> In 2026, a sphere nine metres in diameter and weighing 400 tonnes will be submerged off the coast of California at a depth of 500 to 600 metres. It will have a storage capacity of 0.4 megawatt hours (400 kWh), enough energy to power an average household for several weeks.
What is this going to cost? From a quick search, Tesla Megapacks are now about $250/KWh. With battery costs still falling steadily, those might be considerably cheaper by the time the first 9m sphere hits the water.
And with all the recent anchor-dragging incidents, how many countries would be eager to have their energy storage located far off-shore?
Well concrete and steel are the major material components, fairly cheap, but pumps and turbines require high precision machining, fairly expensive. Lithium will probably be cheaper.
Probably not economical in current conditions, but worth doing to say it was done.
But the pumps and turbines are based on how much power it can process, not how much it can store. Most storage systems have a big downside in that the J that you use once a year costs just as much to store as the one you use every day. This perhaps permits putting a lot of air down there as reserve for when the sun doesn't shine for days on end. Lithium is non-viable for dealing with long periods of darkness (big storm on top of your collectors.) To ensure the lights stay on in the worst case you need weeks worth of storage.
This seems like a workable idea but just electric cranes lifting and lowering weights seems like a simpler approach and I think that has already been proposed.
My guess is there are many ways to balance the grid and the biggest is the utilities not wanting pay.
One thing to consider is the America power grid is in poor shape already and utilities are aiming to avoid modernizing. IE, adding storage to the grid would involve the double of cost of the actual balance equipment and the fixing the old equipment that needs fixing anyway. And utilities are looking avoid both cost.
I think when you lift a cubic meter of water by one meter at a depth of 500 meters, it is like lifting an entire column of 500 cubic meters. It's a huge multiplier effect.
Making a big hollow object that can withstand crushing is much cheaper than something that can contain high pressure. Achieving the same pressure difference on the surface wouldn't be possible with concrete alone.
There is no air involved. They are pumping water out, creating a vacuum. Water rushing back in generates power. No need to compress air or have a hose to the surface.
I'm curious why pumping the water out, creating a vacuum for water to rush in, is better than pumping air in for water to displace.
I had assumed it would be cheaper to have large underwater balloon connected by a hose to a pontoon, and use air. Rather than install and maintain at depth the pumps and a giant concrete sphere able to withstand that sort of pressure.
Have I got the economics wrong? Or is there an efficiency gain from dealing with a liquid rather than compressible gas?
> "The implementation of the pressure hose led to challenges in the dimensioning even for the small prototype, because conventional pressure hoses are usually designed for high internal pressures, but not for high external pressures. In addition, the installation becomes more demanding as the depth of the water increases. Greater water depths are associated with higher pressures, longer pressure hoses with greater friction losses and higher mechanical stresses. The more stable the hose wall is designed, the larger is its bending radius and weight, which in turn makes the installation more difficult."
They tried it with a surface hose and without, and the full charge-dischage cycle were roughly equivalent, so the simpler version without the surface hose is the recommended option.
The article says the spheres will be at 500-600m depth, so something like 50-60 atmospheres of pressure. Compressing air enough to displace water at that pressure would heat it to extreme temperatures which must be very inefficient even with some attempt to recover the heat. Water being incompressible can I assume be moved in and out of the sphere without any heating (and, therefore, energy loss).
One probably wouldn’t want to store energy by adiabatic compression like that. Instead one might try to approximate isothermal compression, perhaps by using multiple stages with heat exchangers to adjust the temperature between stages. IIRC there was a startup, Lightsail Energy, that had an idea of using a water mist as either a heat reservoir or a phase change medium, and I don’t know whether they ever got it working well.
I hadn't thought about heat. While the air will be hot, that heat was already there (the heat energy in 50 liters of air compressed into 1 liter, just like the compressor in a heat pump). I think this makes it even more viable, as the air pump/compressor can use that heat to partially power itself. A water pump can't do that.
