Batteries probably work for this, though prices will have to drop a little more for this.
For seasonal storage, pumped storage and chemical (power to gas) are the more likely solutions.
However, if might turn out that simply overbuilding solar is more cost effective than a lot of seasonal storage.
And yes, you will always need backup generation capacity based on gas.
But those plants can be inefficient and cheaply built, since they won’t run many hours in the year.
I don't see it this way. It's democratized. Each person is slowly going to get batteries as they try to avoid skyrocketing and unfair income based fees that are being introduced. The ultimate goal is to get off the grid, then get your friends off the grid, and so on and so forth. Cheap prismatic batteries from China baby. Price still have a bit to go, but it's very close.
Cheap, slave made batteries from China baby... but we'll just skip over that part, eh?
They're cheap because some bugger ain't getting paid. Then there's the added bonus that that artificial cheapness drives everybody else out of the market.
Depends what your starting assumptions are. Do you think we need enough lithium ion batteries to store 6 months of storage at the standard US energy consumption rate to get through winter? If you do, then you're right: we can't build lithium ion batteries on that scale.
If your assumption is that we need 12 hours of energy storage to get through the night, then it's a much more realistic goal. One could even imagine the batteries being on the same premises where the energy is generated (by solar) and used, in which case it's imposing no load on the energy grid.
If we consider that long-distance high voltage DC lines are an option, we might not even need storage in the first place. In the middle of the night in the US we could be buying solar power from Morocco and Australia. We just need to have the political will to finance the construction of those cables.
Apparently China has plans to build a trans-Pacific HVDC line so they can buy solar power from Chile. Maybe that's just talk and they won't actually complete the project, but at least they're willing to set ambitious goals.
One thing to that I don't have a solid source of numbers to back up but I highly suspect is that eventually the power will be more expensive during the late night than the early morning. As a result a lot of demand will shift to reflect that.
That could be, though it's likely to be location specific. Areas with good hydroelectric might just have steady pricing -- solar during the day, hydro at night.
EV owners often get reduced rates for charging at night; that could go away. That's an argument in favor of electrification of our major highways: if cars and trucks can charge while in motion, that shifts their time of maximum energy consuption to the daylight hours when people do most of their driving, and which better aligns with solar production. You might even have highways more-or-less directly powered by adjacent solar farms. Fortunately highways and solar farms both have reasons to exist in sparsely-populated countryside.
If there's one important thing that all political spectrums can agree on at the moment, it's solar. Some do it for the environemnt, some to be independent from the grid, others to save money, but PV on every roof with local batteries will win. And that's exactly why it scales _better_ than every other solution, because it needs less grid capacity and no central planning.
>And yes, you will always need backup generation capacity based on gas
Always is doing a lot of work in that sentence. I would disagree on the longer term. Deep geothermal for one technology has the capacity to stop us needing gas for such use, and given enough storage, I think the whole system of operation will change.
If you have deep geothermal you don't need the solar panels though. It's more cost-effective to build a geothermal plant that you run at full power all the time (because you sell energy by the kWh) rather then trying to cycle it up and down.
Build enough geothermal to run base night load, use solar+wind+hydro+battery storage for day demand (and install a ton of electrolyzers to catch the unused electricity in overproduction times), keep a small reserve fleet of biogas/power-to-gas powered plants for emergencies, and get as many of the large consumers (as well as private loads such as ACs, EV chargers) as possible to agree on dynamic load shedding.
That does make sense; while you could run the geothermal 24/7, you might actually be better off only running it when you need to, since (as I understand it) geothermal can actually use up all the usable heat in an area, which can take a long time to "recharge" through conduction from surrounding rock.
Fun fact: if you do it right, you can reverse the geothermal heat pump to store energy in the form of heat in the ground in overproduction (i.e. zero or negative electricity rate) times.
Wind helps supplement a lot most of the time. But there are times when there's no sun and no wind. For that, there's currently hydro, nuclear, and eventually batteries.
There are many ideas in the pipeline for grid scale storage to smooth day-to-day renewable intermittency, but we also need seasonal storage. We need the ability to draw energy from storage for extended periods (weeks or months), to cover low energy production during winter.
The seasonal energy storage problem is one that humanity (and, really, life in general) has always faced. Many of our traditional food products are methods of preserving energy from the good times to tide us over in the bad times. We have traditionally adapted our activities for the season. I would imagine that part of the solution will be re-introducing seasonality to industry (e.g. making aluminium mostly in the summer).
