Comment by lven | Hacker News Reader

Comment by lven | Hacker News Reader

Parent Full thread

Not exactly. Nuclear reactors have a difficult time following the load because of Xenon poisoning. Xenon generated during the fission reactions absorbs neutrons that could have been used for fissions. Luckily, it decays away over time. If you turn down the reactor power, you have to wait hours or day for for Xenon buildup to decay so that you can turn the reactor back on. Some reactors manage to load follow more easily by adding lots of excess reactivity (more potent control rods) which is less safe overall. Smaller reactors will have the exact same issue. The amount of Xenon poisoning is proportional to the power density. NuScale reactors run at even higher power density than normal light water reactors, so they will have even worse Xenon poisoning. They won't be load following. One exception where this isn't true is micro gas-cooled reactors that have so low a power density that they have negligible xenon poisoning and can follow loads easily if necessary. Even then, it's not a great idea because of thermal cycling issues.

Some reactors can also more effectively let xenon escape. That is one of the reasons why Alvin Weinberg wanted a liquid molten salt reactor, you can let the xenon bubble out like CO2 in a soft drink.

And if you do that, then the 137Cs ends up in the off gas storage rather than in the molten salt. A good fraction of the fission products escape the salt, which kind of shoots down one of the main selling points of MSRs.

https://gain.inl.gov/SiteAssets/MoltenSaltReactor/Module2-Ov... (see slide 23)

Nuclear can modulate it's output by more aggressively cooling the water. There's no xenon poisoning since the reactors output is the same. This is undesirable because it's essentially wasting fuel, by deliberately reducing the efficiency of the steam turbine. But it can be done, and fuel is not a big driver of nuclear cost.

Modulating nukes' output make each kWh even more expensive, when they are already not competitive.

Most of a nuclear plants' cost is in construction. It doesn't really affect operating costs. Furthermore nuclear is the most competitive decarbonized energy source, because wind and solar rely on fossil fuels.

During the ten years it takes to build a nuke, 100% of the power it is hoped to displace is supplied by fossil fuel. The thousands of tons of concrete are produced by cooking limestone with, again, fossil fuel. And the thousands of tons of steel are refined and smelted with, again, fossil fuel.

Most of a nuke plants' cost is in construction, but that is not because its operating cost is low. It is just insanely expensive to build. Then, its operating cost is high. Each moment it is not producing at 100% rated power, its per-kWh cost increases accordingly. Its operating cost does not decrease proportionally when it operates at below rated maximum power, so the operating cost per kWh is multiplied by the difference. And, operating at below rated capacity, the construction cost is amortized over fewer kWh, again making the per kWh cost greater.

You already well understood all of the above, but evidently hope readers will not.

> During the ten years it takes to build a nuke, 100% of the power it is hoped to displace is supplied by fossil fuel.

By comparison gas turbines will always emit carbon dioxide, and there's no realistic plan to run a solar and wind grid without fossil fuel backing. No, there is no realistic plan to store electricity despite your incessant insistence to the contrary.

> The thousands of tons of concrete are produced by cooking limestone with, again, fossil fuel. And the thousands of tons of steel are refined and smelted with, again, fossil fuel.

Both of these can be replaced with thermochemical processes powered by nuclear power.

> Most of a nuke plants' cost is in construction, but that is not because its operating cost is low. It is just insanely expensive to build. Then, its operating cost is high.

Incorrect, nuclear power is quite cheap once the plants are constructed.

No cement or steel process, over the 60+ years nukes have operated, has been switched over to nuke-powered. There is no reason to think any will be.

Nuclear power operating cost is about commensurate with fossil fuels, which are not competitive. Operating at 50% rated power makes each kWh, marginally, twice as costly. Operating at 50% rated power long term makes each kWh absolutely twice as costly.

At the time when their power cannot be sold at any price sufficient to continue operating, nukes not propped up by tax coercion will be mothballed. Their huge construction cost will end up amortized over only the kWh produced up to that time. So, the finally recognized cost per kWh will balloon to many times over what was promised at construction time.

> No cement or steel process, over the 60+ years nukes have operated, has been switched over to nuke-powered. There is no reason to think any will be.

Isn't this also a barrier to building hydroelectric facilities? They're basically big dams.

Also you believe we'll be able to create massive electrolysis plants to create energy storage for solar and wind? Interesting how you're so confident in massive changes to industrial processes when solar and wind require them, but totally dismissive when other solutions do.

Regardless, metallurgy and cement just need a source of heat and unlike solar and wind which need to convert electricity to heat nuclear plants produce heat directly.

Nukes do not produce output of sufficiently high temperature for cement or steel production.

Hydro-power dams are expensive to build, too, but operating cost is extremely low. New ones will not be competitive with wind & solar, but existing dams will remain useful, where not demolished for ecological or fisheries reasons.

There are no technical impediments to electrolysis. It all just needs to be built out. Efficiency is rising very fast.

Nuclear power does produce enough heat to drive thermochemical water splitting, which can produce hydrogen at better scale than electrolysis (since it avoids issues around electrodes corroding). That hydrogen can in turn be used for metallurgy.

You are correct, most of the cost is in construction. Buildings have finite lifetimes, so each second it doesn't run at 100% capacity you make the average kWh more expensive.

they are when you price in the cost of the destruction of human civilization by climate change if we don't do something soon.

I may be missing something important here, but I'm not a nuclear fission SME. Please do correct me if I'm not making sense.

Doesn't the modularity (multiple 60MW reactors in a single installation) in the NuScale design obviate the "Xenon poisoning" issue, since shutting down one or more reactors doesn't mean halting power generation as it would with a single, larger reactor?

Presumably the reactors can be shut down and powered up independently so addressing the "Xenon poisoning" issue should be just a matter shutting down, then powering up some fraction of the reactors, scheduled to maintain the base load required, no?

I mean that loads can be shared geographically and the reactors can maintain relatively consistent output.

Reply on news.ycombinator.com