[TowerTalk] HV LInes DC

Jim Lux jimlux at earthlink.net
Sun Aug 20 23:22:55 EDT 2006

At 06:22 PM 8/20/2006, k3hx at juno.com wrote:
>I have been fascinated by the discussion of HV DC lines and would be
>grateful if the following questions could be answered.
>I beg the indulgence of the moderator for this off-topic topic.
>1  How is the HV DC contrived.  Generated HV or rectified?

Rectified from HV AC:
as opposed to, say:
  generated directly as DC - don't want brushes, particularly at 500kV
  stepped up as DC by a voltage multiplier.

These days, semiconductor rectifiers are used (thyristors, SCRs).. in the 
"good old days" (1960s) mercury arcs.

A common configuration is two rectifiers, one generating + relative to 
ground, the other generating -.  Most commonly, the rectifiers are fed with 
two 3 phase systems, 60 degrees apart (a so-called 12-pulse converter) so 
the ripple is pretty low (and high frequency), making it easy to filter 
out. (that is, the transformers feeding the rectifier are connected in Wye 
and Delta to get the two shifted phases) The rectifiers are phase 
controlled to control either voltage or current output.

>2  Once it reaches the other end, how it is turned back into AC?
>    Motor-generators,  really big inverters?

Multipulse inverters using thyristors these days.  Mercury arc triggered 
rectifiers in the old days.

One wants to avoid rotating machinery at those voltages, because of the 
insulation hassles. (Standard generators, etc. in power plants put out tens 
of kV, not hundreds)

>3  What are typical voltage ranges and distances travel led?

several hundred kV (typically, something like +400kV and -400kV, relative 
to ground)
1000 km or more
Although, there are short DC links used to interconnect two AC systems

>4  Do the lines rely on the earth for "return" or do they run + and - ?

Normally + and -, but most can run using an earth return if one wire goes 
down, or one of the converters fails. The Pacific Intertie uses (at the 
southern end) a bunch of electrodes in the Pacific Ocean off the coast, 
some 48 km from the end of the line.

As one can imagine, one wants a fairly low "ground rod impedance" if you're 
carrying 1000 Amps or so.

Googling "HVDC link" will find quite a bit of info, including some pictures 
of "valve halls" where  you can see enormous stacks of 1kV water cooled 
thyristors.  This was one of the early applications of optical fibers, for 
triggering the thyristors in the stack.

If you want some practical construction info, you might try googling for a 
Los Alamos National Lab report on High Power Microwave Transmitters by 
William North.

>Tim Colbert  K3HX

More information about the TowerTalk mailing list