OK, my day job is doing HV engineering, not transmission, but high energy stuff.
The author did something kind of equivalent to:
"If we scale a GPU clock to 75 Petahertz, we can make datacenters that fit in bed rooms! Here are the FLOPS calculations to prove it!"
This whole thing is so crazy I don't know where to begin. I applaud the author for jumping into a new subject, but there is _way_ more complexity here than laid out. HV is very difficult to penetrate too because there really isn't much info available online about it.
Those initial dielectric strength numbers are definitely off (maybe they used Wikipedia, which references a value from a 1920 physics book). As from what I can find fused silica has a dielectric strength around 50-100MV/m, which is taken from the AC figure and doubled to get the DC figure (which is fairly typical). Also these numbers are extrapolated, and dielectrics often have non-linear properties. The testers used to get these figures can be a little fickle, and HV is always fickle.
On top of that, in actual HV system design, you really need to be using 25% of the actual dielectric strength for any kind of reliability. So the practical strength of fused silica would ultimately be around ~20MV/m. Which pretty much kills the whole idea right there. Never mind that a single fracture or dielectric breakdown anywhere in the whole glass sheath would require the entire thing to be replaced. Spoiler: You cannot patch HV dielectrics. Trust me, I and many others have tried.
Some other hurdles would be dealing with the insane parasitics, which the author didn't even mention, but are one of if not the most limiting factor in transmission. HVDC lines can have up to 10% ripple, which for the author would be 1.4MV of high frequency ripple. And sea water is conductive! You are basically building a massive capacitor with sea water! The losses would be enormous.
And I don't even want to think about the electronics...14MV is so insane I cannot fathom anything that would be able to reliably handle it. 1MV is already nuts. 800kV is the highest in the world, and that is kinda just a flex.
I swoop in on something like this looking for the first obvious error in units/arithmetic/materials that renders the whole thing ludicrous, but the author has a spreadsheet and it looks like the units are about right. It's an absurdly cheap cable in terms of materials to transmit 10 GW across an ocean. The main things that render it dubious as a practical matter:
- I don't know if operating at 14 million volts is achievable in terms of converter stations. Today's highest voltage HVDC projects operate at 1.1 megavolts and it took years of development to get there from 0.6 megavolts.
- The mechanical practicality of thousands of kilometers of silica clad aluminum. There's a big mismatch in coefficients of thermal expansion and silica is brittle.
Still, this appears to be facially valid in scientific terms if not in engineering terms. That's impressive! It's a really thin intercontinental cable carrying a lot of power.
The whole thing reminded me of this discussion here from 3 years ago:
It has rough numbers for a globe-spanning HVDC cable on the order of a meter in diameter (assumes voltages more like present day commercial HVDC, much thicker conductor to compensate).
> The cable, if snagged by a ship anchor, would catastrophically fail. Not only would it snap, but the internal stresses would propagate the crack along the entire length.
I admire that the author wrote this sentence and continued with the thought experiment anyway
Not quite true. Glass optical fibre is reasonably flexible. More so than many coaxial cables. Just don't go below its minimum bend radius, as it is brittle and will snap.
Glass insulated power cables might work, provided the glass layer is thin enough and its band radius isn't exceeded. One can imagine a cable insulated with many thin layers/strips of glass, which have some movement relative to each other. Multiple layers of insulation is normal practise with plastic insulation, as the failure mode is typically pinholes in the insulation and multiple layers reduced the probability of pin holes going all the way through.
Biggest problem might be a conductor with decent diameter will put a lot of stress on the interior and exterior of a bend. Some ides:
* A multi-standed conductor with each individual conductor insulated. Maybe have high voltage in the interior stands and have a radial voltage gradient (to zero) across the outer strands so no one thin layer of glass is taking the full electric field?
* Could a conductor be insulated with a woven/stranded insulating layer? One can imagine many layers of extremely fine glass fibre finished off with an enclosing layer of something else to keep everything in place. Sort of like a glass insulated coaxial cable.
Stepping back from the technical question of how to lay HVDC undersea, a globally connected power grid seems like a major win for renewable energy. There are a lot of places you’d like to put power plants, and having the infrastructure in place to be able to sell that energy makes it immediately more feasible. We could put nuclear plants anywhere. Solar plants across the Sahara. It would increase the energy available to developing countries without tempting them with dirty fossil fuel plants.
The importance of repairability is underestimated here. All new infrastructure must be built under assumption that there will be multiple attempts at sabotaging it by actors of various level, and multi-megavolt unrepairable cables that can be fully disabled by one smallish unmanned sub don’t win here at all.
