4 ms·

It's a really, really bad idea to extrapolate price reductions in solar/wind into carbon free grids. Unfortunately there's a ridiculous wall of increasing overp

It's a really, really bad idea to extrapolate price reductions in solar/wind into carbon free grids. Unfortunately there's a ridiculous wall of increasing overprovision required as you go deeper and deeper into renewables. Have a look at [0] if you can make the paper fall from the back of a truck, especially at Fig. 3. Assuming absolutely perfect grid able to transmit arbitrary amount of electricity and disregarding any real constraints, going purely by geophysical constraints on the amount of wind and sun available, continental US would need at least x7 as much wind+solar capacity as it does consume on average to have no more than one day of blackouts per year, assuming 12hrs of full consumption stored and immediately dispatchable. The only scenario in which that number goes to about x1 requires 32 days of full storage, which is just impossible with our current tech. And as you can imagine, in any real scenario that number would need to be several times higher. It's an absolutely ridiculous cost barrier that is not talked about enough and that effectively guarantees that we will be burning fossils for the foreseeable future. [0]: https://pubs.rsc.org/en/content/articlelanding/2018/ee/c7ee03029k https://pubs.rsc.org/en/content/articlelanding/2018/ee/c7ee0...

liketochill 5y ago
Not sure why you are downvoted you have a valid point with regards to over provisioning and storage. I wonder how much dispatchable generation has to be kept around to only have two 12 hour blackouts per year due to lack of generation.
- bryanlarsen 5y ago
  3 hours, some overbuilding and a good grid to get to 99.99% reliability (under an hour of blackouts). https://www.nature.com/articles/s41467-021-26355-z https://www.nature.com/articles/s41467-021-26355-z
  - lambdadmitry 5y ago
    I don't see where that number comes from. Also, "good grid" means "perfect national transmission", which is a wildly optimistic assumption. The US is currently split into multiple grids, with ties being able to handle less than 1% of their generation capacity. Moreover, that's just the US. Here is what the link has to say about countries less blessed with landmass and latitude: > Indeed, in smaller countries, substantial gaps (>30% of demand for >20 h per year; pale orange curves in Fig. 4) remain in systems even with 12 h of energy storage and annual generation that is 3x annual demand. Or this, directly referring to extensive overbuilding required to meet demand: > For instance, historical solar and wind resources data in Germany reveal that there were nearly 2 weeks in which dispatchable generation had to cover practically all of the demand because of a period with very low solar and wind power availability (called “dark doldrums”)27. Although with vast enough wind and solar capacity it might still be possible to meet demand in all hours, the required capacity increases exponentially after a point that depends on the renewable resources of that country, and it is this geophysically-dependent point that will largely determine the cost-effectiveness of highly-reliable, renewables-based electricity systems.
Retric 5y ago
Those numbers are wildly off because among other things the use US generates 6% hydroelectric, and they didn’t try to minimize a cost function via location selection etc. Further large hydroelectric dams store months worth of their average output. Also, demand and generation are positively correlated as people use less power on cloudy days and less AC in the winter. You can look at how much oversupply and storage people need to live off grid and realize that’s a worst case situation isn’t 7x oversupply or 30 days of storage it’s roughly 3x oversupply of solar and 3 days of storage. At grid scale that’s roughly 6c/kWh for solar plus batteries which last longer the less their used though that’s also a high when you add other energy sources and grid transmission. The good news is markets are great at minimizing cost functions. In the near term lots of natural gas makes up the difference, but the cost breakdown is such that offsetting 1kWh of natural gas works out even if 50% of solar generation is wasted. Aka 2x oversupply of solar shows up from market forces it’s that cheap. Bring on storage and things start to look very solar heavy even with zero subsidy or price reductions.
- lambdadmitry 5y ago
  Hydroelectric can't ramp up to a large fraction of national demand, for multiple unsurmountable reasons (like, you can't dump the water fast enough for one). Location selection doesn't matter that much as they don't try to optimise cost, they just look at aggregate output. People living off grid still continue to consume the products of grid-enabled industries, from manufacturing to farming to logistics to infrastructure. Heating/cooling and lighting a cabin is much more trivial than paving a road coming to that cabin, feeding its inhabitants, building solar panels for them, or ensuring dense enough population to make research and manufacturing viable. In fact, we don't need to extrapolate minor part of personal consumption to the whole society, we have aggregate energy consumption numbers. "Market forces" can't generate electricity on their own. Instead of taking the bottom-up view, fraught with wishful thinking and unstated assumptions, why not take the top down, working from the potential generation capacity under ideal assumptions? That's exactly what the paper I linked did, and the results are… not great for 100% renewable
  - Retric 5y ago
bryanlarsen 5y ago
The book I linked includes a realistic model incorporating existing hydro, significant rooftop solar, et cetera. It requires a 20% overbuild. https://mitpress.mit.edu/books/electrify https://mitpress.mit.edu/books/electrify
- lambdadmitry 5y ago
  Hydroelectricity is very limited in its maximum instantaneous output, you can't feed half a grid with it, there is just nowhere to put all that water to without massive flooding, so it's a bit of a moot point. Rooftop solar doesn't matter for the top down approach in that paper, they don't make any assumptions about PV location and assume perfect grid, so rooftop PV won't make it any better. So seeing a top down, intentionally optimistic approach arriving at a number much higher than that book's author's, I am very sceptical of the book's results.

