Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

conventional nuclear isn't dispatchable

ramp times are measured not in hours but in days



Completely false

Modern LWR reactors ramp at around 5% per minute.

France & Germany use them for Load following https://en.wikipedia.org/wiki/Load-following_power_plant#Nuc...

Conventional nuclear is vastly more dispatchable than other fuel types


Physically they can ramp down to 50% but generation is pretty much a sunk cost at that point so it makes the bad economics even worse. That would mean that you are paying of the order of $240/MWh instead of $120/MWh.

For comparison solar/wind are about $30-40. Levelizing with pumped storage pushes that up to $60-$70.

IIRC France and Germany almost exclusively use gas for load following. Even at current prices it's vastly cheaper than using a NPP to do it.


When capex costs more than other technologies plus storage, and it reduces your neutron economy and thus the life of your fuel it's just curtailment with extra steps.


Incorrect. NuScale appears to be designed to be dispatchable - for example here is a simulated load-following curve during a day: https://www.researchgate.net/figure/Example-of-NuScale-modul...

From: https://www.researchgate.net/publication/295114246_Integrati...

  The NuScale plant incorporates unique features that enhance its ability to load follow, either due to changes in electricity demand or variable generation by renewable sources on the grid. This is accomplished through a combination of the small unit capacity of a NuScale module (50 MWe gross) and a multi-module approach to the plant design. This design strategy provides a uniquely scalable plant and gives the plant owner considerable flexibility in both the build-out of the plant and also its operation, including for load-following. The key power management options of the NuScale plant for load-following operations, designated NuFollow™, include the following:
  • Taking one or more modules offline for extended periods of low grid demand or sustained wind output,
  • Maneuvering reactor power for one or more modules during intermediate periods to compensate for hourly changes in demand or wind generation, or
  • Bypassing the module’s steam turbine directly to the condenser for rapid responses to load or wind generation variations.
One problem is that the US regulatory agency doesn’t like load-following - apparently only one nuclear reactor in the states does it. It is common in European countries to do some load-following on time periods of hours to 2 days (although like all generation technologies, there are limitations and constraints). A section about France on the topic: https://www.world-nuclear.org/information-library/country-pr...


If you're paying $80/MWh for captial and fixed O&M and $20/MWh for fuel and variable O&M, then you're still payiny 80% for the energy you don't produce.

At that point it's effectively just curtailment.


And with wind and solar, you need storage, which costs a lot and similarly is is not used most of the time.

That said, in New Zealand the story is different, because we have some very large battery banks called hydropower lakes. However when our batteries run dry, the country has a bad time.

Currently storage is over 3TWh[1] and in 2020 NZ hydro generated 24TWh[2].

[1] https://www.energylink.co.nz/publications/hydro-watch

[2] https://wikipedia.org/wiki/Electricity_sector_in_New_Zealand


Storage is needed every night with solar.


That was the parents opening sentence.

Their point is interesting. If you already have hydro and add solar or wind, can you pump back into the lakes during periods of excess, and use the hydro at night?


Yeah this is a thing. If you have enough hydro you can just turn it off entirely when the renewables are going and have a 50/50 mix.

If not, it's called pumped hydro storage (or most precisely it's blie field on river pumped hydro if it's already dammed and on a river). You need to trap the water somewhere so you don't have to pump it too far. This involves building a lower reservoir and adding pumps or modifying the turbines to be two way.


The opening sentence was “is is not used most of the time”


thank you for the update, this does seem to be a real difference between nuscale and conventional nuclear


Not sure if nuscale is wanting to do it in the short term, but many of the SMR concepts include a thermal storage component.

This helps build industry experience with molten salts and would allow it to be actually dispatchable (rather than paying for energy in the form of capex and fixed O&M and then just not producing it)




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: