7 ms·

Instead of nitpicking terminology, let's talk about the issue directly: What do you do when the sun isn't shining and the wind isn't blowing? There isn't enou

Instead of nitpicking terminology, let's talk about the issue directly: What do you do when the sun isn't shining and the wind isn't blowing? There isn't enough batteries on earth to even make a dent in the storage you would need. We don't want to burn any more hydrocarbons. What are our other options? Massive pumped hydro might get us there, but that construction is on a similar time horizon to a new nuclear plant anyway.

bryanlarsen 4y ago
We'd need about 50 TWh of batteries. We can build that faster than we can build a single nuclear plant. And what do nukes do during peak periods of they are supplying base power?
- dahfizz 4y ago
  > We'd need about 50 TWh of batteries. We can build that faster than we can build a single nuclear plant. That's a huge [citation needed]. Global battery production is currently around 0.5 TWh per year[1]. Even if we continue to scale battery production, 50TWh is decades worth of our manufacturing capacity. Also, don't forget that the latest and greatest battery packs last ~10 years. Just the upkeep on your grid scale battery is 5 TWh/year. We don't have anything close to enough manufacturing. [1] https://www.spglobal.com/marketintelligence/en/news-insights/blog/top-electric-vehicle-markets-dominate-lithium-ion-battery-capacity-growth https://www.spglobal.com/marketintelligence/en/news-insights...
  - bryanlarsen 4y ago
    We're doubling battery factory capacity every 12-18 months.
ncmncm 4y ago
Only a negligible fraction of the storage that will be built will be batteries, so talking about batteries is just blowing smoke. Storage cost is falling even faster than renewables cost ever did. There are many different storage options. Each place will adopt the mix of them that makes sense for that place. Until new renewable generating capacity no longer displaces carbon emission, spending on storage would be stupid.
- epistasis 4y ago
  I disagree, we are in the hundreds of GWh/year of lithium ion production now, and it grows 10x roughly every five years. It's unlikely that any other storage tech can build supply chains quickly enough to compete with that growth. I certainly don't see pumped hydro doing this, unless we can stop relying on geography and construction projects. To make pumped hydro economical we would need something like spheres bolted to the ocean/lake floors, that pump water out and in for storage. And the materials costs for that are not falling in the same way as chemical storage.
  - ncmncm 4y ago
    Yet, the material cost for such a sphere, today, would be much less than for the equivalent amount of batteries, even at projected prices. It would only ever experience compressive force, which unreinforced concrete is very good at. Lithium is a poor choice for stationary storage, and competing with growth of electric car usage would be bad policy. Probably molten-metal batteries will turn out better for utility use, where you must have batteries. But there are a lot of other choices besides batteries and pumped hydro.
    - epistasis 4y ago
      On the prices, I haven't heard much optimistic news, the last I heard was at best, parity with current lithium ion prices: https://www.sciencedirect.com/science/article/abs/pii/S0360544217308733 https://www.sciencedirect.com/science/article/abs/pii/S03605... But that was five years ago, and as a manufactured product perhaps they can achieve better pricing over time. Lithium ion has no disadvantages for stationary storage, and it has one absolutely massive advantage: there are lots of higher-margin uses driving production scales up, and manufacturing costs down. If the dropping price of lithium ion batteries suddenly opens up new markets that are so huge, that demand can not be met, prices plateau for a long time, then I hope that other technologies can catch up. But as it is I see little chance for other chemistries or technologies to scale to match lithium ion in the near term future. I'm perhaps hopeful for iron-air batteries from Form Energy, but that's only because their management team has lots of hot shots, and not as much because they have an innate technical advantage. At this point in time, execution is everything, and there's precious few people with the executive capacity and the names that would allow them to attract sufficient capital to compete with the lithium ion juggernauts. (I would love nothing more than to be proven wrong. And I have been proven wrong on my skeptical tech takes in the past, so I know I'm not invincible here.)
epistasis 4y ago
Let's talk about this directly: what do you do when the nuclear reactor goes down because it's time to refuel, or because it's heating up the river too much in a heat wave (eg in France), or because some sensor valve froze shut in an decade-frequent cold snap (Texas)? You get the power from some other generator on the grid at a different geographic location. We can easily get to 80%-90% renewables that way, with current lithium ion storage, and no price decreases. With price decreases, and the huge panopoly of other storage chemistries that are close to shipping, and maybe even competing economically (iron, maybe vanadium flow, maybe zinc air), then we have clear paths to 100%. But even in France, nuclear never beat this 80%-90% mark, and it would absolutely bonkers insane and unrealistic to consider keeping the US's nuclear percentage at its current 20% of generation, because we can't build and we need to replace ~100 reactors very very very soon. We will be lucky if the nuclear industry could even deliver 10% of electricity needs in the future, because they simply can not build and deliver on their promises. Especially with electrification expanding energy consumed as electricity versus fossil fuels in the coming decades. Pumped hydro is also very expensive compared to lots of lithium batteries. We will soon be producing many TWh/year of lithium ion, getting to tens of TWh/year by the early 2030s. We can accelerate that curve, or decelerate it based on investments and regulatory structures. Nuclear is too far behind on tech to catch up. We simply can not grow the workforce or the supply chains of reactors quickly enough to sustain past construction levels at this pace. It would be foolish capital allocation.
- dahfizz 4y ago
  > what do you do when the nuclear reactor goes down because it's time to refuel, or because it's heating up the river too much in a heat wave (eg in France), or because some sensor valve froze shut in an decade-frequent cold snap (Texas)? > You get the power from some other generator on the grid at a different geographic location. Nuclear has a much better uptime then current renewable tech. In the winter, there is zero sunshine in North America for 10+ hours a day. And it just so happens that we hit peak energy usage while the sun is down. For nuclear, you only need storage "peaker plants" to smooth out demand. To go all in on renewalable and solar, you actually need to be able to service 100% of demand, for many hours, from storage. It won't be feasible for decades. I think your estimate for tens of TWh/year by mid 2030s is reasonable. But most of that is for EVs. Tesla has the largest currently running and the largest in-the-pipeline lithium ion battery factories, but all those batteries are going straight into an EV. Even if we dedicated 100% of batteries we can produce to be grid storage, we wouldn't have enough until ~2040. And, by the way, they all need to be replaced by 2050 because the cells all degrade and become dangerous. In reality, the spare production capacity we can dedicate to grid storage will always be small. I'm not convinced it can even catch up with our growing energy usage in our lifetime.

