Originally Posted by
Dairyground
DC is used for long-distance (tens of miles or more) under(salt)water power links such as those between England and France. I have not investigated the rationale for choosing DC, but assume that an AC link would suffer significant losses because of the conductivity of the surrounding medium.
As others have hinted at, it's the capacitance between the undersea cable and the earth that makes AC a problem and DC worthwhile. The capacitance causes transmission losses for AC.
However, to make a DC cable efficient enough to be worthwhile requires very high voltages. Getting the balance right between the working voltage and the reliability of the cable's insulator is the tricky part. In cables such as that, any hint of a breakdown in the insulator is immediately, and explosively, disasterous. Especially as what leaks in is salt water.
It's basically the same argument in favour of using higher pressure hydraulics on an aircraft. Higher pressures are desirable (smaller pipes, more efficiency) but more difficult (preventing leaks).
Back in the 1960s when the first cable was laid, the AC-DC conversion was done with mercury-arc rectifiers; these were terrifying things that would scorch your skin off (they're prodigious UV emitters), and could explode in a cloud of boiling mercury droplets if anything went wrong. Suffice to say it's all solid state electronics these days.
AFAIK the cable has almost never conducted power from the UK to France (the original intention back in the coal rich 1960's UK). It's always flowed the other way (lots of nuclear power).
Superconducting Grid
Back in the 1960s the Central Electricity Generating Board researched the use of superconductors for the UK national grid. They concluded that it would be worthwhile, but would require switching the grid over from AC to DC; superconductors don't eliminate reactive AC loses. That conversion would simply have been overwhelmingly expensive, which is why they didn't do it.
Grid lines strung between pylons tens of meters above the ground are far less lossy at AC than a cable buried in a trench, but it's still a significant source of transmission loss even then. And in case you've ever wondered why grid cables are in bundles of 4, that's to approximate a tube and take some advantage of the skin effect that occurs in AC transmission. 50Hz is a reasonable compomise between AC and DC loses. A lot of power generators actually work at 3000Hz, and the power station has to do a conversion to 50Hz.
Explosively Formed High Temperature Superconductor Grids
These days the attraction of high temperature super conductors is high; liquid nitrogen is all that is required for cooling. Trouble is they're all ceramic; they don't bend, so cannot be rolled out into a trench. There is a technique that has been tried (successfully
AFAIK!) to explosively form the ceramis in-situ; a copper tube is filled with the powdered ingredients, and wrapped in a layer of explosives. This is then rolled out into the trench. The explosives are set off, boom! Et voila, one long rod of high temperature superconductor is formed. I think this technique was used somewhere in the US to connect a (hydro?) generating plant up to the grid without putting a string of pylons through some area of outstanding natural beauty.
I don't think we'll be using that in aircraft any time soon...
DC DC Conversion
Mention has been made of efficient DC-DC conversion being a thing these days, and indeed it is (90% is possible). However it involves converting the DC to AC (say, 10kHz), putting it through a small transformer (hence the 10kHz), and recitfying it back to DC. This is what a switched mode power supply does.