CT linear relationship query 3

[pavement]

Just a quick question guys, the rule of thumb that says the a CT should maintain the linear relationship between the reflections of the CT ratio for up to 20*the rated primary current.Is this with regards to the asymmetrical peak or RMS values? i.e. if I had a bus protected by a OC relay with a prospective asymmetrical fault current of 108kA Pk (76.6kA RMS), would a CT of around 3850:1 suffice?

 

[DanDel]

CT ratios are usually of standard values and are based on the normal loading of a circuit. If the circuit is rated for 2000A, than a 2000:5A CT would normally be used( 5A secondaries are also typical).
The 20x rating is used to ensure linearity within a certain range of the CT that will be used by overcurrent devices. As long as your trip setting is within the 20x of the CT rating, there should be no problem(assuming adequate burden calculations) since anything above that rating will still provide a high enough secondary current for the protective relay to pickup.

 

[pavement]

Cheers Dandel, the 1A secondary comes from the fact that im in the U.K, so basically what youre sayin is that as long as the relay causes the appropriate breakers to trip before the CT reaches the saturation level at which the CT sits down,CT indicative of normal load currents at that point of the system can be used?

Once again many thanks.

 

[buzzp]

Thats a good question.
I have always assumed they were RMS values. Perhaps this is not the case. Maybe the mfg CT curve will shed some light on this for you.

 

[DanDel]

I think you've got it. Most Inst trips are within 6-10x the FLA of the circuit, so 20x will work for most typical applications.
I'm not sure what you mean by 'sits down', but it's important to remember that the CT doesn't stop working at current levels 20x above its rating; it simply fails to respond in a linear fashion.
Think of it this way: there is a maximum secondary output above which no increase in primary current will appreciably change(it is 'saturated'). It doesn't stop producing secondary current, the secondary current it produces is just no longer in a direct ratio with the primary current.

 

[cuky2000]

IEEE/ANSI Standard C57.13 suggests applying CTs for relaying based on the maximum symmetrical rms fault current.

ANSI and IEC are differents, According with ANSI the current should not exceed 20 times the CT current rating and the burden voltage not exceeding the accuracy class voltage of the CT.

There is a rationale for choosing a CT to produce the knee-point on the excitation curve at the maximum symmetrical fault current since the magnetizing reactance is at a maximum. Observe that the knee-point of a typical excitation curve is about 46 percent of excitation voltage corresponding to 10 amperes excitation current.

A rule of-thumb suggests that the C-rating be twice the excitation voltage developed by the maximum fault current, which guarantees operation near the knee-point of the excitation curve for the maximum symmetrical fault.

 

[RalphChristie]

The performance of CTs when they are carrying the load current is not of concern as far as relaying needs are concerned.
The DC-component in fault-current causes the flux linkage to increase considerably above their steady-state peak, driving the CT into saturation. The secondary current of a CT may not represent the primary current faithfully if it goes into saturation, hence relays which depend upon the secondary current are likely to mis-operate during this period.

The ALF (Accuracy limit factor) specifies the maximum current at which the the CT must still conform to accuracy specifications. The ALC (Accuracy limit current) can be determined by multiplying the rated current (primary or secondary) with the ALF.

Therefore: ALC = I rated x ALF

Standard accuracy limit factors for protective transformers are 5, 10, 15, 20 and 30

The accuracy classes to which Protective CTs are manufactured is 5P and 10P. "P" indicates that it is a protective CT. Numbers 5 and 10 are the percentage composite error at rated primary ALC.

Thus a 10P15 CT have a 10% error at 15 times primary current.

For instantaneous overcurrent relays a Class 10P with a rated accuracy limit factor of 5 should be adequate.
For overcurrent relays with inverse-time characteristic, however, class 10P accuracy should be specified together witha rated ALF corresponding to the maximum overcurrent at which reasonable accuracy of the relay time caracteristic is required.
If you want to detect earth faults also(relay connected in the residual circuit of a three-phase set of CTs), a class 5P is required. (Essential to employ CTs with low exciting currents to secure reasonably low earth fault settings)

CT's for most forms of protection are covered by BS3938:1973 and guidance in the application of CTs is given in Appendix B of the British Standard.






Regards

Ralph

 

[scottf]

The ratings are based on RMS and not peak.

If you think in terms of how the tests our done, it's quite clear that ratings are based on RMS.

For the IEEE market, units rated like C800 are tested with a 60 Hz input to show that they can develop 800 V on the secondary, where 800V corresponds to 20 times rated current with an 8 ohm burden and a rated secondary current of 5A.

For the IEC/BS crowd, of course 1A secondaries are the norm and you'll see ratings like 5P20 30 VA, meaning that it must be able to stay within +/-5% accuracy up to 20 times rated current with a 30 VA burden connected.

Sorry to ramble...bottom line...it's based on RMS.