Martin Angove wrote:
In message ,
Tony Williams wrote:

In article ,
John wrote:

OK, I'll rephrase the question: why can all gens not be synced by
a common time standard instead of by the neighbouring mains
waveform?

I think because there is the unavoidable problem
of phase shift along the transmission cables. The
wavelength of 50Hz is about 6000Km, at which the
phase would have shifted a full 360 degrees.
Sounds no problem, but that represents a 10 degree
sync difference only every 167Km or 104 miles.
That might cause a problem on the UK's grid.


You mean that if generator "A" and generator "C" are each 100 miles away
from generator "B", are all synchronised together in *absolute* time by
an external reference and all supply the same line, the waveform
supplied by "A" and "C" will be 10 degrees out of phase with that
supplied by "B" as measured from "B".

A=====B=====C

Sounds nasty.

So am I right in thinking that the way it is actually done is
(effectively) to have one generator start first, and for each of the
others to synchronise with the received waveform at their own locations?
In other words the generators are, in absolute terms, out of phase with
each other but due to wavelength / propagation delay / whatever are
for all practical purposes synchronised.

I can see how this would work for a "bus" topology, but not for anything
involving either rings or a mesh as the length of two or more paths
from generator to generator will be different and hence received
waveforms from each one will be different. How is the grid/supergrid
actually organised in this country? How do they do it in the US where
distances are vastly greater?

Learn something new every day on this ng :-)

Hwyl!

M.

With a mesh, the UK is fairly narrow, so the phase shifts east-west
would be small compared to north-south, I guess.

In principle I imagine one could have no end of phase shift if, and
only if, something... ha, I cant explain it. But I guess its apparent
theres no limit to how much phase shift could be accomodated within a
network, but it would depend very much on its layout and current flows.
You'd have to be careful what you connected to what via what, but in
principle one could presumably operate a grid with a ripple shaped
phase shift, like a ripple in water, with the phase shift adding upto
several whole cycles.

Someone can now explain why this is total twaddle.


NT