A stack of optically powered 15kV mosfets, to get to 14MV, sounds absurdly awesome. 933+ mosfets that you're trying to drive in series, egads. But neat weird idea.
> A 15 kV SiC MOSFET gate drive with power over fiber based isolated power supply and comprehensive protection functions
I distantly remember reading about someone stress testing a submarine drone tether at higher than rated voltages, seeing what practical voltage they could get out of it. I distantly recall there being a lot of concern about like corona arching or something with the sea water? That was a fun paper. I don't ever if it was only because they exceeded the insulation value, but I feel like there were some notable challenges to running high voltages in salt water that I'm not quite remembering.
HVDC cables are kind of an often overlooked solution to net zero. Moving power over long distances, across timezones is kind of a super power. The main obstacle to scaling this from a few GW to tens/hundreds of GW is cost. Just by laying more cables can you increase capacity between regions and their ability to share excess power to each other. But each cable is a multi billion dollar project. Which means that there is only a little bit of capacity to move power around but not a lot. For example Europe can import a few GW of African solar in the middle of the winter. But it could probably need hundreds when it is dark and not windy there.
Likewise cross Atlantic cables have been talked about but so far don't exist. Same with getting power from the East coast US to the West coast and vice versa. The east coast goes dark while the west coast is still producing lots of solar. And in the morning on the west coast, it's afternoon on the east coast. There is a bit of import/export between California (solar) and Canada (wind / hydro). But it could be much more.
Cables have another important function: they can be used to charge batteries. Batteries allow you to timeshift demand: e.g. charge when the sun is out, discharge when people get home in the evening. And off peak, the cables aren't at full capacity anyway meaning that any excess power can easily be moved around to charge batteries locally or remotely. Renewables, cables and batteries largely remove the need for things like nuclear plants.
Yes it gets dark and cloudy sometimes but those are local effects and they are somewhat predictable. And if the wind is not blowing that just means it is blowing elsewhere. Wind flows from high pressure to low pressure areas. Globally, there always are high and low pressure areas. If anything, global warming is causing there to be more wind, not less. So, global wind energy production will always maintain a high average even if it drops to next to nothing locally. Likewise, global solar production moves around with the sun rise and sun set and seasons but never drops to zero everywhere. If it's night where you are, it isn't on the other side of the planet. If it's winter where you are, it isn't at -1 * your latitude.
If long distance cables get cheap and plentiful, that's a really big deal because this allows for moving around hundreds of gwh of power. HVDC allows doing that over thousands of kilometers across oceans, timezones, and continents. Cheaper HVDC lowers the cost of that power.
It's an interesting take to be sure. I suspect that the lack of flexibility is going to be the real killer.
You'd probably have to build offshore platforms on either side to bring the cables up and terminate them and now that's a nightmare, saltwater/salty air and electronics don't mix well.
Or you're going to have to trench very deeply for the first few miles.
Either way you're stuck with something that really doesn't want to be bent.
I think the "glass is great insulation" is a good insight and perhaps a composite glass fiber/polymer sheath would really increase the V/m without the brittleness.
This is the kind of transmission line design I've seen proposed for use on the Moon - where hydrocarbons are basically nonexistent, but aluminium and silicon are abundant.
Glass insulated cable sounds like a tech that should be prototyped on smaller scales - and could be somewhat useful on those smaller scales.
> The cable, if snagged by a ship anchor, would catastrophically fail. Not only would it snap, but the internal stresses would propagate the crack along the entire length.
I can’t this writeup seriously with comments like this. There is no mention of any attempt to calculate the allowable bend radius. Also, quenching a glass tube in a continuous process? Does that work?
The Moore-like fall of solar+battery costs took away solar satellites, solar convection plants, submarine power cables and (widely deployed though) sun tracking hardware. Labour costs are becoming a bigger proportion so some installations plop panels on the ground than slant them to south (in northern hemisphere).
It's true that you can switch very high voltages with MOSFETs in series. But the next step after switching is a transformer that needs to handle 14 MV between the primary and lower-voltage secondary winding. I don't think anyone has built something like that before. Given the dielectric strength of transformer oil, the primary windings need to be 500mm away from both the secondary windings and the core, which seems like it'd be hard to do while getting good inductive coupling.
It’s interesting. I think the real way to do this is gradually scale up. Crossing the ocean is hard mode. Instead start by something much shorter and land based. Then you at least have a stable platform to work on and can focus on the other hard problems
If I (and only I) owned such a cable from Europe to the US, how much money could I make by buying cheap solar energy from the bright side and selling it to the dark side of the Atlantic?
