Hacker News new | ask | show | jobs
by mackwic 3930 days ago
I suggest you to learn a bit about power grid engineering. Basically, the fact that we can't store electricity means that when you power on your toaster, somewhere a power plant has to increase its power to compensate.

But the nuclear power plant takes up to 2 days to startup. It's not sustainable for the daily life where consumption change every hour.

That's why there is the need for more agile power plant, where fuel, gas, and coal are suitable. Gas being the most agile but the less efficient, fuel and coal being used as the middle between gas and nuclear, depending of the price of each resource.

This is an engineering constraint before anything else.

5 comments

I know that nuclear power is base-load power, not peaking power. Infrastructure is something of a passing interest of mine. :)

Last I checked, the majority of the US's power comes from coal-fired power plants. If there was the political will and public understanding, we could likely replace pretty much all of that with nuclear plants, reserving coal-fired, or oil-fired (or pumped-storage hydro, or flywheel storage, or...) plants to meet peak power demands.

Well, I have to say I don't know the US infrastructure enough to have a bold opinion, but I know heavy infrastructure like these often have a sound design respectively to their constraints (which can be absurd).

The European power grid is very tight and well-meshed, this is why we can let our nuclear power plant and sell electricity to Switzerland, Germany or Netherlands or buy it.

I guess it's not as well interconnected in the US, so that states can't easily lend power which disable the use of nuclear power plant for many states ?

I think that the US either has a national power grid, or has one for the West and East halves of the country. Obviously, throwing power across the whole damn continent is far less efficient than just throwing it across a state.

I've -very recently- heard people making the claim that Texas has its own grid, but that's news to me. Maybe the situation in Texas is that they have all the power gen capacity that they need within the state, and are also hooked into the national grid. shrug

> I know heavy infrastructure like these often have a sound design respectively to their constraints (which can be absurd).

Anti-nuke hysteria is toxic to sound planning and design. :(

There are three in the US.

East, West and Texas.

https://sites.google.com/site/theuspowergrid/

> I think that the US either has a national power grid, or has one for the West and East halves of the country

It would makes sense.

> Obviously, throwing power across the whole damn continent is far less efficient than just throwing it across a state.

For now. With the future high-temperature supra-conductive cabling we will be able to do wonders of efficiency on large distance.

That can be fixed with pricing. Charge more during peak, less during surplus. France relies on nuclear for about 80% of its power. Power is cheaper at night than during the day. As a result, almost all appliances like dishwashers, washing machines, dryers and water heaters come with timers that allow you to use the cheaper power.
The surge pricing is true in the UK too:

https://en.m.wikipedia.org/wiki/Economy_7

Who are you convinced we can't store electricity? Pumped storage for storing potential energy has worked long before we even had electricity.

Sure, the whole process of storing electricity is lossy. Thermodynamics guarantees it. But the point isn't to try and store enough generation capacity to meet our demand. You just need enough storage capacity to deal with the peaks.

There is of course the argument of "pumped storage isn't practical for meeting all of our peaking demands". Of course it isn't. Neither is nuclear suitable for meeting all of our baseline demands. They are both just tools available.

It's not that we can't store electricity - we just can't store enough of it to turn renewables into usable tool for meeting baseline demand. We may be able to do that in the future, with battery technology improving and some clever shenanigans like using electric vehicles as grid storage - but we need something now, and there's no other alternative for doing it green than nuclear power.
> ...we just can't store enough of it to turn renewables into usable tool for meeting baseline demand...

/me puts on pedant hat and mad engineer jacket:

Are you sure about this? The US is huge and has an enormous amount of uninhabited space. If we covered -say- the north-western quarter of Nevada in the best batteries available today, would that cover base and peaking power during slack production time for -say- the surrounding states?

/me removes pedant hat

(Do bear in mind that I feel both that anti-nuke hysteria is hugely damaging to the planning and deployment of new nuclear power plants in the US, and that it's fairly clear that of the currently available non-hydroelectric base-load generation tech, nuclear power is the only good option. [Though, solar power beamed down from orbit is a really intriguing idea.])

I was never arguing that we would meet baseline demand with "renewables" stored in some sort of mechanism.

But if you're really convinced this problem is unsolvable, lets just do this: we'll generate enough power (with fission, or whatever you want) to meet peak demand all the time. As long as we're near the coast there is a convenient way to get rid of all that surplus energy: desalination plants. I can think of a few areas that would jump at this idea, like California. Even if you wind up with too much freshwater, just start dumping it back into the ocean. The salt is generally worthless and is dumped back into the ocean. This will actually help offset the fact that the oceans are desalinating due to the ice caps melting.

Electricity prices will go up as a result of this, but you'll be subsidizing the fact that the grid is now greener as a result.

Nuclear can do load following pretty well (as does coal) as long as there plant supports it - reasonably recent ones are required to by regulations. It is used in that mode commonly eg in France and Germany. The economics are another thing of course, you don't save much in operating costs by throttling output.

Stopping to 0% is not needed in normal load variation, thanks to toaster usage patterns.

nuclear could replace the entire base load. If you factor in CO2 emission externality costs, it might even be cheaper to run nuclear above baseload. Build solar to help with peak usage and rely on gas for peaking as well.

The surplus could be used to pump water up damns or charge batteries. Or even to just pull co2 out of the air to compensate for car emissions.

It could provide very cheap night power for electric car owners. Worst case we could just waste the energy.

Once you accept that co2 emissions are an existential threat to the world, small inefficiencies are a small cost.