7 ms·

I don’t think most people deny that solar is a good idea. Although isn’t it a statistic that all batteries we’ve ever manufactured would power the world for wha

by valval 2y ago

I don’t think most people deny that solar is a good idea. Although isn’t it a statistic that all batteries we’ve ever manufactured would power the world for what, 15 minutes?

toomuchtodo 2y ago
So, think in systems; humans are building a machine (the entire supply chain for solar and batteries) that builds a machine (the global energy grid). By the end of this year, the world will be deploying ~660GW of solar capacity per year. Within 18 months, trajectories of manufacturing capacity bring that to 1TW per year. It is unlikely that number goes anywhere but up. Now on to batteries. Again, let us assume this is a manufacturing floor, and battery manufacturing continues to scale up globally. > BNEF is tracking 7.9 TWh of annual battery manufacturing capacity announced for the end of 2025. That’s compared to demand projections of 1.6 TWh, and even that assumes steady EV demand growth and very rapid growth in batteries for storage applications. Even half that total announced capacity would be enough to equip almost every car sold in the world next year with a 50 kWh battery pack. [My note: Global light vehicle auto sales are ~90M units per year] We are, trying to be as rational and with as little hyperbole as possible, at the hockey stick inflection point looking almost straight up. Edit: It also appears the US is throwing another $3B into battery projects. https://news.ycombinator.com/item?id=41603409 https://news.ycombinator.com/item?id=41603409 ("HN: US announces $3B in funding for new battery projects") https://www.bloomberg.com/news/newsletters/2024-07-09/china-s-batteries-are-now-cheap-enough-to-power-huge-shifts https://www.bloomberg.com/news/newsletters/2024-07-09/china-... | https://archive.today/DklaA https://archive.today/DklaA ("Bloomberg: China’s Batteries Are Now Cheap Enough to Power Huge Shifts") https://www.bloomberg.com/news/newsletters/2024-04-12/china-already-makes-as-many-batteries-as-the-entire-world-wants https://www.bloomberg.com/news/newsletters/2024-04-12/china-... | https://archive.is/8Dy4D https://archive.is/8Dy4D ("Bloomberg: China Already Makes as Many Batteries as the Entire World Wants") (i track and provide reporting on global energy electrification velocity and clean energy policy on a volunteer basis for a cleantech startup)
- toomuchtodo 2y ago
  Additional citation: https://www.abc.net.au/news/2024-09-21/major-climate-agencies-call-global-emissions-peak/104016030 https://www.abc.net.au/news/2024-09-21/major-climate-agencie... (“The rise of solar power and China's staggering EV growth may have pushed global emissions into decline”)
WJW 2y ago
That's only because we haven't made a lot of batteries, historically speaking. Their price per kWh was too high with the technology of the time. With modern battery technology there is nothing stopping us from building dramatically more capacity.
MichaelNolan 2y ago
Basically all the batteries that have ever been built (by volume), have been built in the last 3 years. The party is just getting started. So while it might only be 15 minutes today, it's going to 1000x in just a few years. https://rmi.org/the-rise-of-batteries-in-six-charts-and-not-too-many-numbers/ https://rmi.org/the-rise-of-batteries-in-six-charts-and-not-...
- aidenn0 2y ago
  That chart makes me annoyed that all UPSs under $1000 use SLA batteries. Where's my compact LiFePO UPS or same-sized UPS that lasts for an hour?
  - MichaelNolan 2y ago
    How many kwh do you need? These are just over $1000 - https://signaturesolar.com/eg4-lifepower4-v2-lithium-battery-48v-100ah-server-rack-battery-ul1973-ul9540a-10-year-warranty/ https://signaturesolar.com/eg4-lifepower4-v2-lithium-battery...
    - cogman10 2y ago
      Holy cow! 5kwh for $1k is amazing!
    - aidenn0 2y ago
      That is not a UPS.
      - MichaelNolan 2y ago
        What’s the difference between a regular battery with a smart inverter and a UPS? It’s just that the UPS is an all in 1 box right? Or is there something else that a ups has?
        aidenn0 2y ago
        Firstly, the linked battery doesn't have a smart inverter, so still not a UPS. An online UPS will also have a AC-DC converter (rectifier) in addition to the DC-AC converter (inverter). It usually also has fans to cool those since they are running continuously, even when AC power is available. An offline UPS is basically an online UPS with a relay, so that if the AC power is working, all of the UPS circuitry is bypassed. This reduces the usage of the DC electronics and increases the efficiency when power-loss is rare (since the AC->DC->AC path will always dissipate some power). A line-interactive transformer is like an offline UPS, but with voltage conditioning (e.g. via an autotransformer) to handle voltages that are a bit low or a bit high. Consumer UPSs are (almost all?) either offline or line-interactive.