First thought:
10 GW * $0.03/kWh 4 hours/day = $1.2Mio per day [0]
1. The technical solution relies heavily on fantasy.
2. It is not needed. We have no significant power transmission across the low lying fruit of continental America or Eurasia, and those lines are built! Why bother crossing an ocean?
3. Why not cross Greenland and the North Sea and its islands? Under sea cables are expensive.
Continuous melted silica coating is fine, but how does one account for all the movement, bends and vagaries of the high seas, especially for something that is so brittle?
I wonder if the glass sheath could be replaced with bundled glass fibers in a dielectric gel? Would that cross section allow for a much greater distance for current to trace through the gel? Seems like maybe it would give a 2x advantage, or maybe glass ribbons could be made instead for a micro braided insulation?
For the glass to be the insulator we need, I'm assuming the author envisions a solid tube, with no airgaps (can't do fibre braid as that would allow gaps which means loss of insulation, or you'd need oil to fill the gaps.)
This means huge bend radius in the order of hundreds of meters. Not only that but laying it on the ocean bed would require trenching and full support to stop localised bending.
Now to the manufacture:
> The cable is then quenched in water to surface harden it, before it moves out of the back of the ship and falls to the ocean floor over a length of many kilometers (due to very low curve radius).
So that'll cause the tube to break. Glass builds up hige amounts of stresses when it cools down quickly (see prince ruperts drop) so needs an annealing step. ( https://en.wikipedia.org/wiki/Annealing_(glass) )
Moreover changes in temperature mean that using aluminum is probably going to cause the glass to shatter when the temperature changes. which means that you either need https://en.wikipedia.org/wiki/Kovar or somehow make expansion joints every n meters.
Finally that cable is going to be heavy, so unless you make it around the same densisty as salt water, it'll have so much weight it'll snap as soon as you try and dump it into the sea.
apart from that, looks good. well apart from the units are wrong to start with.
TLDR:
you'd need 5x the width of Polyethylene to achieve the same level of insulation at high voltages. but as silica tube doesn't bend and shatters really easily, cant be transported and has a slow extrusion rate, it seems logical to just use PE.
Everything looks nice but something very important was not considered in all of this.
High voltage and high current means Z-pinch - the conductor itself is going to compress itself, thus resulting in basically delaminating from the glass sheathing. This is why we have rubber/petroleum-based flexible sticky insulators on cabling like that, it can somewhat flex/shrink with the conductor and is more likely to stay attached and less likely to get damaged.
Just transmit laser power down fiber optics at that point. Either way you're going to need semiconductor switching (it's IGBTs all the way down baby!) nothing electromechanical is going to handle that kind of load.
I watched a video recently that talked about how China is really the only country to have developed and built UHVDC power transmission. Some look at this and say how it's a failure of everyone else. My immediate thought was: "this solves aproblem only China has" and that turned out to be correct.
China produces most of its power in the west of the country between solar farms, the Three Gorges Dam and so on. Most of the population is 2000 miles away in the east of the country. For over a billion people, the cost of more efficient long-distance transmission make economic sense.
Someone asked "could Australia do this to transmit solar power from the West coast to the east coast in peak hours?". Technically? Yes. Practically? No. Why? It's obviously expensive with far fewer people but also all that space in between is uninhabited. So if you ever need to maintain it (which you will) you have to send people out into the wilderness to do it. China doesn't have that problem because it's not really unpopulated anywhere, at least not to the scale Australia is.
My point here is that you should always ask for something like this "what problem does it solve?" And the answer for more efficient long-distance power transmission is "almost nobody".
I think power grids are going to go in the other direction and become increasingly localized rather than nationalized.
Worldwide power grid with glass insulated HVDC cables
(omattos.com)126 points by londons_explore 22 hours ago | 122 comments
Comments
The author did something kind of equivalent to:
"If we scale a GPU clock to 75 Petahertz, we can make datacenters that fit in bed rooms! Here are the FLOPS calculations to prove it!"
This whole thing is so crazy I don't know where to begin. I applaud the author for jumping into a new subject, but there is _way_ more complexity here than laid out. HV is very difficult to penetrate too because there really isn't much info available online about it.
Those initial dielectric strength numbers are definitely off (maybe they used Wikipedia, which references a value from a 1920 physics book). As from what I can find fused silica has a dielectric strength around 50-100MV/m, which is taken from the AC figure and doubled to get the DC figure (which is fairly typical). Also these numbers are extrapolated, and dielectrics often have non-linear properties. The testers used to get these figures can be a little fickle, and HV is always fickle.
