Measured as the amount of GW installed in 2040 that are produced by fusion reactors of either private or public companies or organizations located in one of these territories
1,000
3.00@NivlacM bold theory, presupposes that once fusion gets on the grid it will be cheaper than other sources of electricity in total levelized cost. extra spicy considering that deployed fusion power units by 2040 will likely be a lot closer to FOAK than NOAK.
c.f.: https://manifold.markets/SimoneRomeo/when-will-a-nuclear-fusion-reactor?tab=comments#4vi261mb80j
@pyrylium the key to bold predictions is knowing nothing about the topic at hand. Fusion in my head is just free energy. No cost, just put a Fusion label on a box and it can power a house
@SimoneRomeo are we counting theoretical installed capacity (regardless of realized power generation) or actual produced fusion MW for the grid in 2040?
@SimoneRomeo strictly speaking, the dispatched power from a power plant will always be less than the nameplate capacity (https://en.wikipedia.org/wiki/Nameplate_capacity) because the availability factor (https://en.wikipedia.org/wiki/Availability_factor) will always be less than 100%. There are many reasons why this may be the case -- nuclear fusion is not an intermittent power resource like solar or wind but equipment will always require some amount of downtime for maintenance and inspection. There may also be economic reasons for this -- if your plant is expensive to run (say, it runs on expensive helium-3) then it may only be profitable to operate during peak demand hours (https://en.wikipedia.org/wiki/Peaking_power_plant). A good overall metric for this is capacity factor (https://en.wikipedia.org/wiki/Capacity_factor), which assesses the actual energy deployed to the grid as a fraction of the theoretical maximum deployable energy based on the nameplate capacity.
To give an example: if China builds a 2 GW fusion power plant but only operates it at 1% capacity, and the US builds a 1 GW power plant but runs it at 100% capacity, who would you resolve in favor of? The word "installed" in your description implies nameplate capacity (i.e. China), but the word "produced" implies actual deployed power (i.e. the US).
@pyrylium thanks for the explanation! Which figure will be easier to define? If it's easier to assess the actual energy deployed, let's stick with this one. If we won't probably have clear data and we can only know the nameplate capacity, let's use this instead
@SimoneRomeo as far as ease of definition goes, nameplate capacity will almost certainly be widely publicized information where deployed power may or may not be publicly accessible (especially in countries like China). Just keep in mind that nameplate capacity is as much a business/political metric as it is an engineering metric -- on some level you are just accepting a power company's self-assessed capabilities at face value. This can get pretty controversial -- the concentrated solar power industry is (imho) a good example of a sector that looks great in nameplate potential on paper but falls apart if you dig into actual dispatched power. Just something to keep in mind from your neighborhood energy nerd (:
@SimoneRomeo it depends on the local policy & regulatory environment (e.g. for all its flaws, ERCOT is and was a leader in real-time power production data transparency: https://www.ercot.com/mktinfo/rtm -- this will almost certainly not be the case for Chinese power markets). Short of this, companies may voluntarily publicize this information, or it may be provided to shareholders. It might be possible to deduce the amount of fusion power on the grid from other data by negation, but this strikes me as unlikely and prone to error especially for smaller energy deployments.