5 ms·

A big problem with solar and wind is that they aren't dispatchable. A 100 MWe nuclear power plant fitting in the space of small local power plant isn't equivale

by p_l 4y ago

A big problem with solar and wind is that they aren't dispatchable. A 100 MWe nuclear power plant fitting in the space of small local power plant isn't equivalent of a 100MW solar or wind plant - for my local area, the equivalent solar/wind plant is about a combination of few Gigawatts of nameplate capacity of solar and wind plants (allowing for very optimistic storage tech, it gets "down" to 1180 MW solar + 410 MW wind + 840 MWh battery at 170MW output + 50MW hydrogen turbine backed by 10186 MWh stored hydrogen. It's also 1.5x the cost of plonking EPR-based NPP.) Depending on smoothing from elsewhere is going to be also tricky - already considerable amounts of money are being paid out for curtailment because the grid can't take it.

entropicgravity 4y ago
The larger the geography of the grid (tied with HVDC) the more supply and demand average out. This minimizes the need for pumped storage (and its variations). To give just one example the UK is building an underwater HVDC directly between Morocco (solar) and the UK. So the known solutions are already available they just have to be implemented.
- p_l 4y ago
  Except those solutions tend to ignore the cost of failures and that even with interconnects balancing the grid is not trivial - just inside european grid the problems are considerable, and there's still the issue of interconnects failing. And even then, you now need to overbuild not just for replacement of local plants, but also for replacement of plants on the other end of the continent. Great if you're selling gas or gas power plants, I guess, not so great if you want zero CO2 emissions especially with how powerhungry the replacement techs can be.
pfdietz 4y ago
Nuclear isn't dispatchable. If it's not always running the cost/kWh escalates dramatically. Solar and wind with storage are somewhat dispatchable, because you can decide when to discharge the storage systems.
- p_l 4y ago
  It's much more dispatchable than renewables, it's just cheaper to run it always at full capacity. Solar/wind/storage combination requires overbuilding in range of 15-25 times the nameplate capacity. Personally I'm of the opinion that we should have nuclear+renewables with less focus on storage and more on opportunistic production of, for example, green hydrogen - not for grid storage but for all other uses like steel production, off-grid power systems (including cars/trains that have needs beyond battery capability) etc. Current approach with renewables always being graded on their lowest possible price point and silently ignoring the growing gas generation required to smooth them and larger and larger grid instability is not great.
  - pfdietz 4y ago
    Ah, so solar/wind is not dispatchable because it's more costly you claim, not because one can't actually do it? You don't need 15-25x overbuilding. That is nonsense. Perhaps you're assuming no storage whatsoever, and are oversizing the renewables to deliver sufficient power instantaneously? What terrible and foolish engineering. The round trip efficiency of power -> hydrogen -> power is maybe 40%, so if you sent the entire renewable output through hydrogen the overbuilding (in the sense of how much energy you'd have to produce / how much energy was delivered to the grid) would be 2.5. And of course you'd send only a bit of energy through hydrogen; most would be delivered either directly to the grid or through diurnal storage at higher round trip efficiency. If I go to https://model.energy/ https://model.energy/, click on the US, and solve for 2030 cost assumptions and 2011 historical weather data, I find the renewable + storage system to provide a synthetic baseload output to the US involves 0.1% solar curtailment, 23.8% wind curtailment, and delivers nearly 3/4 of the renewable power directly to the grid. If I narrow it down to just, let's say, Texas, the wind curtailment goes up to 32.1%. All this is a far, far cry from 15-25x overbuilding.
    - p_l 4y ago
      You use Texas (or whole of USA). Which is considerably more southern than most of Europe, and definitely more southern than Poland. IIRC Texas is also much more flat, which benefits availability of wind power. The 15-25x numbers I get from model.energy, in fact - lowest I get is 15x assuming way too optimistic ideas about availability of salt caverns in Poland, 25x is with mainly battery power, the numbers for no storage are too hilarious too discuss. Are there luckier locations that can get better numbers? Sure. I'm not against building renewables and storage, far from it. I am, however, against building new fossil fuel plants, including gas backup for renewables, and would rather we take the minimum average power use, fill it with nuclear, then get renewable/storage Virtual Power Plants to fill in the peaks (with some load following from reactors if necessary). If necessary, we can find new power sinks to make it more economical, in fact I'd love if we had large scale green hydrogen and synthetic fuel production backed by hydrogen electrolysers taking in overproduction. The real problem isn't cost of nuclear, it's that we still leave profit as the main guiding principle (to the point that building new solar might crash in some states) and not optimizing for 0 emissions.

by p_l 4y ago

entropicgravity 4y ago

The larger the geography of the grid (tied with HVDC) the more supply and demand average out. This minimizes the need for pumped storage (and its variations). To give just one example the UK is building an underwater HVDC directly between Morocco (solar) and the UK. So the known solutions are already available they just have to be implemented.

p_l 4y ago
Except those solutions tend to ignore the cost of failures and that even with interconnects balancing the grid is not trivial - just inside european grid the problems are considerable, and there's still the issue of interconnects failing. And even then, you now need to overbuild not just for replacement of local plants, but also for replacement of plants on the other end of the continent. Great if you're selling gas or gas power plants, I guess, not so great if you want zero CO2 emissions especially with how powerhungry the replacement techs can be.

pfdietz 4y ago

Nuclear isn't dispatchable. If it's not always running the cost/kWh escalates dramatically. Solar and wind with storage are somewhat dispatchable, because you can decide when to discharge the storage systems.

p_l 4y ago
It's much more dispatchable than renewables, it's just cheaper to run it always at full capacity. Solar/wind/storage combination requires overbuilding in range of 15-25 times the nameplate capacity. Personally I'm of the opinion that we should have nuclear+renewables with less focus on storage and more on opportunistic production of, for example, green hydrogen - not for grid storage but for all other uses like steel production, off-grid power systems (including cars/trains that have needs beyond battery capability) etc. Current approach with renewables always being graded on their lowest possible price point and silently ignoring the growing gas generation required to smooth them and larger and larger grid instability is not great.
- pfdietz 4y ago
  Ah, so solar/wind is not dispatchable because it's more costly you claim, not because one can't actually do it? You don't need 15-25x overbuilding. That is nonsense. Perhaps you're assuming no storage whatsoever, and are oversizing the renewables to deliver sufficient power instantaneously? What terrible and foolish engineering. The round trip efficiency of power -> hydrogen -> power is maybe 40%, so if you sent the entire renewable output through hydrogen the overbuilding (in the sense of how much energy you'd have to produce / how much energy was delivered to the grid) would be 2.5. And of course you'd send only a bit of energy through hydrogen; most would be delivered either directly to the grid or through diurnal storage at higher round trip efficiency. If I go to https://model.energy/ https://model.energy/, click on the US, and solve for 2030 cost assumptions and 2011 historical weather data, I find the renewable + storage system to provide a synthetic baseload output to the US involves 0.1% solar curtailment, 23.8% wind curtailment, and delivers nearly 3/4 of the renewable power directly to the grid. If I narrow it down to just, let's say, Texas, the wind curtailment goes up to 32.1%. All this is a far, far cry from 15-25x overbuilding.
  - p_l 4y ago
    You use Texas (or whole of USA). Which is considerably more southern than most of Europe, and definitely more southern than Poland. IIRC Texas is also much more flat, which benefits availability of wind power. The 15-25x numbers I get from model.energy, in fact - lowest I get is 15x assuming way too optimistic ideas about availability of salt caverns in Poland, 25x is with mainly battery power, the numbers for no storage are too hilarious too discuss. Are there luckier locations that can get better numbers? Sure. I'm not against building renewables and storage, far from it. I am, however, against building new fossil fuel plants, including gas backup for renewables, and would rather we take the minimum average power use, fill it with nuclear, then get renewable/storage Virtual Power Plants to fill in the peaks (with some load following from reactors if necessary). If necessary, we can find new power sinks to make it more economical, in fact I'd love if we had large scale green hydrogen and synthetic fuel production backed by hydrogen electrolysers taking in overproduction. The real problem isn't cost of nuclear, it's that we still leave profit as the main guiding principle (to the point that building new solar might crash in some states) and not optimizing for 0 emissions.