audunw 2y ago
What’s insane is that battery grid storage has already overtaken pumped hydro storage, and practically all of that is in just the last few years. We haven’t really even ramped up production of battery chemistries that are tailor made for grid storage yet. Batteries like sodium-ion, molten metal or various flow batteries have basically no limits to how much we can scale them, and a really low potential price floor.
cogman10 2y ago
> Although isn’t it a statistic that all batteries we’ve ever manufactured would power the world for what, 15 minutes Is it really that much? That's incredible! You have to realize that battery manufacturing capacity has been growing at an exponential rate. 15 minutes of global power capacity will quickly turn into 30 minutes, 1 hour, 2 hours, etc. 8 hours of global storage is enough to survive the night with just solar alone. 15 minutes is already a huge win. The more lead time we have for production ramping, the cheaper and more efficient we can make the plants that provide backup power.
_greim_ 2y ago
What's messing with people's heads when it comes to predicting solar is the perfectly rational tendency to base projections on past and current data. I think it qualifies as a black swan event for this reason.
- cogman10 2y ago
  Ditto with batteries. I very frequently see people talk about batteries with notions from the 2000s about expense and manufacturing capability. Batteries have become dirt cheap to make and we are very quickly getting to the point where the power electronics are actually the more expensive part of a battery deployment rather than the battery capacity itself.
  - 7952 2y ago
    Yes you often here this with nuclear advocates. Saying that grid scale battery storage is a pipe dream. The real defining fact about solar and battery storage is that it is very amenable to mass production and scaling. Small modular components that can be produced in a factory and require minimal maintenance.
- pfdietz 2y ago
  And you see people who confidently spout conclusions that have long been rendered obsolete. It's not just that they don't understand growth, it's that their conclusions ossified a decade or two ago. They're basing them on far past data, not recent past data.
  - namaria 2y ago
    Good. Other people, the majority of people being wrong, can be very profitable for those who are right. I'm in no crusade to make people think like me. I am on a path of leveraging my knowledge.
jillesvangurp 2y ago
Let's counter that with some real statistics: Batteries are being produced at TWH/year scale now. Probably crossing into 1.5-2 TWH/year this year. That's new batteries added every year. This is projected to grow over the next few years. Most of those batteries have thousands of cycles of lifetime. Tens of thousands for some of the newer chemistries (e.g. sodium ion). Overall electricity production is 25-30 TWH per year; for the entire planet. So 25000 TWH. That's going to go up of course but it's a nice number to work with. Those batteries can cycle multiple times per day. Lets assume once a day (they'd be able to do multiple of course). So you get about 365 TWH of power cycled per day out of 1 TWH of batteries (discharged and charged again). You'd charge them with cheap solar, wind, geothermal, and whatever else you have available; whenever you have that available. About 70 TWH of battery would be enough to get to 25000 TWH per year. We'll have that probably in the early 2040s. Once we hit tens of TWH of battery produced per year that won't take long. Of course electricity production and demand will increase as well. Of course, we'll have plenty of other power sources. So, most of those batteries will just be sitting there fully charged most of the time and we won't actually be cycling most of them all that much. That's a lot of potential energy. Combine that with cables, dynamic pricing, demand shaping, and a few other things and the implied need for long term storage melts away.
- valval 2y ago
  Where do the raw materials for that production ramp come from? Are they abundant?
  - toomuchtodo 2y ago
    https://www.sustainabilitybynumbers.com/p/lithium-electric-vehicles https://www.sustainabilitybynumbers.com/p/lithium-electric-v... https://www.economist.com/science-and-technology/2024/09/19/most-electric-car-batteries-could-soon-be-made-by-recycling-old-ones https://www.economist.com/science-and-technology/2024/09/19/... | https://archive.today/OM62h https://archive.today/OM62h

by valval 2y ago

I don’t think most people deny that solar is a good idea. Although isn’t it a statistic that all batteries we’ve ever manufactured would power the world for what, 15 minutes?