At the rate that solar costs are falling we really don't. We just over-build and energy during the day is nearly zero cost. There is no problem with curtailing excess solar production, or opportunistically use that daytime power curve to run non time critical industrial processes.
Then you would have a lot of excess energy during the summer and this energy would then look for an use. Metal and cement production are very likely candidates and seasonality would be then dictated by the market.
No one knows exactly where economy of scale will land us, but much like Moore's law the trend in the data is there. Of course any major change in these industries are going to have knock on effects. We have been automating the steel production process in 1st world countries for years simply due to the cost of human labor. I expect that trend will continue and those plant operators will effectively end up being paid to work on other things on non operation days.
I think green hydrogen production is the obvious use case, but smelting, carbon capture and other things you would do with "free" power all make sense.
California installed about 5GW of batteries in the last three years.
Today natural gas peaked at 16GW. Batteries at the same time were delivering 3.8GW. During the day renewables which is mostly but not all solar peaked at 18.5GW.
So currently shifting my guess about 2 percent of the daily power generation. So not huge. But that didn't even exist three tears ago.
Let's see: it takes 9.8 joules to lift 1 kg by 1m. Let's say the tower is 100m high. The video says the blocks are 30 tons. 1 joule is 0.000278 watt hours.
9.8 * 30000 * 100 * 0.000278 * 0.0001 = 0.81732 kilowatt hours per 30 ton brick.
They claim they can store 100 mwh of energy. So, that's
100000 / 0.81732 = about 122,351 bricks.
From the top of the building it looks like it stores about 22x24 cubes per level, assuming the parts that aren't following the grid pattern are the lifts. So, 528 cubes per level. That would mean they need 231 levels of upper cube storage at an average height of 100 meters above the lower cube storage area, and it looks like they've only got about 29 total so far.
The math seems suspicious, but maybe my assumptions are wrong, or I made a calculation error. The structure might continue underground quite some distance, or the 30 ton cubes in the video aren't what they're using on the actual structure, or what looks like a single cube in the images are actually 8 or 27 cubes stacked into a bigger cube.
I'd be curious how this compares to the cost of LFP batteries, which are somewhere around $100 per kwh or so (in China). My guess is it's probably a lot more expensive than batteries or pumped hydroelectric storage.
Assuming your 278 factor is correct in value and scale (I didn't check)
9.8 * 30000 * 100 * 0.000278 * 0.0001 = 0.81732 kilowatt hours per 30 ton brick.
should(?) read
9.8 * 30000 * 100 * 0.000278 * 0.001 = 8.1732 kilowatt hours per 30 ton brick.
surely? 0.1 is a tenth, 0.01 is one hundreth, looks like you threw in an extra shift for luck.
> My guess is it's probably a lot more expensive than batteries or pumped hydroelectric storage.
Location, location, location!!
Not everywhere is suitable for pumped hydroelectric - relatively few places have dams or are dam suitable.
This at least fits in with modern city aesthetics and might be integrated and expanded to include dual use electric vehicle parking functionality ... (with vaguely interesting optimal usage algorithms).
Ah, so I did. 8.1 kwh per brick sounds like a more substantial amount of energy.
They'd need about 12,235 bricks at a 100m height, which is closer to practical.
I'm still kind of skeptical. I mean, pumped hydroelectric storage is pretty cheap. It's hard to compete with that; but then maybe there are uses for a big gravity battery in a dense urban area where hydroelectric isn't an option and maybe even batteries are a potential public health and safety risk. Maybe I've spent too much time alternating between Kicad and Factorio lately, but the concept of plopping down a bypass capacitor the size of a large building to buffer out "noise" on your power grid seems kind of like a normal thing to do.
Yeah, thing about Switzerland is that a lot of global companies are based there and market products and services globally that aren't swiss clocks, chocolates, dams in the alps, lederhosen, etc.
This particular company has multiple gravity battery designs that it's shopping globally and building the first industrial scale installation in China (checks map) some distance from Switzerland.
So, again, not everywhere is suitable for pumped hydroelectric and (obviously one would have hoped) it's okay to develop products for places that have no dams while living in a place that has dams.
"a day or two" would handle probably 90% of situations -- or 100% if there's enough solar overcapacity to power everything even on winter days (which probably involves good power grid connections to places like Arizona and Mexico).
I think we need to close all the fossil fuel plants eventually, but keeping them online to be used maybe a few days or weeks a year is the most responsible way to use them.