Let's examine a vertical core through it: We have a top made of concrete, below that there is air, below that the ocean. This is a pure compression load on the concrete determined only by the amount of air volume, not the depth and thus the pressure the air is under. Make sure the mass of the concrete exceeds the mass of the water the device displaces.
In addition you need a skirt around the sides to keep the air from escaping. It experiences an outward force at the top and an inward force at the bottom, but both are once again based on the air column, not the pressure.
There's only one part of the system that actually must be beefy--the connection to the surface which will always be pressurized to the depth of the storage.
This is compressed air storage, but without the big waste that normally entails as the tank pressure changes. And without the big pressure vessel. I don't know what the round trip efficiency will be, compressed air usually is abysmal because a compressor will be designed for a given pressure and a turbine will be designed for a given pressure. Tank pressure below the compressor pressure is wasted energy, tank pressure above the turbine pressure is wasted energy. But this uses fixed pressure, they won't be mismatched.
This is the first mechanical system that I've seen proposed that sounds sensible. (Check the energy density for all the lifting approaches--abysmal.)
That doesn't match the described mechanism at all. The load on the concrete is absolutely related to depth, and immense. They describe one option with a connection to the surface, but that's not required. It's not really compressed air, more like pumped hydro with an small lower reservoir but a huge height difference.
That sounds like one important consideration — concrete has poor tensile strength.
Also, aside from being under water, this is functionally a lot like pumped hydro, which is an established, well developed technology. Compressed air energy storage has been tried, but as far as I know, it has never really been a success.
Is it compressed air? The article says, and I quote:
> To store energy, excess electricity is used to pump water out of the sphere, creating a relative vacuum.
To release energy, we open the valve: the water, pushed by the external pressure, rushes into the sphere and turns the turbine, producing electricity.
If you take air at ~50 bar and release it into the atmosphere and try to extract energy from the water replacing it spinning a turbine, you are throwing away most of the energy stored in the compressed air.
In any case, the math works out for pumped hydro. Ask Google:
(4.5m)^3 * 4/3 * pi * 500m * 1 g/mL * 9.8m/s^2 in kWh
This gives 520 kWh, which is consistent with the article’s claims.
edit: One can ballpark the storage capacity of compressed air storage, too. Assuming isothermal compression (which would be a nice ideal case and is not easy to achieve unless one compresses very slowly), the work is nRT • ln (volume ratio). nRT = PV measured at any point in the process, which is conveniently exactly the calculation above: 520kWh. For the pumped hydro model, I lazily assumed that they pumped all the way to vacuum, which is obviously wrong (some water would boil), but it makes almost no difference. But here we need to compare the actual pressures, and the pressure ratio (equivalently volume ratio) is around 50 between sea level and 500m deep. So multiply by ln 50 to get around 2MWh.
But that’s the actual work done in a perfect isothermal process. In the real world, the starting and ending states will be around the same temperature, but the process will be far from isothermal, so a good deal more than 2MWh will be used to compress the gas and a good deal less will come back out.
They tested two approaches: "Open ventilation" (hose to the surface) and "Closed ventilation" (no air sources).
"The most important finding was that an air-connection to the surface is not needed, reducing the technical effort significantly." [1]
"Charging requires more time and energy with closed ventilation. In return the higher pressure difference results in a higher turbine power and energy during the discharging phase…ventilation is not beneficial in this regard." [2]
This sounds like an actually reasonable thing, as long is there's a large exclusion zone around it, and they don't put a second one anywhere near it.... a catastrophic failure of one would take them all out, possibly killing all the marine life nearby, and damaging any submarines in the area.
The typical catastrophic failure would be a failure of "hull" integrity - i.e. the concrete sphere - at the point of relative vacuum. That could cause an implosion and collapse of the sphere.
That implosion would have a localised short-term effect on water currents as water moves to fill the void, which may affect wildlife in the very immediate area, at the point of implosion (e.g. fish swimming directly next to the sphere may be sucked in and knocked into or trapped by the collapsing concrete).
I can't envisage a failure mode which would cause an explosion or cascade to other spheres in the locality.
