I have a situation where we are trying to add ground fault to some existing substations having a secondary Main-tie-Main arrangement. Intent is to bring the system up to code and provide some protection to the owner.
System is 480/277v 3ph 4w, 3000 to 4000 ampere mains. The swgr is old and neutral bussing run such that installing neutral ct's is near impossible from a shutdown situation. In addition the Neutral bussing has multiple branches.
In a perfect world, we would install ct's on the mains and tie (ABC+N) and connect the tie in a differential fashion. So it appears that this can not be done easily.
I have encountered ground fault protection on existing buildings, 3ph 4w, 480/277, where CT's were put on the phases only, none on the neutral. A GE IAC relay was used to trip the main; Inst pickup was set to 1200 amperes, time delay set to 1 second, Inst and TOC contacts wired in series. This basically mimics a GFault curve found on a typical low voltage breaker. From a nusiance trip perspective, if the loads are relatively balanced, which they usually are, there is no problem. Also, the need for the differential connection to the tie completely goes away. (The fault is sensed at the main breaker that feeds into the fault only, no issue with ground current going up into the other side of the sub N-G bond and across the tie.)
What thoughs does the group have on this in terms of Code and practical application.
Thanks,
Jay Petersen
System is 480/277v 3ph 4w, 3000 to 4000 ampere mains. The swgr is old and neutral bussing run such that installing neutral ct's is near impossible from a shutdown situation. In addition the Neutral bussing has multiple branches.
In a perfect world, we would install ct's on the mains and tie (ABC+N) and connect the tie in a differential fashion. So it appears that this can not be done easily.
I have encountered ground fault protection on existing buildings, 3ph 4w, 480/277, where CT's were put on the phases only, none on the neutral. A GE IAC relay was used to trip the main; Inst pickup was set to 1200 amperes, time delay set to 1 second, Inst and TOC contacts wired in series. This basically mimics a GFault curve found on a typical low voltage breaker. From a nusiance trip perspective, if the loads are relatively balanced, which they usually are, there is no problem. Also, the need for the differential connection to the tie completely goes away. (The fault is sensed at the main breaker that feeds into the fault only, no issue with ground current going up into the other side of the sub N-G bond and across the tie.)
What thoughs does the group have on this in terms of Code and practical application.
Thanks,
Jay Petersen