I wish it would be this way but unfortunately the worst solar months are really bad if you are on the northern hemisphere, in my place it's 3% of the yearly production in December and same in January.
It might be like that in some places, but not everywhere. HVDC lines are pretty efficient. From this map it looks like Arizona, New Mexico, Texas, Florida, and presumably Regular Mexico could be major producers for the rest of the US.
The most low hanging fruit for seasonal storage is most likely thermal storage connected to district heating. At least technologically, not 100% sure on economics, but given the scalability laws (heat loss scales to square and heat capacity to cube of the size of the storage), I am optimistic there as well.
Sure, battery production volumes are increasing rapidly. Going from hundreds to thousands of gwh annually. Basically twh territory.
32 GW at a capacity factor of about 0.25 (about right for solar, I think) pumps out about 70twh annually. Sounds like a lot but the US consumes about 4000twh per year. So this only represents about 1.7% of its overall energy production. The US is still catching up with other countries that source a larger percentage from solar already (e.g. China, Germany). I think globally it's around 5% currently. But undeniably the IRA is delivering big time in terms of economic stimulation and the bootstrapping of what are no doubt going to be very profitable businesses. And as a side effect, a lot of renewable energy capacity is coming online very rapidly.
Those car batteries could be cycled daily of course. There's will be an order of magnitude more battery on the roads (just in cars) than you can charge with this amount of solar. Of course those cars don't discharge all that much. Mostly they don't drive anywhere near their maximum range on a daily basis and are near fully charged most of the time.
So, that's going to be tens of twh that is ready to be used, in principle. We'll probably never utilize most of that available capacity for grid usage. Not because we can't but because there are more economical forms of storage than car batteries that are also becoming available. In other words, we'll have lots of redundant storage options. Domestic battery, grid battery, truck batteries, buses, ships, cars, etc. Tens, hundreds of twh of capacity.
Aside from energy storage, which the other comments touch, there is also the use of variable pricing to drive demand during peak solar production hours. Factories and other big users would have their most intensive work aligned to when energy is cheapest during the day. That way any needed storage is minimized.
Problem is peak electricity usage is 4-9pm when people come home from work, turn on the A/C, TVs, stoves, lights, etc. Residential users, not factories. Moving that peak is not feasible unless you're somehow changing the standard 9 to 5 workday. So raising peak hours prices too much will just hurt a majority of people
California has already mandated migrating plans to variable pricing, because there's so much solar. They call it time-of-use plans, and that's where I'm pulling 4-9pm from. Weekend peak hours are also priced lower because more people are home during the day and the peaks are lower.
Burning gas or coal 24 hours a day because "the sun only shines for 8 hours" is the kind of gigabrain solution that the "what about when the sun doesn’t shine and the wind doesnt blow" crowd like the most.
The real answer; burn something, and enjoy the 8 hours when the sun is shining to get electricity that is cheaper.
> Solar deployment is now happening at a roughly $500B annualized rate.
Which technology deployments were larger than this? The US's aircraft production during WWII seems to have peaked at maybe $400B (inflation-adjusted). Global datacenter construction appears to be maybe $200B/year.
And India at #4 is one to watch. There are a lot of people there who will benefit from cheap energy.
It is a bit funny how Germany seems to be floundering thanks to the Energiewende. In hindsight it looks like a clear mistake to have gone in so hard without waiting until it made economic sense.
Take a look at a map, India is far, far more to the south than Germany. I have the feeling that some Germans would see solar power as a panacea even if they lived on the north pole.
Germany will have around 12 GW new solar power installed in 2023 (with about 1/4th of the US population), so a little bit more per capita than the US. It is important to know that almost all of Germany is more to the north than the most northern parts of the US (except Alaska), so solar power in Germany is much less efficient and in winter almost completely useless.
Yeah, shortening energy supply when you already have a supply crunch is very good way to kill the economy. The only question is if this was done by stupidity or malice.
The shutdown had nothing to do with renewables. It was done by conservatives who, in the same year they took this step, proudly proclaimed that they stopped the buildout of PV in Germany.
Do you have more information about that? If you average the power over a month, then yes it might be true in the plot, but I suspect it's less smooth in real-time.
It doesn't average out on an intraday basis, but neither does consumption, which has its peaks during the day.
I know someone working at an energy provider heavy on renewables, there they use gas turbines (=biomass) for compensating heavy fluctuations, because those are online in a few minutes.
