I have been looking into the adjusted pickup method and associated shift factors used when multiple source bus configurations are being coordinated [ex. parallel generator systems two or more] and was hoping someone could point me to an IEEE standard or some industry articles on this to better understand this on shifting TCC’s around a specific location or device. On one example where two different sized generators were modeled it shifted the smaller units beyond the TCC of the larger unit indicating a longer trip time for that device for a downstream fault condition. It looks that the shift factor calculation looks at the generator current to the fault / the fault current at the location specified, assuming that larger gen breaker sees more current and trips faster than the smaller unit, this is what I am seeing and was looking for some backup information but searching around it appeared the information was pretty scarce and was hoping someone could point me toward some. THANKS.

Mike, I am not entirely sure of what you are asking. I am not aware of a relay that shifts the curve. I have set relays to have several settings groups based on differing conditions E.G. - setting protection based on one generator when only one is connected but changing if more are tied in.

I don't believe that you will find any reference to this in the literature. It is just a more accurate method to determine selectivity.

I have recently completed an update to our paper mill's power system ETAP model. ETAP TCC's have a feature called 'normalized mode' which shifts TCC curves based on the fault current that each device see's for a specified faulted location. They have an easy way to change the fault location so that it's possible to see how the relays coordinate in all cases. I found that in some TCC's, it was not possible to achieve selectivity without using this normalized view mode.

EasyPower software lets you fault a bus that is included in the TCC. It then clips each TCC curve at the thru-fault level seen by each device. I believe this might produce the same results but a bit harder to study the end effects.

See if this helps explain the situation. If I understand your question correctly, the shift is for different devices “seeing” different currents for one event. i.e. if a fault is downstream from a generator distribution switchboard/switchgear, the individual breakers for each of the two generators would see each generator’s individual short circuit current. However, the common distribution bus would see the short circuit current of both generators.

The result is that when compared to each other, the generator breakers and the downstream breakers do not “appear” to operate as a TCC would indicate. Each generator would see their respective current and the distribution board would see both.

As an example, let’s say each generator produces 50 percent of the total short circuit current and the bus sees 100 percent. (from both generators) The generator overcurrent device would see 50 percent less current than the bus so the bus device could operate faster than its time current curve indicates relative to the generator device. Therefor the curve needs shifted to account for the devices seeing different currents. This is often referred to as “apparent coordination”

This is similar to what happens with the primary and secondary of a transformer. Each side of the transformer would have different currents and so do the primary and secondary device’s. As an example, a 480 v to 240 v transformer has a 2 to 1 ratio and the secondary current is twice as much as the primary current for events on the secondary. You can’t plot the primary and secondary devices as normal, one needs “shifted” by the ratio to view the “apparent” coordination.

I don’t know of any standard but this is a pretty common requirement for network (multiple sources/paths) in a coordinate study. And then we can talk about how this affects arc flash calculations….

Jim,
Exactly, on one TCC it shows the generator TCC’s with a footnote that the curves indicated a “non-running” condition, then there was a curve for each generator operating in an island mode showing coordination with either generator or downstream devices. The sequence of operation for the parallel switchgear stated that if one generator was able to carry the load the other shall go offline so at any time you may only have one of two generators operating or both in parallel, a final TCC indicated “Parallel Operation” it included a “Curve Shift” where the curves were shifted on the TCC to reflect “apparent coordination” based on each of the generators contribution to the specified faulted bus downstream. Whereas like you indicated each generator contributed a fraction of the total fault current at the bus therefor each generator breaker under the fault condition only sees its contribution to the fault and will operated differently based upon that. In this case one generator was significantly larger [almost 2x] so when you looked at the “apparent” coordination the larger generator TCC was to the left of the smaller digging deeper it appears this is because of the higher fault current it would see causes its breaker to operates before the smaller gen breaker that sees less of the fault current thus longer time delays. It looks as if it compared the contributed faulted current / calculated fault current at the specified location or relay to determine the shifting factor for the apparent coordination. I did not know if there was any literature published by IEEE / ANSI, etc. describing this in more detail…I appreciate all the responses. As for the I.E. not sure how you calculate because the I.E. will differ with the operating mode of the generators that are automatically controlled and if you are working on a bus is a different area how are you to know the current operating condition.. The best solution no hot work while operating on generator power, or I would think installing reverse power relays [currently not on the existing EPS] for each of the generators and set the reverse power pick-ups equal to each other than using these settings in conjunction with the operating time of the breakers…..just a thought

Sounds like quite a project. I just threw the arc flash question out there to stir things up. It actually could be solved for the different scenarios buy using a piece wise solution. i.e. calculated the incident energy up to the point where the first device trips, then calculate for the next interval with what sources remain on line etc. However it's not perfect since there is also a generator decrement so the fault current is also decreasing from each unit. Good luck!