Ok please don't blow me up if this is a dumb question: I have a plant with an ungrounded delta-delta main power transformer. The main and feeder breakers on this 2,400 volt system all have ground fault protection. The protection is very touchy, and nuescance trips are not uncommon. Fault isolation is very poor as well. In the old days there was only ground fault annunciation, but no trip device. I am researching what the actual code requirements are to see if we should perhaps consider defeating the trip functions and only have alarm notification of a detected fault. All code requirements that I can find in the NEC say that this is only required for solidly grounded wye systems up to 600 volts. Furthermore, it specifically gives an excemption to continous operating industrial sites where uncontrolled outages could present other hazards. So am I reading this right that my ungrounded delta 2400 volt system isn't required to have ground fault protection?
Ground fault protection on an ungrounded delta system at 2,400 volts
- Thread starter Tfret
- Start date
- Status
electrofelon
Senior Member
- Location
- Cherry Valley NY, Seattle, WA
Note that you are speaking of two different requirements. The first is GFP in 230.95. Note that this detects and interrupts the service. The other issue is "ground detectors" under 250.21(B) which could be just a notification system. I am not real familiar with MV systems, but it appears that the NEC may not require either for your service. whether it is wise to have neither, is another question.
- Location
- San Francisco Bay Area, CA, USA
- Occupation
- Electrical Engineer
As electrofelon said, those sections ONLY apply to systems under 1000V, so for a 2400V system, there are no requirements for GF protection or even detection. 250.20.C says that grounding of systems over 1000V shall be PERMITTED, but that's not the same as being required.
The main PURPOSE of using an ungrounded delta system is exactly what you have described as the "used to be" scenario. You use(d) an ungrounded delta so that the first ground fault did NOT cause the entire facility to shut down, because what happened was that you went from an ungrounded delta to a corner grounded delta, and life (or rather, productivity) went on. When I worked at a steel mill, we had simple "ground lights" on the wall of the electric shop that told us IF there was a ground fault somewhere, then we had to go find it. But nothing shut down, because the cost of that was astronomical.
But honestly, a LOT of people were uncomfortable with that concept, thinking, incorrectly, that a GF protection system would protect people. They will NOT. A typical GFP relay for a 2400V system will trip at something like 5A, which is way more than lethal at much lower voltages, at 2400V it just defrays some of the cost of cremation. Nevertheless, skittish engineers would step into a facility with an ungrounded system and immediately implement GFP and start causing all kinds of problems. So this is a discussion that all of the stakeholders need to participate in, and I would suggest getting an EE involved who truly understands the issues, not someone who has a knee-jerk reaction to seeing anything ungrounded.
The main PURPOSE of using an ungrounded delta system is exactly what you have described as the "used to be" scenario. You use(d) an ungrounded delta so that the first ground fault did NOT cause the entire facility to shut down, because what happened was that you went from an ungrounded delta to a corner grounded delta, and life (or rather, productivity) went on. When I worked at a steel mill, we had simple "ground lights" on the wall of the electric shop that told us IF there was a ground fault somewhere, then we had to go find it. But nothing shut down, because the cost of that was astronomical.
But honestly, a LOT of people were uncomfortable with that concept, thinking, incorrectly, that a GF protection system would protect people. They will NOT. A typical GFP relay for a 2400V system will trip at something like 5A, which is way more than lethal at much lower voltages, at 2400V it just defrays some of the cost of cremation. Nevertheless, skittish engineers would step into a facility with an ungrounded system and immediately implement GFP and start causing all kinds of problems. So this is a discussion that all of the stakeholders need to participate in, and I would suggest getting an EE involved who truly understands the issues, not someone who has a knee-jerk reaction to seeing anything ungrounded.
Ground protection? If the provided ground protection is through current based elements it would be difficult. As there is too few or sometimes no current would flow during ground faults.
You could used voltagebased protection elements. But fault location will not be easy.
Or you could install zigzag or wye.delta transformer to provide a means as ground source.
I do understand the system, but by no means am an expert. The feeder breakers and all downstream MV breakers use a zero sequence CT and a GKC protective relay. The feeders are set at either 4 or 2 amps. All down stream relays set to 2 amps. The main uses a grounded wye-open delta circuit where the open delta connects to an over voltage relay and a parallel resistor. It is supposed to trip at 4 amps. The issue is that the charging currents in the sheilded cables are close to 2 amps. Any unusual circumstance, like a nearby lighting strike, or even a sudden heavy load change on a circuit, can influence the charging currents enough to trip a relay. And seldom does a single relay trip. Usually at least 2 and often more relays trip. So trying to determine if a true fault caused the problem is extrememly difficult. I believe as someone else pointed out, the whole purpose of installing an ungrounded delta is to ensure robustness and system reliability even during a ground fault. I have an idea about changing the trip functions to indicator lamps only. But before I raise this suggestion I just need to be sure there is no code violation in doing so.
zog
Senior Member
- Location
- Charlotte, NC
Sounds like you need to update your coordination study and settings
We have a current coordination study, and I have the system fully modled in EasyPower. The problem is that trying to set such low fault currents (the bottom two settings on the GKC relays) across the entire system, there simply isn't enought space between the settings for reliable coordination. But again, my desire is to determine with confidence if these devices are "required" by law or code. Once i have the answer to that question, I can pose making changes and determine what works best for the plant.
- Location
- Wisconsin
- Occupation
- PE (Retired) - Power Systems
Did your coordination study really examine the complete GF issue? Looking at a typical 50/51N TCC plot is not sufficient as it rarely addresses transient conditions nor even some steady state conditions, like system charging current.
Told you so, it is difficult to have an overcurrent coodination in case. Power system models in skm, etap, edsa, easy power cant model enough charges ug transients.
Since you have already pts connected in wye grounded open delta with OV relay, then these are devices assign for ground faults.
In this case, you need to coordinate your OV relay with all OC ground relays. Still it would be difficult.
Also, is the open delta pt also wired to the feeder relays, aside from having a zero seq CT?
Sent from Mars
Since you have already pts connected in wye grounded open delta with OV relay, then these are devices assign for ground faults.
In this case, you need to coordinate your OV relay with all OC ground relays. Still it would be difficult.
Also, is the open delta pt also wired to the feeder relays, aside from having a zero seq CT?
Sent from Mars
- Location
- Wisconsin
- Occupation
- PE (Retired) - Power Systems
GF tripping is generally not required, by the NEC, except for 480Y/277 solidly grounded systems.
GF alarming is usually required on ungrounded systems irrespective of voltage.
electrofelon
Senior Member
- Location
- Cherry Valley NY, Seattle, WA
Jim:
Per the nec? My reading of it was, surprisingly, that there is no gf detector required for over 1000 volts.
- Location
- Wisconsin
- Occupation
- PE (Retired) - Power Systems
You are correct. The pertinent reference is 250.21.
- Status