Voltage transients at the end of a distribution system

[PhaseShift]

I have heard that voltage transients have the most effect or cause the worst damage at the farthest point in a power system. I would guess that this would have something to do with the impedance being greater the longer the distance but I'm not exactly sure.

Can anyone explain if this is true, or why this is?

 

[ibew441dc]

I have heard that voltage transients have the most effect or cause the worst damage at the farthest point in a power system. I would guess that this would have something to do with the impedance being greater the longer the distance but I'm not exactly sure.

Can anyone explain if this is true, or why this is?

I can only comment based on my opinion and have no formal reference to offer......I would say that high voltage transients would do the most damage to a connected load, which is the furthest point on the system. If a wire rated 600 volts gets hit with 12kv for 1/2 a second, my guess is it may be okay. But if a load rated 120 volts got hit with 12kv I would think it would get smoked.

 

[mivey]

Transients will be worse at a dead end. A transient wave will act much like water rushing down a gutter. When it reaches the end, it will double back over on itself and you can have voltage doubling.

FWIW, transients will also be worse at any point of discontinuity.

TDR meters make use of wave theory to find discontinuities in the line so you can locate trouble points or other discontinuity points of interest like taps, splices, etc.

 

[PhaseShift]

Transients will be worse at a dead end. A transient wave will act much like water rushing down a gutter. When it reaches the end, it will double back over on itself and you can have voltage doubling.

FWIW, transients will also be worse at any point of discontinuity.

TDR meters make use of wave theory to find discontinuities in the line so you can locate trouble points or other discontinuity points of interest like taps, splices, etc.

So it will be the worst at any point with a dead end due to reflection? Not necessarily the furthest point, but any dead end.

I remember seeing something once that the transient wave when it doubles back has a magnitude that can depend on the impedeance of the system compared to the load? Is there any truth to system impedances having any effect on these transient waves. I belive the article had something to do with reflective wave theory.

 

[mivey]

So it will be the worst at any point with a dead end due to reflection? Not necessarily the furthest point, but any dead end.

I remember seeing something once that the transient wave when it doubles back has a magnitude that can depend on the impedeance of the system compared to the load? Is there any truth to system impedances having any effect on these transient waves. I belive the article had something to do with reflective wave theory.

Yes. The transient wave also changes when there is a discontinuity. This would include a change in impedance, like at a transformer, tap, splice, etc.

 

[mivey]

I remember seeing something once that the transient wave when it doubles back has a magnitude that can depend on the impedeance of the system compared to the load?

If the load impedance matches the line's characteristic impedance, the incident (incoming) wave will not be reflected.

An open circuit (infinite impedance) produces a reflected wave with the same amplitude & phase as the incident wave.

A short circuit (zero impedance) produces a reflected wave with the same amplitude but opposite phase as the incident wave.

The rest fall somewhere in between, but the ratio of the complex amplitude of the reflected and incident waves at the load is given by (had to go look it up):

(Z_load - Z_line) / (Z_line + Z_load) and is equal to:
0 for a matched load
-1 for a short circuit
+1 for an open circuit

It might be interesting to note that the wavelength of a 60 Hz signal is just over 3100 miles.

 

[ELA]

If the load impedance matches the line's characteristic impedance, the incident (incoming) wave will not be reflected.

An open circuit (infinite impedance) produces a reflected wave with the same amplitude & phase as the incident wave.

A short circuit (zero impedance) produces a reflected wave with the same amplitude but opposite phase as the incident wave.

The rest fall somewhere in between, but the ratio of the complex amplitude of the reflected and incident waves at the load is given by (had to go look it up):

(Z_load - Z_line) / (Z_line + Z_load) and is equal to:
0 for a matched load
-1 for a short circuit
+1 for an open circuit

It might be interesting to note that the wavelength of a 60 Hz signal is just over 3100 miles.

Now that you went and looked it up what did you originally mean by a dead end? An open or a short :wink:
I think that a transient on a power line is way more complex than simply applying transmission line theory that might normally be applied to communications or data lines.

