Hacker News new | ask | show | jobs
by popol12 1519 days ago
Sounds inefficient How big are the loss on such a high distance connection ?
5 comments

High Voltage Direct Current (HVDC) is much more efficient at transmission over long distances than AC. It can throw off your intuition on these things.

In China, there is a 3300 km line like this moving 12GW of power. This is a little bit longer, but not by much.

https://en.wikipedia.org/wiki/High-voltage_direct_current

The reason AC is often believed to be more efficient is because higher voltages of either type are more efficient (less loss to electrical resistance), and it's much easier to step voltage up and down using AC.
Transformers are very efficient and easy to make for sure, and can ‘easily’ step up to millions of volts.

Longer distance runs of AC can have a lot of losses though due to inductance, especially when they’re close to large conductive masses (like seawater or earth).

DC requires expensive conversion equipment to get back to AC for the grid.

AC is cheap for short runs. Usually AC vs DC high voltage starts to be equally cost effective around the several hundred miles of buried lines, and a clear winner around the 1000 mile range (efficiency and capex wise).

HVDC can also allow grid balancing when disparate grids are connected (different frequencies and voltages for instance).

Another advantage of DC over AC is that DC doesn't have to be synchronized wrt frequency, phase, or power-factor.
The line you're talking about is part of China's UHVDC grid, which is all on land. The Chinese have invested more than $100 billion into this over the past twenty years, because they have to, all their power generation is far from their cities.

The longest submarine HVDC line today is something like 600 km. There are no UHVDC submarine lines.

3.5 % per 1,000 km, according to https://en.wikipedia.org/wiki/High-voltage_direct_current, which comes out to around 10 %.
14% (1,035^3,8)
Delivered power is 0.965^3.8 of input power, 87.3%, so 12.7% loss.

Using the 3.5% per 1000 km factor. Plus, of course, conversion and voltage stepping losses at both ends.

Right.
If you do not really know the loss why does it "sound inefficient"? I am pretty sure they've done all the due diligence and if the project is a go then the losses are acceptable.

I am more interested (again if the project is a go) what happens if / when some "friendly" people will mess with the cable.

Worth noting the project isn't even funded yet.

https://doomberg.substack.com/p/20000-volts-under-the-sea?s=...

Wikipedia quotes losses of 3.5% per 1,000 km.
3.5% per 1000km.