You talk about it like a gentle process. The proposed unit discussed in the wikipedia article stores about as much energy as 10 tons of TNT, and catastrophic failure could easily be similarly rapid. It's certainly something that would require further analysis (I don't have a good intuition for what the 'blast radius' would be underwater, especially in terms of threatening another unit).
The energy source in TNT is the chemical potential within the explosive, which is greater than the potential outside, so the reaction is explosive.
In this example, the energy source is the pressure of the water column, which is equipotent with the depth, and applying equally to all external faces of the sphere. It'll be dramatic for the sphere, but fairly inconsequential for anything not immediately adjacent.
I wonder what the plan is to deal with biofouling, it's a notoriously challenging thing to deal with on surface ships with barnacles etc clogging intakes. Maybe it's less of a problem on the sea floor?
@750 meters, there is no light, and anything that grows down there ,does it very slowly
but your right about anybody thinking this will not be a problem at smaller scales closer to the surface
Similar technology is bieng used to pump air into abandoned mines and caverns,
Also gravity batteries, where an electric train is
run up hill and the used as a generator on the downhill run, also used in abandoned mine shafts.
And somewhere there is a mine useing giant electric trucks that produce a surplus of electricity through regenerative braking, as the trucks are filled with rock, high on a mountain, and brake the whole way down, but climb back up empty, which
uses less energy than what was generated, not significant, other than as highlight of what is possible.
> In 2026, a sphere nine metres in diameter and weighing 400 tonnes will be submerged off the coast of California at a depth of 500 to 600 metres. It will have a storage capacity of 0.4 megawatt hours (400 kWh), enough energy to power an average household for several weeks.
What is this going to cost? From a quick search, Tesla Megapacks are now about $250/KWh. With battery costs still falling steadily, those might be considerably cheaper by the time the first 9m sphere hits the water.
And with all the recent anchor-dragging incidents, how many countries would be eager to have their energy storage located far off-shore?
Well concrete and steel are the major material components, fairly cheap, but pumps and turbines require high precision machining, fairly expensive. Lithium will probably be cheaper.
Probably not economical in current conditions, but worth doing to say it was done.
But the pumps and turbines are based on how much power it can process, not how much it can store. Most storage systems have a big downside in that the J that you use once a year costs just as much to store as the one you use every day. This perhaps permits putting a lot of air down there as reserve for when the sun doesn't shine for days on end. Lithium is non-viable for dealing with long periods of darkness (big storm on top of your collectors.) To ensure the lights stay on in the worst case you need weeks worth of storage.
This seems like a workable idea but just electric cranes lifting and lowering weights seems like a simpler approach and I think that has already been proposed.
My guess is there are many ways to balance the grid and the biggest is the utilities not wanting pay.
One thing to consider is the America power grid is in poor shape already and utilities are aiming to avoid modernizing. IE, adding storage to the grid would involve the double of cost of the actual balance equipment and the fixing the old equipment that needs fixing anyway. And utilities are looking avoid both cost.
I was actually wondering why the whole thing needed the complexity of being underwater and felt like I was just an idiot, so thanks.
From the article it sounds like the only reason to deal with the hassle of the water is so you can put these somewhere no one ever sees?
I think when you lift a cubic meter of water by one meter at a depth of 500 meters, it is like lifting an entire column of 500 cubic meters. It's a huge multiplier effect.
Making a big hollow object that can withstand crushing is much cheaper than something that can contain high pressure. Achieving the same pressure difference on the surface wouldn't be possible with concrete alone.
They specifically talk about putting them deep. There's only one thing to be gained by going deep: pressure.
It sounds like this is simply compressed air storage, except using seawater to contain it so you don't need a pressure vessel.
There is no air involved. They are pumping water out, creating a vacuum. Water rushing back in generates power. No need to compress air or have a hose to the surface.
I'm curious why pumping the water out, creating a vacuum for water to rush in, is better than pumping air in for water to displace.
I had assumed it would be cheaper to have large underwater balloon connected by a hose to a pontoon, and use air. Rather than install and maintain at depth the pumps and a giant concrete sphere able to withstand that sort of pressure.
