AC to DC Conversion and Long Distance Power Transmission

I first read of this work on CR4 and Globalspec soon after joining. It is apparently the way to go when interfacing different quality grid systems, and is apparently used to get phase and hertz power of different quality between the Mexican and US Grid systems.

I am sorry I cannot refer you quickly to the blog thread where this was all explained, but it must have been sometime in 2007, for it was early in my participation on this site and in this forum.

I believe the main reason is to be able to take the raw DC power, and make it run through the partner grid at 60 hertz three phase instead of 220 50 hertz, single phase, or whatever.

All mature grids I have worked with in my niche gave me 123 volts out the panel box at 60 hertz very nicely, except for the boxes in Pennsylvania, which were whatever they gave you, mature or not, scary as hell.

There are others much better than myself who can give you a definitive answer, but that is what I decided the better experienced, like Sparkstation attempted to impart to me on your subject.

I was also under the impression that the Hi Voltage DC transmissions did get to be less sensible if they ran too long, and thought they were best perfected for something like 20 to 50 miles, though 100 kilometers is what, 68 miles or so?

Again I want to emphasis that this is an area where I have not had the pleasure of hands on experience, and will defer to those that have.

another advantage of using DC is that there is no problem with the overvoltage occuring when the powerline is partially loaded.

In long AC transmission if partially loaded, at the end of the line there are problems with overvoltage of about 2-3 times the rated voltage of the powerline due to capacitive effect. The insulators in this case needs to be slightly higher than √2.

Hello aikhh:

You hit the nail on the head when you said power lost during transmission. I read an article several years ago about the Russians doing an ultra high-voltage DC transmission system, unfortunately I did not remember everything out of that article, so I had to cheat and do a little Google.

For short distances the conversion costs from AC/DC and back out way the gains in efficiency, however for longer distances particularly underground or underwater where AC losses tend to be higher is more economical to transmit electricity in the form of high-voltage DC and convert it back to AC.

The following link will explain it far better than I ever could.

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

The UK and France, among other places, are linked with HVDC cables, as the UK's AC power grid and France's AC power grid are not synchronised. Use of DC through the link, and converting it at the far end, overcomes this issue.

Australia does it between the mainland and Tasmania and an article was written in the Australian electronics magazine, Silicon Chip. I think it was about 18 months ago.

Tony

DC also doesn't have problems with power factor changes due to the self inductance of the line.

As others have mentioned, you need a fairly long line and considerable power to justify changes from AC/DC then DC/Ac at the ends of the line.

The inherant inductance of the line, in long distance spans, is one of the primary reasons for HVDC transmission. When you have a VERY long cable, the inductive reactance becomes noticably high. Every time the AC wave form passes through zero you have just discharged that megasize inductor, now you have to try and recharge it again. As you can imagine, from basic electrical theorem, it takes a finite amount of power to overcome this inductive reactance. Usually these losses are just considered a necesary evil (kind of a cost of doing business) but when you are dealing with a super long intertie, these losses can be so high that they justify the expense, and efficiency loss, of the rectification and subsequent DC to AC inversion, involved with an HVDC interlink.

On long AC transmission lines, it's actually capacitive reactance rather than inductive but nevertheless it requires the supply of reactive power by the generators in order to maintain power factor and voltage stability. Of course, generation owners do not like to provide excess reactive power because they generally receive no compensation for it. Transmission lines are inherently capacitive while the loads are predominantly inductive. On long AC transmission lines it is often necessary to install reactive compensation in the form of series compensation capacitor banks for the same reasons and to reduce resistive power losses (which represent real unmetered power that must be provided). That's the primary reason DC transmission is most efficient for long EHV and UHV transmission lines.

Contrary to an earlier post, stranded conductors are used for DC transmission lines and, while skin effect is a consideration for AC transmission, the magnitude depends upon the conductor construction and the number of stranded layers in the conductor.

Me too, since from what the others say, it works better over longer distances, than shorter.

The skin effect you mention is caused by inductance (see Lenz's Law) not capacitance. The eddy currents are in direct opposition to the current that causes them and this equals loss.

I thought that I would look at an example (it only confirms what you write):

Basslink is a monopolar link run at 400-kV with 1250-Amps nominal maximum current - so the minimum load impedance would be equivalent to 320-Ohms. I imagine the pole is screened with an Earth conductor - and for 320 Ohm the inner diameter of the screen would need to be at least 210 times the diameter of the inner. The effect on total cable diameter means that achieving this impedance of would blow the costs out of the water (!).

I also noticed that things would be even worse for a mismatched link as long as 1-km, because this is approaching the resonant 1/4-wavelength; resonant effects would more than double the apparent capacitance of the cable. So I doubt that AC would be practical over such a link even if it was possible to synchronise the networks.

(DC still has it's problems, of course)

Fyz

"1-km, because this is approaching the resonant 1/4-wavelength."

1 kilometer? At what frequency?

Jon

My tryping - the given distance was 1000-km, see below

Hi fyz

If you'd remembered to log in, I'd have given you a GA for that one. The simple example said more than a multitude of words.

Of course, at 50 Hz the 1/4 wavelength is actually 1500km, while for 60 Hz it is 1250km, but what's a few O's between friends!

Regards

Sceptic

Thanks. Presentation problems as ever - I was of course calculating for the specified distance of 1000-km = 1 Mm (megametre)

In this case the calculations were based on "approaching" being 2/3 of 1500-km (i.e. 50-Hz and air-spaced cable). If I'm correct that undersea power cables would not use air-spacing, the slower velocity of dielectric-loaded cables means that resonance in a practical system would be much more severe than my illustration suggested.

Fyz

Hi Physicist?

Gave you the GA you should have got previously.

I think you are right about the effect of the undersea cables on resonant length. If it gets near resonance, losses would really skyrocket.

AC current and voltage along a large distance radiates electromagnetic energy, as any antenna does.

Along a HVDC line there exist a stationary electric field and a stationary and orthogonal magnetic field. As both fields are stationary, there is no electromagnetic radiation, thus improving efficiency.

This is a benefit that adds to the avoidance of the skin effect, which obligates to use wider diameter wires in AC modes.

Jaime Soto Figueroa

Chile