liketochill 5y ago

Not sure why you are downvoted you have a valid point with regards to over provisioning and storage. I wonder how much dispatchable generation has to be kept around to only have two 12 hour blackouts per year due to lack of generation.

bryanlarsen 5y ago
3 hours, some overbuilding and a good grid to get to 99.99% reliability (under an hour of blackouts). https://www.nature.com/articles/s41467-021-26355-z https://www.nature.com/articles/s41467-021-26355-z
- lambdadmitry 5y ago
  I don't see where that number comes from. Also, "good grid" means "perfect national transmission", which is a wildly optimistic assumption. The US is currently split into multiple grids, with ties being able to handle less than 1% of their generation capacity. Moreover, that's just the US. Here is what the link has to say about countries less blessed with landmass and latitude: > Indeed, in smaller countries, substantial gaps (>30% of demand for >20 h per year; pale orange curves in Fig. 4) remain in systems even with 12 h of energy storage and annual generation that is 3x annual demand. Or this, directly referring to extensive overbuilding required to meet demand: > For instance, historical solar and wind resources data in Germany reveal that there were nearly 2 weeks in which dispatchable generation had to cover practically all of the demand because of a period with very low solar and wind power availability (called “dark doldrums”)27. Although with vast enough wind and solar capacity it might still be possible to meet demand in all hours, the required capacity increases exponentially after a point that depends on the renewable resources of that country, and it is this geophysically-dependent point that will largely determine the cost-effectiveness of highly-reliable, renewables-based electricity systems.

Retric 5y ago

Those numbers are wildly off because among other things the use US generates 6% hydroelectric, and they didn’t try to minimize a cost function via location selection etc. Further large hydroelectric dams store months worth of their average output. Also, demand and generation are positively correlated as people use less power on cloudy days and less AC in the winter. You can look at how much oversupply and storage people need to live off grid and realize that’s a worst case situation isn’t 7x oversupply or 30 days of storage it’s roughly 3x oversupply of solar and 3 days of storage. At grid scale that’s roughly 6c/kWh for solar plus batteries which last longer the less their used though that’s also a high when you add other energy sources and grid transmission. The good news is markets are great at minimizing cost functions. In the near term lots of natural gas makes up the difference, but the cost breakdown is such that offsetting 1kWh of natural gas works out even if 50% of solar generation is wasted. Aka 2x oversupply of solar shows up from market forces it’s that cheap. Bring on storage and things start to look very solar heavy even with zero subsidy or price reductions.

lambdadmitry 5y ago
Hydroelectric can't ramp up to a large fraction of national demand, for multiple unsurmountable reasons (like, you can't dump the water fast enough for one). Location selection doesn't matter that much as they don't try to optimise cost, they just look at aggregate output. People living off grid still continue to consume the products of grid-enabled industries, from manufacturing to farming to logistics to infrastructure. Heating/cooling and lighting a cabin is much more trivial than paving a road coming to that cabin, feeding its inhabitants, building solar panels for them, or ensuring dense enough population to make research and manufacturing viable. In fact, we don't need to extrapolate minor part of personal consumption to the whole society, we have aggregate energy consumption numbers. "Market forces" can't generate electricity on their own. Instead of taking the bottom-up view, fraught with wishful thinking and unstated assumptions, why not take the top down, working from the potential generation capacity under ideal assumptions? That's exactly what the paper I linked did, and the results are… not great for 100% renewable
- Retric 5y ago

bryanlarsen 5y ago

The book I linked includes a realistic model incorporating existing hydro, significant rooftop solar, et cetera. It requires a 20% overbuild. https://mitpress.mit.edu/books/electrify https://mitpress.mit.edu/books/electrify

lambdadmitry 5y ago
Hydroelectricity is very limited in its maximum instantaneous output, you can't feed half a grid with it, there is just nowhere to put all that water to without massive flooding, so it's a bit of a moot point. Rooftop solar doesn't matter for the top down approach in that paper, they don't make any assumptions about PV location and assume perfect grid, so rooftop PV won't make it any better. So seeing a top down, intentionally optimistic approach arriving at a number much higher than that book's author's, I am very sceptical of the book's results.