Have I got the economics wrong? Or is there an efficiency gain from dealing with a liquid rather than compressible gas?
https://www.sciencedirect.com/science/article/am/pii/S235215... (PDF)
The article says the spheres will be at 500-600m depth, so something like 50-60 atmospheres of pressure. Compressing air enough to displace water at that pressure would heat it to extreme temperatures which must be very inefficient even with some attempt to recover the heat. Water being incompressible can I assume be moved in and out of the sphere without any heating (and, therefore, energy loss).
(I'm guessing, of course)
One probably wouldn’t want to store energy by adiabatic compression like that. Instead one might try to approximate isothermal compression, perhaps by using multiple stages with heat exchangers to adjust the temperature between stages. IIRC there was a startup, Lightsail Energy, that had an idea of using a water mist as either a heat reservoir or a phase change medium, and I don’t know whether they ever got it working well.
I hadn't thought about heat. While the air will be hot, that heat was already there (the heat energy in 50 liters of air compressed into 1 liter, just like the compressor in a heat pump). I think this makes it even more viable, as the air pump/compressor can use that heat to partially power itself. A water pump can't do that.
Tension vs compression loading probably.
There isn't much loading at all.
Let's examine a vertical core through it: We have a top made of concrete, below that there is air, below that the ocean. This is a pure compression load on the concrete determined only by the amount of air volume, not the depth and thus the pressure the air is under. Make sure the mass of the concrete exceeds the mass of the water the device displaces.
In addition you need a skirt around the sides to keep the air from escaping. It experiences an outward force at the top and an inward force at the bottom, but both are once again based on the air column, not the pressure.
There's only one part of the system that actually must be beefy--the connection to the surface which will always be pressurized to the depth of the storage.
This is compressed air storage, but without the big waste that normally entails as the tank pressure changes. And without the big pressure vessel. I don't know what the round trip efficiency will be, compressed air usually is abysmal because a compressor will be designed for a given pressure and a turbine will be designed for a given pressure. Tank pressure below the compressor pressure is wasted energy, tank pressure above the turbine pressure is wasted energy. But this uses fixed pressure, they won't be mismatched.
This is the first mechanical system that I've seen proposed that sounds sensible. (Check the energy density for all the lifting approaches--abysmal.)
That doesn't match the described mechanism at all. The load on the concrete is absolutely related to depth, and immense. They describe one option with a connection to the surface, but that's not required. It's not really compressed air, more like pumped hydro with an small lower reservoir but a huge height difference.
That sounds like one important consideration — concrete has poor tensile strength.
Also, aside from being under water, this is functionally a lot like pumped hydro, which is an established, well developed technology. Compressed air energy storage has been tried, but as far as I know, it has never really been a success.
The volumes involved are way too low for pumped hydro. This is compressed air--but without the big downsides it usually entails.
Is it compressed air? The article says, and I quote:
> To store energy, excess electricity is used to pump water out of the sphere, creating a relative vacuum. To release energy, we open the valve: the water, pushed by the external pressure, rushes into the sphere and turns the turbine, producing electricity.
If you take air at ~50 bar and release it into the atmosphere and try to extract energy from the water replacing it spinning a turbine, you are throwing away most of the energy stored in the compressed air.
In any case, the math works out for pumped hydro. Ask Google:
(4.5m)^3 * 4/3 * pi * 500m * 1 g/mL * 9.8m/s^2 in kWh
This gives 520 kWh, which is consistent with the article’s claims.
edit: One can ballpark the storage capacity of compressed air storage, too. Assuming isothermal compression (which would be a nice ideal case and is not easy to achieve unless one compresses very slowly), the work is nRT • ln (volume ratio). nRT = PV measured at any point in the process, which is conveniently exactly the calculation above: 520kWh. For the pumped hydro model, I lazily assumed that they pumped all the way to vacuum, which is obviously wrong (some water would boil), but it makes almost no difference. But here we need to compare the actual pressures, and the pressure ratio (equivalently volume ratio) is around 50 between sea level and 500m deep. So multiply by ln 50 to get around 2MWh.
