----------------------------

George Tsakalos said:

Ben, first of all thanks for the answer, it covers me completely. I

thought i was right to suspect that a capacitor can not produce any

harmonics (a passive element cannot produce anything right?)

My special interest is in capacitors that are pole-mounted in MV networks.

Is there any theory or "algorithm" that states where we should place

capacitors to correct the cosö in the span of a MV line?

And by the way, can we use filters instead to filter out anything else

than the 60 or 50 Hz signal and stop worrying about any high harmonic

currents? You mentioned the inductance in series solution with the

capacitors. But, wouldn't that work in the opposite way we're trying to

go? Wouldn't that reduce the amount of Var the capacitors are "feeding"

the network?

Thanks in advance again,

George Tsakalos

----

The capacitors, in the situation that you mention, do correct the power

factor However, their main purpose is to improve the voltage profile.

Suppose that, on a feeder, the voltage must be kept within bounds of +/-

5%. and if no capacitors are used. it can happen that the voltage will be at

the upper bound at the source end and below the lower bound at the other.

Then a capacitor bank at an intermediate point can help by improving the PF

at that point, resulting in a lower voltage drop to that point (beyond that

point it has no effect).

Instead of having a source at 1.05 pu and a mid point at 0.99 pu and end of

line at 0.93pu, with capacitors at the mid point, the profile might be

1.05->1.03->0.97pu at heavy load and yet be OK at light load.

The problem with a set algorithm is that the load distribution is not the

same in all cases. it may be necessary to actually determine a voltage

profile by measurement at different times of day and home in on correction

for the worst case by considering various amounts of compensation at

different points. This could be an interesting optimisation problem. There

probably are many "rule of thumb" procedures out there as this is a common

problem.

The inductance in series is already there- that is the problem. It is this

inductance that leads to higher voltage drops in the first place.