Wouldn't the greatest area of effect have a lot to do with where the transient was introduced into the system? Other more important factors might be the rise time and duration of transient and how well that transient couples across transformer windings. What type of suppression devices are encountered along the lines etc.

 

[Besoeker]

I have heard that voltage transients have the most effect or cause the worst damage at the farthest point in a power system. I would guess that this would have something to do with the impedance being greater the longer the distance but I'm not exactly sure.

Can anyone explain if this is true, or why this is?

I suppose the answer is "it depends".
Here are two waveforms I measured.
They are for two identical pump motors driven by two identical PWM inverters. The difference was in the distance of the cable runs, P4 being much shorter than P2

At first sight, P2 may look worse. It certainly has considerable overshoot which could cause damage. But on P4, the rate of rise of voltage is much faster (about 3000V/us) which can, and in this case did, cause motor winding insulation failure. Which was what I was investigating.

 

[Besoeker]

It might be interesting to note that the wavelength of a 60 Hz signal is just over 3100 miles.

Interesting but maybe not so relevant given that transients are often sub-cycle events so inferred frequency components would be higher.

 

[fmtjfw]

I suppose the answer is "it depends".
Here are two waveforms I measured.
They are for two identical pump motors driven by two identical PWM inverters. The difference was in the distance of the cable runs, P4 being much shorter than P2

At first sight, P2 may look worse. It certainly has considerable overshoot which could cause damage. But on P4, the rate of rise of voltage is much faster (about 3000V/us) which can, and in this case did, cause motor winding insulation failure. Which was what I was investigating.

From the graphs I see a rise of 0-600V in P2 within 1 horizontal unit while P4 takes several divisions, Also the P2 initial wave front appears steeper. What part of the P4 waveform has a faster rise time? Am I mis-seeing or mis-understanding?

/s/ Jim WIlliams

 

[gar]

090516-0959 EST

fmtjfw:

Look at the the time scaling.

.

 

[mivey]

Now that you went and looked it up what did you originally mean by a dead end? An open or a short :wink:

We don't short the ends of our electric power transmission & distribution lines for normal operation, but they probably terminate near an impedance (tip: they don't short them in your neck of the woods either).

I think that a transient on a power line is way more complex than simply applying transmission line theory that might normally be applied to communications or data lines.

FYI, the reflection equation is indeed used in studying transients in electric power systems. I was not implying the analysis is simple. In fact, while it can be quite complex, and is more often than not done by computer, that does not invalidate the under-lying principles.

Wouldn't the greatest area of effect have a lot to do with where the transient was introduced into the system?.

What if the transient was introduced out in the middle of a cross-country line where no one was at? Who would care but the insulators and conductors? If your transformer receives a direct lightning strike, yeah, you are probably going to figure out that is a bad spot. But if a tree falls in the forest...the transient can produce a traveling wave that leaves the forest and that is what I was talking about.

Other more important factors might be the rise time and duration of transient and how well that transient couples across transformer windings. What type of suppression devices are encountered along the lines etc

Yes, there are a lot of factors that must be considered.

 

[mivey]

Interesting but maybe not so relevant given that transients are often sub-cycle events so inferred frequency components would be higher.

Maybe not real relevant, but when I was looking up the reflection formula (it always seemed easy to get the "+" and "-" in the wrong place) I saw the wavelength formula, it just reminded me of low-frequency wavelengths and how submarines would use them for communication because the waves could penetrate the water to a greater depth. One of our profs said there was a long antenna up the side of a mountain in Hawaii (up the side, on the top, I can't remember exactly ?) that he used as an example.

Anyway, I thought would calculate the 60 Hz wavelength because I could not remember how long it was. I just thought it was interesting, but it may not be so interesting to someone else.

 

[Besoeker]

090516-0959 EST

fmtjfw:

Look at the the time scaling.

.

Thanks gar.
For clarity and consistency, I should have have used the same time axis scaling for both.