But that’s the actual work done in a perfect isothermal process. In the real world, the starting and ending states will be around the same temperature, but the process will be far from isothermal, so a good deal more than 2MWh will be used to compress the gas and a good deal less will come back out.
[dead]
Article claims pumping water out to a "relative vacuum"
How do they do this without causing the pumps to cavitate?
Edit: here's the paper https://www.sciencedirect.com/science/article/abs/pii/S23521...
I assume that “relative” is doing some heavy lifting. I bet they’re bringing in air from the surface at a bit over 1atm to replace the removed water.
They tested two approaches: "Open ventilation" (hose to the surface) and "Closed ventilation" (no air sources).
[1] https://en.wikipedia.org/wiki/Stored_Energy_at_Sea[2] https://www.sciencedirect.com/science/article/am/pii/S235215... (PDF)
This sounds like an actually reasonable thing, as long is there's a large exclusion zone around it, and they don't put a second one anywhere near it.... a catastrophic failure of one would take them all out, possibly killing all the marine life nearby, and damaging any submarines in the area.
The typical catastrophic failure would be a failure of "hull" integrity - i.e. the concrete sphere - at the point of relative vacuum. That could cause an implosion and collapse of the sphere.
That implosion would have a localised short-term effect on water currents as water moves to fill the void, which may affect wildlife in the very immediate area, at the point of implosion (e.g. fish swimming directly next to the sphere may be sucked in and knocked into or trapped by the collapsing concrete).
I can't envisage a failure mode which would cause an explosion or cascade to other spheres in the locality.
You talk about it like a gentle process. The proposed unit discussed in the wikipedia article stores about as much energy as 10 tons of TNT, and catastrophic failure could easily be similarly rapid. It's certainly something that would require further analysis (I don't have a good intuition for what the 'blast radius' would be underwater, especially in terms of threatening another unit).
The energy source in TNT is the chemical potential within the explosive, which is greater than the potential outside, so the reaction is explosive.
In this example, the energy source is the pressure of the water column, which is equipotent with the depth, and applying equally to all external faces of the sphere. It'll be dramatic for the sphere, but fairly inconsequential for anything not immediately adjacent.
The mantis shrimp kills prey using cavitation. I'm not sure the shock wave from such an energetic implosion would be "fairly inconsequential".
The implosion would result in a subsequent cavitation event, which could have significant energy gain.
The wikipedia article has a lot more detail on the proposal: https://en.wikipedia.org/wiki/Stored_Energy_at_Sea
The target depth is 750m (compared to 100m for the pilot in Lake Constance).
I wonder what the plan is to deal with biofouling, it's a notoriously challenging thing to deal with on surface ships with barnacles etc clogging intakes. Maybe it's less of a problem on the sea floor?
@750 meters, there is no light, and anything that grows down there ,does it very slowly but your right about anybody thinking this will not be a problem at smaller scales closer to the surface Similar technology is bieng used to pump air into abandoned mines and caverns, Also gravity batteries, where an electric train is run up hill and the used as a generator on the downhill run, also used in abandoned mine shafts. And somewhere there is a mine useing giant electric trucks that produce a surplus of electricity through regenerative braking, as the trucks are filled with rock, high on a mountain, and brake the whole way down, but climb back up empty, which uses less energy than what was generated, not significant, other than as highlight of what is possible.
Is there any reason to need a concrete sphere? Couldn't a robust, durable flexible bag do the trick?
My hydroflask, when compressed, will push water out :)
If you pump water out of a flexible bag you are not storing energy, just shrinking the bag.
20-year maintenance intervals for something pumping seawater? Impressive. Way better than a chemical battery.
Seawater may well be what derails this project. The test site was in a fresh water lake.
I don't think you're pumping sea water - I think you're pumping air.
They're pumping water, see technical paper.
Cool idea. It’s like pumped water storage but backwards
[dead]