bryanlarsen 4y ago

We'd need about 50 TWh of batteries. We can build that faster than we can build a single nuclear plant. And what do nukes do during peak periods of they are supplying base power?

dahfizz 4y ago
> We'd need about 50 TWh of batteries. We can build that faster than we can build a single nuclear plant. That's a huge [citation needed]. Global battery production is currently around 0.5 TWh per year[1]. Even if we continue to scale battery production, 50TWh is decades worth of our manufacturing capacity. Also, don't forget that the latest and greatest battery packs last ~10 years. Just the upkeep on your grid scale battery is 5 TWh/year. We don't have anything close to enough manufacturing. [1] https://www.spglobal.com/marketintelligence/en/news-insights/blog/top-electric-vehicle-markets-dominate-lithium-ion-battery-capacity-growth https://www.spglobal.com/marketintelligence/en/news-insights...
- bryanlarsen 4y ago
  We're doubling battery factory capacity every 12-18 months.

ncmncm 4y ago

Only a negligible fraction of the storage that will be built will be batteries, so talking about batteries is just blowing smoke. Storage cost is falling even faster than renewables cost ever did. There are many different storage options. Each place will adopt the mix of them that makes sense for that place. Until new renewable generating capacity no longer displaces carbon emission, spending on storage would be stupid.

