewables, you only need half as much storage.

Less than half, because you can use pricing to shift demand.

Suppose you have 50 GW of nominal demand, but solar generation is at 10% of normal because it's overcast. You can get demand down to 40GW through price incentives. If you had 50 GW of solar which is down to 5 GW, you would need to cover 35 GW from storage. If you had 25 GW of solar which is down to 2.5 GW, and 25 GW of nuclear, you only have to cover 12.5 GW from storage. Less than half as much.

Baseload tends to be about 40% of max load in most grids. i.e. having base production doesn't really change the economics of storage or over-provisioning etc. On the other hand because there is always, in this theoretical grid 60% overproduction potential your base load has real difficulty making money about 1/3rd of the time.

Baseload, is really minimal load in the grid. There are many ways to produce the minimal amount. Real difficult part is to provide the max load demand on the grid.

Basically current nuclear technology is inflexible in economic terms (also quite a bit in practical terms), shown historically by most grid storage being deployed when nuclear was being built out.

Real engineering reviewed a paper that claimed we have cheap enough renewables to do ~80% of electricity from them right now. For the rest 20% we would need massive price drops on the battery side but just that 80% alone would cause a pretty massive decrease in price due to economies of scale.

So, diversify and do both?

Exactly.

And for extra points, this phases out oil and coal faster. 1TWh of solar replaces 1TWh of coal, but 1TWh of solar and 1TWh of nuclear replaces 2TWh of coal.

Reply on news.ycombinator.com