77 comments
[ 2.7 ms ] story [ 156 ms ] threadFor seasonal storage, pumped storage and chemical (power to gas) are the more likely solutions. However, if might turn out that simply overbuilding solar is more cost effective than a lot of seasonal storage.
And yes, you will always need backup generation capacity based on gas.
But those plants can be inefficient and cheaply built, since they won’t run many hours in the year.
https://www.caiso.com/TodaysOutlook/Pages/supply.html#sectio...
Very scalable. It’s happening right now and accelerating.
Cheap, slave made batteries from China baby... but we'll just skip over that part, eh?
They're cheap because some bugger ain't getting paid. Then there's the added bonus that that artificial cheapness drives everybody else out of the market.
https://www.nytimes.com/2022/06/20/business/economy/forced-l...
If your assumption is that we need 12 hours of energy storage to get through the night, then it's a much more realistic goal. One could even imagine the batteries being on the same premises where the energy is generated (by solar) and used, in which case it's imposing no load on the energy grid.
If we consider that long-distance high voltage DC lines are an option, we might not even need storage in the first place. In the middle of the night in the US we could be buying solar power from Morocco and Australia. We just need to have the political will to finance the construction of those cables.
Apparently China has plans to build a trans-Pacific HVDC line so they can buy solar power from Chile. Maybe that's just talk and they won't actually complete the project, but at least they're willing to set ambitious goals.
EV owners often get reduced rates for charging at night; that could go away. That's an argument in favor of electrification of our major highways: if cars and trucks can charge while in motion, that shifts their time of maximum energy consuption to the daylight hours when people do most of their driving, and which better aligns with solar production. You might even have highways more-or-less directly powered by adjacent solar farms. Fortunately highways and solar farms both have reasons to exist in sparsely-populated countryside.
Always is doing a lot of work in that sentence. I would disagree on the longer term. Deep geothermal for one technology has the capacity to stop us needing gas for such use, and given enough storage, I think the whole system of operation will change.
Heat pumps are pretty awesome.
The seasonal energy storage problem is one that humanity (and, really, life in general) has always faced. Many of our traditional food products are methods of preserving energy from the good times to tide us over in the bad times. We have traditionally adapted our activities for the season. I would imagine that part of the solution will be re-introducing seasonality to industry (e.g. making aluminium mostly in the summer).
Also consider the workforce differences too, since presumably this would also need to be on leave for a portion of the year.
Today natural gas peaked at 16GW. Batteries at the same time were delivering 3.8GW. During the day renewables which is mostly but not all solar peaked at 18.5GW.
So currently shifting my guess about 2 percent of the daily power generation. So not huge. But that didn't even exist three tears ago.
For simple math you need to lift/drop 1 tonne a 367 km to store 1 MWh of energy.
https://www.energyvault.com/project-cn-rudong
Apparently you can use multiple weights in parallel ...
https://vimeo.com/647372871
9.8 * 30000 * 100 * 0.000278 * 0.0001 = 0.81732 kilowatt hours per 30 ton brick.
They claim they can store 100 mwh of energy. So, that's 100000 / 0.81732 = about 122,351 bricks.
From the top of the building it looks like it stores about 22x24 cubes per level, assuming the parts that aren't following the grid pattern are the lifts. So, 528 cubes per level. That would mean they need 231 levels of upper cube storage at an average height of 100 meters above the lower cube storage area, and it looks like they've only got about 29 total so far.
The math seems suspicious, but maybe my assumptions are wrong, or I made a calculation error. The structure might continue underground quite some distance, or the 30 ton cubes in the video aren't what they're using on the actual structure, or what looks like a single cube in the images are actually 8 or 27 cubes stacked into a bigger cube.
I'd be curious how this compares to the cost of LFP batteries, which are somewhere around $100 per kwh or so (in China). My guess is it's probably a lot more expensive than batteries or pumped hydroelectric storage.
9.8 * 30000 * 100 * 0.000278 * 0.0001 = 0.81732 kilowatt hours per 30 ton brick.
should(?) read
9.8 * 30000 * 100 * 0.000278 * 0.001 = 8.1732 kilowatt hours per 30 ton brick.
surely? 0.1 is a tenth, 0.01 is one hundreth, looks like you threw in an extra shift for luck.
> My guess is it's probably a lot more expensive than batteries or pumped hydroelectric storage.
Location, location, location!!
Not everywhere is suitable for pumped hydroelectric - relatively few places have dams or are dam suitable.
