Numerically computed zero sequence current

[HiSet]

Modern protective relays can use either numerically calculated Io or actual neutral current fed via residual connecting of three phase CTs.I not sure if there is any difference between the two? Is one better than other?

 

[Alex68]

The best way I know is to use a thoroidal CT to avoid the sum of the errors of the phase CTs.

Whereas the Io is >= the primary current of the CTs, the error are negligible and you can use either the calculated or the residual current without big differences.

Differences may arise if the numerical relay has only one A/D converter and the sampling frequence is too low. So the current are sampled at "very" different moments and the numerical vectorial summation may be affected by errors. Otherwise the residual connected CTs are your solution.

 

[davidbeach]

Numeric relays, while only using one A/D, use sample and hold circuits so that all samples represent the same instant in time. The computed zero sequence current should be about as accurate as that measured from a residual connection of phase CTs. A single zero sequence CT, also known as a toroidal or window CT will provide a more accurate result, although the increased accuracy may or may not have any significance.

 

[jghrist]

One reason for using the calculated zero-sequence settings, at least for Schweitzer relays, is that the default settings all use the calculated zero-sequence (G) settings instead the residual (N) settings. Schweitzer application guides that I've seen use the G settings. If you use the N settings, you have to take care to change every reference. For example, the ground enabled pushbutton on distribution line protection SEL-351S relays controls the 51G1 element with the default settings. If you use the residual 51N1 element in the trip equation without changing the ground enabled logic, it won't work as expected. Other settings like event report triggering have to be changed as well.

I think that theoretically, there should be an preference to using the directly measured residual CT current instead of using a calculated value, but there probably is no practical advantage.

 

[davidbeach]

Using the calculated zero-sequence saves the need to run the residual connection back through the relay. The results should be very close if the residual connection is compared to the calculated zero-sequence.

I'd be careful using any canned logic from any relay manufacturer (and I work for one of them). I think the canned logic is a great point of departure and an excellent way of seeing how someone who knows the relay very well would solve a particular protection problem. But was the protection problem they solved exactly the protection problem you need solved? Maybe, maybe not, but in either case how will you answer the question as to why the program was what it was if something goes wrong? Probably best to make the whole program your own.