CT Saturation 1

[UKEng]

Could someone please explain exactley what happens when a CT goes into saturation, e.g. does the secondary current drop off as the primary current still increase or does it flatten out, does it deviate from a sine wave.

Thank you

 

[nitro]

Chuncky hope this helps.

CT Saturation

DC Saturation
When a fault occurs the current usually contains a DC component. The total flux required to produce the offset drives the CT into saturation.

Flux reaches saturation during the positive cycle and the exciting reactance decreases shunting the primary current thus distorting the secondary.

During the negative cycle of the primary current, the core becomes unsaturated.

As the DC component decays the negative cycle of the primary current and flux become greater and the core eventually runs out of saturation during a complete cycle returning to steady state.

The DC component time constant is the X/R ratio of the primary circuit.

 

[electricpete]

The output of the ct will read low in portions of the waveform where the ct is in saturation.

As was mentioned above, this will happen once per ac cycle if the saturation is induced by dc offset. This will happen twice per ac cycle if the saturation is simply due to high ac current magnitude and burden.

I posted an example waveform of ct saturation during induction motor start at the following link. You'll also see an analysis which shows that the distortion ("glitch&quot occurs at current zero crossing assuming a purely resistive burden.

thread237-9886

 

[scottf]

Chunky-

Nitro is correct about saturation during a fault.

Of course you can also drive a CT into saturation by driving too much current through it under a steady state condition. ANSI/IEEE C57.13, defines saturation for protection cores to be where the error exceeds 10%. In practice, saturation is said to occur at the point that the secondary current versus primary current fails to be linear.

Normally, a CTs performance is evaluated by the excitation curve, which is a plot of secondary terminal voltage versus excitation current. Saturation is said to start at the "knee" region of the curve and it becomes that small increases in the secondary terminal voltage must be brought about by large increases in excitation current.

In "steady-state saturation", the waveform will be sinusoidal (in CTs).

Hope this helps...

 

[jwerthman]

APT (now owned by NXTphase has available a great little spreadsheet program for analyzing CT performance. You can vary the current, X/R ratio, burden, and then it will plot the secondary output of the CT. It makes several simplifying assumptions, but it seems to be fairly good. It's available for download on their website at

http://www.nxtphase.com

(click on APT section from the home page and then look for downloads/other). A very similar program (maybe the same one) is available on one of the IEEE websites, but I can't remember where.

 

[electricpete]

scottf

I hope that once again you will permit me a respectful disagreement. I know you have a lot of experience with instrument transformers but one of your statements sounds incorrect to me.

I don't believe that the output waveform of a ct in saturation can be sinusoidal under any real-world circumstances. During different parts of a cycle the flux will sometimes be above the knee of the curve and sometimes below the knee of the curve. This gives a varying error current (magnetizing current) during different parts of the cycle. And the error current will not be a simple percentage of the input current. The output is therefore distorted and nonsinusoidal, even for a sinusoidal input.

 

[jwerthman]

I have to agree with electricpete. By definition, any transformer in saturation is no longer linear and cannot reproduce a sinusoid at the secondary.

 

[scottf]

It really depends on how much in saturation it is. If during the cycle, it is above and then below the knee you are somewhat correct. However, if it doesn't not cross the knee (which is common for deep saturation), then the output is sinusoidal. Don't forget about the role that memory/hystoresis plays during the changing from positive to negative cycles.

 

[electricpete]

For steady state ac current input, the flux will reverse direction twice per cycle and must pass through zero twice per cycle. during the period of time when flux is low near the zero crossing, the ct is not in saturation, and the error current is much smaller than when the flux is high and ct is in saturation.

Since you mention deep saturation, I think I know what you're driving at. In deep saturation the fraction of a cycle that the ct spends below saturation is very small. The majority of the waveform might look sinsusoidal except that small fraction of cycle.