This at least fits in with modern city aesthetics and might be integrated and expanded to include dual use electric vehicle parking functionality ... (with vaguely interesting optimal usage algorithms).
I'm still kind of skeptical. I mean, pumped hydroelectric storage is pretty cheap. It's hard to compete with that; but then maybe there are uses for a big gravity battery in a dense urban area where hydroelectric isn't an option and maybe even batteries are a potential public health and safety risk. Maybe I've spent too much time alternating between Kicad and Factorio lately, but the concept of plopping down a bypass capacitor the size of a large building to buffer out "noise" on your power grid seems kind of like a normal thing to do.
Sure but one of the prototypes was based in Switzerland... probably one of the best place on earth for dams.
This particular company has multiple gravity battery designs that it's shopping globally and building the first industrial scale installation in China (checks map) some distance from Switzerland.
So, again, not everywhere is suitable for pumped hydroelectric and (obviously one would have hoped) it's okay to develop products for places that have no dams while living in a place that has dams.
Do you have a point to make?
I think we need to close all the fossil fuel plants eventually, but keeping them online to be used maybe a few days or weeks a year is the most responsible way to use them.
It's really badly uneven.
https://www.solar-electric.com/learning-center/solar-insolat...
In the Northwest we also have pretty good hydroelectric to handle the cold rainy months. Every region is in its own specific situation.
Alaska would be a rough place for solar. I don't know if they do much hydroelectric or wind, but it seems like they'd be well situated for either.
If nothing else works, there's always nuclear.
32 GW at a capacity factor of about 0.25 (about right for solar, I think) pumps out about 70twh annually. Sounds like a lot but the US consumes about 4000twh per year. So this only represents about 1.7% of its overall energy production. The US is still catching up with other countries that source a larger percentage from solar already (e.g. China, Germany). I think globally it's around 5% currently. But undeniably the IRA is delivering big time in terms of economic stimulation and the bootstrapping of what are no doubt going to be very profitable businesses. And as a side effect, a lot of renewable energy capacity is coming online very rapidly.
Those car batteries could be cycled daily of course. There's will be an order of magnitude more battery on the roads (just in cars) than you can charge with this amount of solar. Of course those cars don't discharge all that much. Mostly they don't drive anywhere near their maximum range on a daily basis and are near fully charged most of the time.
So, that's going to be tens of twh that is ready to be used, in principle. We'll probably never utilize most of that available capacity for grid usage. Not because we can't but because there are more economical forms of storage than car batteries that are also becoming available. In other words, we'll have lots of redundant storage options. Domestic battery, grid battery, truck batteries, buses, ships, cars, etc. Tens, hundreds of twh of capacity.
California has already mandated migrating plans to variable pricing, because there's so much solar. They call it time-of-use plans, and that's where I'm pulling 4-9pm from. Weekend peak hours are also priced lower because more people are home during the day and the peaks are lower.
https://www.eia.gov/electricity/gridmonitor/dashboard/electr... Last weekend's lowered peaks are a bit less noticable because of the holiday Monday though.
The real answer; burn something, and enjoy the 8 hours when the sun is shining to get electricity that is cheaper.
Who produce it?
Why other place say sahari or even … global seems an odd word here.
At the same time, I'm not sure if you need any rare element in the manufacturing process that you can source from specific countries only.
> Solar deployment is now happening at a roughly $500B annualized rate.
Which technology deployments were larger than this? The US's aircraft production during WWII seems to have peaked at maybe $400B (inflation-adjusted). Global datacenter construction appears to be maybe $200B/year.
More here: https://about.bnef.com/blog/renewable-energy-investment-hits...
https://nitter.net/patrickc/status/1698496805162353029
China leads in steepest trend, but US is second (though much lower down)
https://ourworldindata.org/grapher/solar-electricity-per-cap...
https://ourworldindata.org/grapher/solar-electricity-per-cap...
It is a bit funny how Germany seems to be floundering thanks to the Energiewende. In hindsight it looks like a clear mistake to have gone in so hard without waiting until it made economic sense.
https://en.wikipedia.org/wiki/Solar_power_by_country
https://www.visualcapitalist.com/mapped-solar-power-by-count...
Good thing they were thinking ahead and shut down all nuclear plants they had. Fricking unbelievable
It doesn't average out on an intraday basis, but neither does consumption, which has its peaks during the day.
I know someone working at an energy provider heavy on renewables, there they use gas turbines (=biomass) for compensating heavy fluctuations, because those are online in a few minutes.