CT Saturation

I told your previous anonymous avatar that I did not know what the L400 means.

If it was me, I would look in the manuals, drawings and data archive of the plant [Schedule of equipment, Data Sheets??] to see if there is any information about the item. There might be a test magnetisation curve, which would be compared with periodic tests in service.

If this fails, speak to the local Brown Boveri dealer, tell him you have this CT, give him all type numbers and serial numbers which are marked on it and ask him if he recognises the 2.5L400 and knows what it means and if he could provide more data.

You now write of protection service.....in European practise, protection CTs are usually specified as 5P10 or 5P20. P means protection, while 5 means 5% error at the number of times rated continuous current which follows the P.

http://en.wikipedia.org/wiki/Current_transformer

gives an example of a current transfo with 2.5L400 designation, its specification is that it will give 400V into its secondary load at 20 times rated continous current. Note this 400V is not a saturation voltage. The definition of 2.5L400 would suggest that the magnetising current was over 2.5% x 100 amps = 2.5 amps with 400V applied to the secondary "on the test bench" for a 5 amp secondary current at rated primary current.

As you write, this means that the maximum burden would be 4 ohms for the specified accuracy. This would mean the rated [continous] VA would be 5 amps x 5 Amps x 4 ohms = 100VA [I²R], for a 5 amp secondary.

Knee-point saturation voltage is defined in the above Wikipedia reference.

There are problems in calculating any relation between knee point and the 400V secondary voltage...

1. Roughly, magnetising current will have a 90 degree lag to a resistive secondary current. A simple vector diagram suggests magnetising current is about 0.2 x 100 = 20 amps for 2.5 amps magnitude error at 100 amps secondary.
2. Secondary burden current will not be purely resistive in phase due to leakage reactance (the CT has its own resistance and inductance in the secondary current path). So it will not be at 90 degrees to magnetising current.
3. While the secondary current may be reasonably sinusoidal, the magnetising current will be very "peaky". So a vector diagram assuming sinusoidal values is not reliable.
4. The ratio testing is done with r.m.s value reading instruments. Summing a near sinusoid current with a peaky one at about 90 degrees lag and estimating the rms current is not an easy calculation.
   
5. The degree of saturation at 400V is not known.
   

So, you would be wrong to assume 400V is the saturation voltage.