Sepam relays and ANSI 50 & 51

[radug]

Good afternoon,

I have a question regarding ANSI 50 & 51 settings in SEPAM relays:

If I define a IDMT curve (IEEE or IEC) for ANSI 51 and then I add a DT setting, with, for example a 0,4 sec time delay. Which will be the tripping characteristic?

I mean, I thought that the DT current setting blocked the IDMT curve, so that for currents equal or greater to the DT setting, tripping was done in the time defined in DT (0,4 sec in this case) regardless of the tripping time of the IDMT curve (I mean, even if the tripping time according to the IDMT curve was for example 0,1 sec).

Now, a client has told me that tripping will be done following IDMT curve if IDMT curve tripping time is less than the DT tripping time.

Which is the correct tripping time?

I searched Sepam manuals and was unable to find info about this.

Will the same tripping characteristic happen with other relays? For example, ABB REF/RET/REM?


Thanks.

 

[davidbeach]

If the output ORs the inverse and definite time then the client is correct. If the output ANDs the inverse and definite time then you are correct. Don't know about the SEPAM, but in many relays either is possible depending on how the user programs the relay.

 

[jghrist]

In the SEPAM relay, both elements are operative and the fastest element trips first. If the high current part of the IDMT element is faster than the DT setting, then the IDMT will trip. I discovered this while trying to figure out why my TCC curves in SKM came out looking notched. The DT element only was effective for a short range of currents.

 

[radug]

Thank you for your replies,

I finally spoke with a Merlin Gerin technical support engineer and was told that Sepam relays do not "cancel" the 51 curve when a 50 DT with a time delay is put in the settings. He told me that the final tripping characteristic is the sum of both curves, that they behave as if they were two different and independent relays. He also told me that all the other manufacturer relays that he knows about behave the same way.

This will make me revise other coordination studies. I usually did a 51 inverse + 50 time delayed configuration for coordination. For example, to coordinate 3 steps, one with no time delay, the next with a 0,3 time delay and the other with a 0,6 time delay.