Hacker News Reader

Comment by henrikschroder | Hacker News Reader

Nuclear power plants can't replace oil- or coal-powered plants, because nuclear plants can't adjust output over the day to fit actual usage. Nuclear power plants are great for providing the baseline power production, but you need something adjustable to handle the morning and evening peaks, like oil or coal or wind or hydro or really big batteries.

So no, you can't do the math for how many nuclear plants you would need to get off oil.

View on HN

France is using some of their nuclear plants to provide peak load, and they adjust the power output of most of their plants over the day to better fit the actual usage. Their reactors are not ideal for this; there are better ways to design nuclear plants for load-following. But it can certainly be done.

Alternately, you could make enough nuclear capacity to provide your peak power levels all the time, and dump excess energy into some energy-hungry industrial process, like aluminum smelting, or synthetic fuel production, or ammonia synthesis.

So yes, you can do the math. It's just that there are a bunch of different ways to do that math, and most of them involve a lot of cleverness, large budgets, careful engineering, and a fair dose of uncertainty.

Alternately, you could make enough nuclear capacity to provide your peak power levels all the time, and dump excess energy into some energy-hungry industrial process, like aluminum smelting, or synthetic fuel production, or ammonia synthesis

Heck, you don't have to do anything with it, just disconnect your turbines from your generators.

Technically, sure. Economically, since most of the cost of nuclear power is the capital cost of building the plants, you want to be running them as close to 100% capacity as you can, all the time. If you can smelt some aluminum while you're at it, that really helps.

Good idea on the face but the energy needs to be dumped somewhere and dumping that much excess energy is tough and costly. Used to work in the industry, if the turbines tripped or anything else happened along the way, you SCRAM immediately as a safety precaution.

And they buy massive amounts of coal produced electric energy from germany during peak load times in winter.

Sure they can. Hybrids that charge at night, that get cheaper rates to charge when the utility sends a signal, pumped storage, only running hydroelectric during the day, running aluminum smelters at night, thermal storage, or worst case, just not using the extra electricity. There is stuff to build, but we're not going to build all of those nuke plants overnight either.

Yes you can.

Liquid-Fluoride Thorium Reactors can be started and stopped without incident, in a few hours. I believe their output can also be adjusted in operation. They're safe and efficient enough to build one into a 40-foot trailer that can be started, stopped, and moved on a moment's notice.

My favorite idea for this sort of thing is to put a LFTR on a barge (or, better, in a submarine), using seawater for cooling, and ship them out to any coastal area where they're needed. Build them in factories, assemble them in shipyards where such things are relatively easy, and handle the periodic maintenance and refueling in a centralized location.

Here's where it gets really fun: you can use the waste heat for desalination. There are a lot of places that sit next to a coastline but are short on electricity and fresh water. Russia is trying to make a lot of money from this, using a variant of the reactors they've been using on their nuclear ice-breaker ships.

And if you're using something like a LFTR that's compact enough to fit in a submarine, the water will protect you from hurricanes, earthquakes, aerial attack, and so on. It's a pretty clever scheme. Here's a thread on the Energy From Thorium forums discussing it:

http://energyfromthorium.com/forum/viewtopic.php?f=43&t=...

Can we not solve this problem by storing energy in the longer term?

It's true some 'burst' capacity is important, but can you not get a lot closer to 'average load' than the baseline by using the energy? Pumping water into a reservoir at night, and getting the power back with a small hydro plant during the day comes to mind. We smooth out an even worse problem with wind and solar, which is spiky AND out of sync with demand sometimes. I believe flywheels are used to store energy in the short term sometimes. You did mention 'really big batteries', implying that we would need too many, but we aren't limited to chemical forms of energy storage.

I think this is a solved problem, or at least a fairly solvable problem, and nuke plants can provide not only the 'baseline' but close to the bulk average needs for power. We can decouple the power generation and the need for "something adjustable" and benefit.

Here's the electricity demand curve for the UK for the last 8 days:

http://www.nationalgrid.com/uk/Electricity/Data/Realtime/Dem...

Electricity demand during the night is about half of what it is during the day. If you were to have nuclear baseline production that covers peak consumption, you would have to find a way to store about 100000MWh of energy that gets produced during the night and then distribute that during the following day, or you would just waste it, somewhere, somehow.

(Disclaimer: Back-of-the-envelope math, I'm assuming the scale on that graph is in MW, it's only for the UK, not the US)

Yes, it's not an impossible problem. Yes, you can even out the peaks by changing behaviour. Yes, you could perhaps figure out a way to store and release that much energy each day.

But my point still stands, replacing fossil-based electricity production with nuclear is not trivial, it's a lot more complicated than summing up the total output of all fossil-based plants and dividing it by the average output of a new nuclear plant as DanielBMarkham argues.

I'm also a big fan of nuclear. Many countries would probably benefit from having their baseline production completely covered by nuclear plants. But how to deal with peak demand of electricity without fossil-based production and in an economically viable way, that is not a solved problem yet.

One of the reasons demand is light at night is that we run our central heating systems on gas -- mostly imported from Norway or parts east, these days. Per kWh, gas heating is a lot cheaper for the consumer. But if we had a serious base load nuclear infrastructure, we could provide much cheaper off-peak electricity for domestic heating, which would smooth out the demand curve significantly.

You're thinking of solid-fuel uranium reactors.

Actually, thorium reactors have a natural load-following capability due to their negative thermal coefficient of reactivity. You suck power out of them too fast, they cool, get denser, and the reaction rate increases, increasing power output.

In addition, shutting down a liquid salt reactor is literally a matter of pulling, actually melting, a plug. Simply overheat the core, melting the freeze plug, and the fuel falls into a subcritically shaped catchment tray. Reheat and reload and the reactor is up again. They're very easy to start and stop.

There's a wealth of information on the Net about these guys. I hope they get built!

Reply on news.ycombinator.com