On top of that, in actual HV system design, you really need to be using 25% of the actual dielectric strength for any kind of reliability. So the practical strength of fused silica would ultimately be around ~20MV/m. Which pretty much kills the whole idea right there. Never mind that a single fracture or dielectric breakdown anywhere in the whole glass sheath would require the entire thing to be replaced. Spoiler: You cannot patch HV dielectrics. Trust me, I and many others have tried.
Some other hurdles would be dealing with the insane parasitics, which the author didn't even mention, but are one of if not the most limiting factor in transmission. HVDC lines can have up to 10% ripple, which for the author would be 1.4MV of high frequency ripple. And sea water is conductive! You are basically building a massive capacitor with sea water! The losses would be enormous.
And I don't even want to think about the electronics...14MV is so insane I cannot fathom anything that would be able to reliably handle it. 1MV is already nuts. 800kV is the highest in the world, and that is kinda just a flex.
- I don't know if operating at 14 million volts is achievable in terms of converter stations. Today's highest voltage HVDC projects operate at 1.1 megavolts and it took years of development to get there from 0.6 megavolts.
- The mechanical practicality of thousands of kilometers of silica clad aluminum. There's a big mismatch in coefficients of thermal expansion and silica is brittle.
Still, this appears to be facially valid in scientific terms if not in engineering terms. That's impressive! It's a really thin intercontinental cable carrying a lot of power.
The whole thing reminded me of this discussion here from 3 years ago:
https://news.ycombinator.com/item?id=31971039
It has rough numbers for a globe-spanning HVDC cable on the order of a meter in diameter (assumes voltages more like present day commercial HVDC, much thicker conductor to compensate).
I admire that the author wrote this sentence and continued with the thought experiment anyway
Not quite true. Glass optical fibre is reasonably flexible. More so than many coaxial cables. Just don't go below its minimum bend radius, as it is brittle and will snap.
Glass insulated power cables might work, provided the glass layer is thin enough and its band radius isn't exceeded. One can imagine a cable insulated with many thin layers/strips of glass, which have some movement relative to each other. Multiple layers of insulation is normal practise with plastic insulation, as the failure mode is typically pinholes in the insulation and multiple layers reduced the probability of pin holes going all the way through.
Biggest problem might be a conductor with decent diameter will put a lot of stress on the interior and exterior of a bend. Some ides:
* A multi-standed conductor with each individual conductor insulated. Maybe have high voltage in the interior stands and have a radial voltage gradient (to zero) across the outer strands so no one thin layer of glass is taking the full electric field?
* Could a conductor be insulated with a woven/stranded insulating layer? One can imagine many layers of extremely fine glass fibre finished off with an enclosing layer of something else to keep everything in place. Sort of like a glass insulated coaxial cable.
> A 15 kV SiC MOSFET gate drive with power over fiber based isolated power supply and comprehensive protection functions
https://ieeexplore.ieee.org/document/7468138
I distantly remember reading about someone stress testing a submarine drone tether at higher than rated voltages, seeing what practical voltage they could get out of it. I distantly recall there being a lot of concern about like corona arching or something with the sea water? That was a fun paper. I don't ever if it was only because they exceeded the insulation value, but I feel like there were some notable challenges to running high voltages in salt water that I'm not quite remembering.
https://news.ycombinator.com/item?id=42513761 ("Undersea power cable linking Finland and Estonia suffers damage", 112 comments)
It's been half a year and it still[0] hasn't been fixed yet.
How does anyone, really, imagine building planetary infrastructure where a trivial amount of asymmetric warfare can take the whole thing down?
[0] https://yle.fi/a/74-20164957 ("Fingrid said that the EstLink 2 connection should be back online on June 25, earlier than expected")
Likewise cross Atlantic cables have been talked about but so far don't exist. Same with getting power from the East coast US to the West coast and vice versa. The east coast goes dark while the west coast is still producing lots of solar. And in the morning on the west coast, it's afternoon on the east coast. There is a bit of import/export between California (solar) and Canada (wind / hydro). But it could be much more.
Cables have another important function: they can be used to charge batteries. Batteries allow you to timeshift demand: e.g. charge when the sun is out, discharge when people get home in the evening. And off peak, the cables aren't at full capacity anyway meaning that any excess power can easily be moved around to charge batteries locally or remotely. Renewables, cables and batteries largely remove the need for things like nuclear plants.
Yes it gets dark and cloudy sometimes but those are local effects and they are somewhat predictable. And if the wind is not blowing that just means it is blowing elsewhere. Wind flows from high pressure to low pressure areas. Globally, there always are high and low pressure areas. If anything, global warming is causing there to be more wind, not less. So, global wind energy production will always maintain a high average even if it drops to next to nothing locally. Likewise, global solar production moves around with the sun rise and sun set and seasons but never drops to zero everywhere. If it's night where you are, it isn't on the other side of the planet. If it's winter where you are, it isn't at -1 * your latitude.