epistasis 4y ago
I disagree, we are in the hundreds of GWh/year of lithium ion production now, and it grows 10x roughly every five years. It's unlikely that any other storage tech can build supply chains quickly enough to compete with that growth. I certainly don't see pumped hydro doing this, unless we can stop relying on geography and construction projects. To make pumped hydro economical we would need something like spheres bolted to the ocean/lake floors, that pump water out and in for storage. And the materials costs for that are not falling in the same way as chemical storage.
- ncmncm 4y ago
  Yet, the material cost for such a sphere, today, would be much less than for the equivalent amount of batteries, even at projected prices. It would only ever experience compressive force, which unreinforced concrete is very good at. Lithium is a poor choice for stationary storage, and competing with growth of electric car usage would be bad policy. Probably molten-metal batteries will turn out better for utility use, where you must have batteries. But there are a lot of other choices besides batteries and pumped hydro.
  - epistasis 4y ago
    On the prices, I haven't heard much optimistic news, the last I heard was at best, parity with current lithium ion prices: https://www.sciencedirect.com/science/article/abs/pii/S0360544217308733 https://www.sciencedirect.com/science/article/abs/pii/S03605... But that was five years ago, and as a manufactured product perhaps they can achieve better pricing over time. Lithium ion has no disadvantages for stationary storage, and it has one absolutely massive advantage: there are lots of higher-margin uses driving production scales up, and manufacturing costs down. If the dropping price of lithium ion batteries suddenly opens up new markets that are so huge, that demand can not be met, prices plateau for a long time, then I hope that other technologies can catch up. But as it is I see little chance for other chemistries or technologies to scale to match lithium ion in the near term future. I'm perhaps hopeful for iron-air batteries from Form Energy, but that's only because their management team has lots of hot shots, and not as much because they have an innate technical advantage. At this point in time, execution is everything, and there's precious few people with the executive capacity and the names that would allow them to attract sufficient capital to compete with the lithium ion juggernauts. (I would love nothing more than to be proven wrong. And I have been proven wrong on my skeptical tech takes in the past, so I know I'm not invincible here.)

epistasis 4y ago

Let's talk about this directly: what do you do when the nuclear reactor goes down because it's time to refuel, or because it's heating up the river too much in a heat wave (eg in France), or because some sensor valve froze shut in an decade-frequent cold snap (Texas)? You get the power from some other generator on the grid at a different geographic location. We can easily get to 80%-90% renewables that way, with current lithium ion storage, and no price decreases. With price decreases, and the huge panopoly of other storage chemistries that are close to shipping, and maybe even competing economically (iron, maybe vanadium flow, maybe zinc air), then we have clear paths to 100%. But even in France, nuclear never beat this 80%-90% mark, and it would absolutely bonkers insane and unrealistic to consider keeping the US's nuclear percentage at its current 20% of generation, because we can't build and we need to replace ~100 reactors very very very soon. We will be lucky if the nuclear industry could even deliver 10% of electricity needs in the future, because they simply can not build and deliver on their promises. Especially with electrification expanding energy consumed as electricity versus fossil fuels in the coming decades. Pumped hydro is also very expensive compared to lots of lithium batteries. We will soon be producing many TWh/year of lithium ion, getting to tens of TWh/year by the early 2030s. We can accelerate that curve, or decelerate it based on investments and regulatory structures. Nuclear is too far behind on tech to catch up. We simply can not grow the workforce or the supply chains of reactors quickly enough to sustain past construction levels at this pace. It would be foolish capital allocation.

dahfizz 4y ago
> what do you do when the nuclear reactor goes down because it's time to refuel, or because it's heating up the river too much in a heat wave (eg in France), or because some sensor valve froze shut in an decade-frequent cold snap (Texas)? > You get the power from some other generator on the grid at a different geographic location. Nuclear has a much better uptime then current renewable tech. In the winter, there is zero sunshine in North America for 10+ hours a day. And it just so happens that we hit peak energy usage while the sun is down. For nuclear, you only need storage "peaker plants" to smooth out demand. To go all in on renewalable and solar, you actually need to be able to service 100% of demand, for many hours, from storage. It won't be feasible for decades. I think your estimate for tens of TWh/year by mid 2030s is reasonable. But most of that is for EVs. Tesla has the largest currently running and the largest in-the-pipeline lithium ion battery factories, but all those batteries are going straight into an EV. Even if we dedicated 100% of batteries we can produce to be grid storage, we wouldn't have enough until ~2040. And, by the way, they all need to be replaced by 2050 because the cells all degrade and become dangerous. In reality, the spare production capacity we can dedicate to grid storage will always be small. I'm not convinced it can even catch up with our growing energy usage in our lifetime.