toomuchtodo 2y ago

So, think in systems; humans are building a machine (the entire supply chain for solar and batteries) that builds a machine (the global energy grid). By the end of this year, the world will be deploying ~660GW of solar capacity per year. Within 18 months, trajectories of manufacturing capacity bring that to 1TW per year. It is unlikely that number goes anywhere but up. Now on to batteries. Again, let us assume this is a manufacturing floor, and battery manufacturing continues to scale up globally. > BNEF is tracking 7.9 TWh of annual battery manufacturing capacity announced for the end of 2025. That’s compared to demand projections of 1.6 TWh, and even that assumes steady EV demand growth and very rapid growth in batteries for storage applications. Even half that total announced capacity would be enough to equip almost every car sold in the world next year with a 50 kWh battery pack. [My note: Global light vehicle auto sales are ~90M units per year] We are, trying to be as rational and with as little hyperbole as possible, at the hockey stick inflection point looking almost straight up. Edit: It also appears the US is throwing another $3B into battery projects. https://news.ycombinator.com/item?id=41603409 https://news.ycombinator.com/item?id=41603409 ("HN: US announces $3B in funding for new battery projects") https://www.bloomberg.com/news/newsletters/2024-07-09/china-s-batteries-are-now-cheap-enough-to-power-huge-shifts https://www.bloomberg.com/news/newsletters/2024-07-09/china-... | https://archive.today/DklaA https://archive.today/DklaA ("Bloomberg: China’s Batteries Are Now Cheap Enough to Power Huge Shifts") https://www.bloomberg.com/news/newsletters/2024-04-12/china-already-makes-as-many-batteries-as-the-entire-world-wants https://www.bloomberg.com/news/newsletters/2024-04-12/china-... | https://archive.is/8Dy4D https://archive.is/8Dy4D ("Bloomberg: China Already Makes as Many Batteries as the Entire World Wants") (i track and provide reporting on global energy electrification velocity and clean energy policy on a volunteer basis for a cleantech startup)

toomuchtodo 2y ago
Additional citation: https://www.abc.net.au/news/2024-09-21/major-climate-agencies-call-global-emissions-peak/104016030 https://www.abc.net.au/news/2024-09-21/major-climate-agencie... (“The rise of solar power and China's staggering EV growth may have pushed global emissions into decline”)

WJW 2y ago

That's only because we haven't made a lot of batteries, historically speaking. Their price per kWh was too high with the technology of the time. With modern battery technology there is nothing stopping us from building dramatically more capacity.

MichaelNolan 2y ago

Basically all the batteries that have ever been built (by volume), have been built in the last 3 years. The party is just getting started. So while it might only be 15 minutes today, it's going to 1000x in just a few years. https://rmi.org/the-rise-of-batteries-in-six-charts-and-not-too-many-numbers/ https://rmi.org/the-rise-of-batteries-in-six-charts-and-not-...

aidenn0 2y ago
That chart makes me annoyed that all UPSs under $1000 use SLA batteries. Where's my compact LiFePO UPS or same-sized UPS that lasts for an hour?
- MichaelNolan 2y ago
  How many kwh do you need? These are just over $1000 - https://signaturesolar.com/eg4-lifepower4-v2-lithium-battery-48v-100ah-server-rack-battery-ul1973-ul9540a-10-year-warranty/ https://signaturesolar.com/eg4-lifepower4-v2-lithium-battery...
  - cogman10 2y ago
    Holy cow! 5kwh for $1k is amazing!
  - aidenn0 2y ago
    That is not a UPS.
    - MichaelNolan 2y ago
      What’s the difference between a regular battery with a smart inverter and a UPS? It’s just that the UPS is an all in 1 box right? Or is there something else that a ups has?
      - aidenn0 2y ago
        Firstly, the linked battery doesn't have a smart inverter, so still not a UPS. An online UPS will also have a AC-DC converter (rectifier) in addition to the DC-AC converter (inverter). It usually also has fans to cool those since they are running continuously, even when AC power is available. An offline UPS is basically an online UPS with a relay, so that if the AC power is working, all of the UPS circuitry is bypassed. This reduces the usage of the DC electronics and increases the efficiency when power-loss is rare (since the AC->DC->AC path will always dissipate some power). A line-interactive transformer is like an offline UPS, but with voltage conditioning (e.g. via an autotransformer) to handle voltages that are a bit low or a bit high. Consumer UPSs are (almost all?) either offline or line-interactive.