If long distance cables get cheap and plentiful, that's a really big deal because this allows for moving around hundreds of gwh of power. HVDC allows doing that over thousands of kilometers across oceans, timezones, and continents. Cheaper HVDC lowers the cost of that power.
You'd probably have to build offshore platforms on either side to bring the cables up and terminate them and now that's a nightmare, saltwater/salty air and electronics don't mix well.
Or you're going to have to trench very deeply for the first few miles.
Either way you're stuck with something that really doesn't want to be bent.
I think the "glass is great insulation" is a good insight and perhaps a composite glass fiber/polymer sheath would really increase the V/m without the brittleness.
Glass insulated cable sounds like a tech that should be prototyped on smaller scales - and could be somewhat useful on those smaller scales.
I can’t this writeup seriously with comments like this. There is no mention of any attempt to calculate the allowable bend radius. Also, quenching a glass tube in a continuous process? Does that work?
First thought: 10 GW * $0.03/kWh 4 hours/day = $1.2Mio per day [0]
I am not sure about my assumptions...
[0]: https://www.wolframalpha.com/input?i=10+GW+*+%24+0.03%2FkWh+...
1. The technical solution relies heavily on fantasy.
2. It is not needed. We have no significant power transmission across the low lying fruit of continental America or Eurasia, and those lines are built! Why bother crossing an ocean?
3. Why not cross Greenland and the North Sea and its islands? Under sea cables are expensive.
4. Why not cross the Bearing Strait?
500 MV/m is 0.5 MV/mm, so it's 300x worse insulator than XLPE plastic per article.
Would be a bummer if we build the worldwide insulated network, only to find out it's not insulated enough ツ)_/¯
For the glass to be the insulator we need, I'm assuming the author envisions a solid tube, with no airgaps (can't do fibre braid as that would allow gaps which means loss of insulation, or you'd need oil to fill the gaps.)
This means huge bend radius in the order of hundreds of meters. Not only that but laying it on the ocean bed would require trenching and full support to stop localised bending.
Now to the manufacture:
> The cable is then quenched in water to surface harden it, before it moves out of the back of the ship and falls to the ocean floor over a length of many kilometers (due to very low curve radius).
So that'll cause the tube to break. Glass builds up hige amounts of stresses when it cools down quickly (see prince ruperts drop) so needs an annealing step. ( https://en.wikipedia.org/wiki/Annealing_(glass) )
Moreover changes in temperature mean that using aluminum is probably going to cause the glass to shatter when the temperature changes. which means that you either need https://en.wikipedia.org/wiki/Kovar or somehow make expansion joints every n meters.
Finally that cable is going to be heavy, so unless you make it around the same densisty as salt water, it'll have so much weight it'll snap as soon as you try and dump it into the sea.
apart from that, looks good. well apart from the units are wrong to start with.
TLDR:
you'd need 5x the width of Polyethylene to achieve the same level of insulation at high voltages. but as silica tube doesn't bend and shatters really easily, cant be transported and has a slow extrusion rate, it seems logical to just use PE.
as I understand it, nobody is doing cable laying this way - and this dream of 14MV cable is kinda hinges on that
High voltage and high current means Z-pinch - the conductor itself is going to compress itself, thus resulting in basically delaminating from the glass sheathing. This is why we have rubber/petroleum-based flexible sticky insulators on cabling like that, it can somewhat flex/shrink with the conductor and is more likely to stay attached and less likely to get damaged.
Just transmit laser power down fiber optics at that point. Either way you're going to need semiconductor switching (it's IGBTs all the way down baby!) nothing electromechanical is going to handle that kind of load.
China produces most of its power in the west of the country between solar farms, the Three Gorges Dam and so on. Most of the population is 2000 miles away in the east of the country. For over a billion people, the cost of more efficient long-distance transmission make economic sense.
Someone asked "could Australia do this to transmit solar power from the West coast to the east coast in peak hours?". Technically? Yes. Practically? No. Why? It's obviously expensive with far fewer people but also all that space in between is uninhabited. So if you ever need to maintain it (which you will) you have to send people out into the wilderness to do it. China doesn't have that problem because it's not really unpopulated anywhere, at least not to the scale Australia is.
My point here is that you should always ask for something like this "what problem does it solve?" And the answer for more efficient long-distance power transmission is "almost nobody".
I think power grids are going to go in the other direction and become increasingly localized rather than nationalized.