audunw 2y ago

What’s insane is that battery grid storage has already overtaken pumped hydro storage, and practically all of that is in just the last few years. We haven’t really even ramped up production of battery chemistries that are tailor made for grid storage yet. Batteries like sodium-ion, molten metal or various flow batteries have basically no limits to how much we can scale them, and a really low potential price floor.

cogman10 2y ago

> Although isn’t it a statistic that all batteries we’ve ever manufactured would power the world for what, 15 minutes Is it really that much? That's incredible! You have to realize that battery manufacturing capacity has been growing at an exponential rate. 15 minutes of global power capacity will quickly turn into 30 minutes, 1 hour, 2 hours, etc. 8 hours of global storage is enough to survive the night with just solar alone. 15 minutes is already a huge win. The more lead time we have for production ramping, the cheaper and more efficient we can make the plants that provide backup power.

_greim_ 2y ago

What's messing with people's heads when it comes to predicting solar is the perfectly rational tendency to base projections on past and current data. I think it qualifies as a black swan event for this reason.

cogman10 2y ago
Ditto with batteries. I very frequently see people talk about batteries with notions from the 2000s about expense and manufacturing capability. Batteries have become dirt cheap to make and we are very quickly getting to the point where the power electronics are actually the more expensive part of a battery deployment rather than the battery capacity itself.
- 7952 2y ago
  Yes you often here this with nuclear advocates. Saying that grid scale battery storage is a pipe dream. The real defining fact about solar and battery storage is that it is very amenable to mass production and scaling. Small modular components that can be produced in a factory and require minimal maintenance.
pfdietz 2y ago
And you see people who confidently spout conclusions that have long been rendered obsolete. It's not just that they don't understand growth, it's that their conclusions ossified a decade or two ago. They're basing them on far past data, not recent past data.
- namaria 2y ago
  Good. Other people, the majority of people being wrong, can be very profitable for those who are right. I'm in no crusade to make people think like me. I am on a path of leveraging my knowledge.

jillesvangurp 2y ago

Let's counter that with some real statistics: Batteries are being produced at TWH/year scale now. Probably crossing into 1.5-2 TWH/year this year. That's new batteries added every year. This is projected to grow over the next few years. Most of those batteries have thousands of cycles of lifetime. Tens of thousands for some of the newer chemistries (e.g. sodium ion). Overall electricity production is 25-30 TWH per year; for the entire planet. So 25000 TWH. That's going to go up of course but it's a nice number to work with. Those batteries can cycle multiple times per day. Lets assume once a day (they'd be able to do multiple of course). So you get about 365 TWH of power cycled per day out of 1 TWH of batteries (discharged and charged again). You'd charge them with cheap solar, wind, geothermal, and whatever else you have available; whenever you have that available. About 70 TWH of battery would be enough to get to 25000 TWH per year. We'll have that probably in the early 2040s. Once we hit tens of TWH of battery produced per year that won't take long. Of course electricity production and demand will increase as well. Of course, we'll have plenty of other power sources. So, most of those batteries will just be sitting there fully charged most of the time and we won't actually be cycling most of them all that much. That's a lot of potential energy. Combine that with cables, dynamic pricing, demand shaping, and a few other things and the implied need for long term storage melts away.

valval 2y ago
Where do the raw materials for that production ramp come from? Are they abundant?
- toomuchtodo 2y ago
  https://www.sustainabilitybynumbers.com/p/lithium-electric-vehicles https://www.sustainabilitybynumbers.com/p/lithium-electric-v... https://www.economist.com/science-and-technology/2024/09/19/most-electric-car-batteries-could-soon-be-made-by-recycling-old-ones https://www.economist.com/science-and-technology/2024/09/19/... | https://archive.today/OM62h https://archive.today/OM62h