480 VAC, 3∅ Distribution Question

[electrofelon]

I was going to reply to this, but looks like OP clarified some in last post. Appears to be an ungrounded system (at least the 480 volt system is) and is grounding done "locally" at each building? But likely only to non current carrying components leaving the system still ungrounded.

Can make for a real mess with stray currents when you have a ground fault on phase A in one building, a fault on phase B in another building and a fault on phase C in a third building.

Yes I agree different faults in different buildings can be a mess when there is only a high impedance path between them

 

[Bretzel]

So, it appears from the comments, clearing the present fault will make the system slightly safer but not safe. Correct?

 

[mbrooke]

So, it appears from the comments, clearing the present fault will make the system slightly safer but not safe. Correct?

Correct. The system needs an EGC and ground detector lights.

 

[kwired]

Correct. The system needs an EGC and ground detector lights.

Or intentionally ground one of the conductors.

Is this 480 directly utilized by any load or is it always transformed to 120/240 - 208?

Technically should still have separate EGC from grounded conductors per today's codes, but when it was originally installed likely would have been ok to just run grounded conductor to each building and then connect to a grounding electrode at each building and run separate EGC's after first disconnect at each building.

 

[Wire-Smith]

=

The 480 secondary side of the transformers has only two terminals. There are only three wires leaving the transformer building. T1 connects across leg A and B of the 3 plant distribution wires; T2 connects across leg B and C; T3 connects across legs C and A. There is a stud welded to the bottom of the side of each transformer can. This stud connects to a conductor which bonds each can to the next and then to the same ground peg as the primary. There is no distribution ground or neutral wire from here downstream.
View attachment 20968
=

thats ungrounded delta

 

[Wire-Smith]

So, it appears from the comments, clearing the present fault will make the system slightly safer but not safe. Correct?

it is often actually argued that ungrounded systems are safer. grounded systems kill people all the time as well.

people with grounded systems often drive ground rods at equipment for ground fault return current path as well(incorrectly and much more hazardous on grounded system). with ungrounded system you don't need return path for non current carrying metal parts. everything (conduit, enlcosures, motor frames, etc.) just needs bonded together.

 

[kwired]

thats ungrounded delta

Somehow I missed that, but yes as he describes it, it is a delta secondary. A wye secondary would have one line from each transformer all tied together (the neutral point) and the "phase" conductors would each tie to the remaining three terminals.

it is often actually argued that ungrounded systems are safer. grounded systems kill people all the time as well.

people with grounded systems often drive ground rods at equipment for ground fault return current path as well(incorrectly and much more hazardous on grounded system). with ungrounded system you don't need return path for non current carrying metal parts. everything (conduit, enlcosures, motor frames, etc.) just needs bonded together.

Safer to an extent. A ground fault in the system doesn't energize non current carrying components, it just creates a grounded reference point and now the system is grounded. If you ignore fault indicators and continue to run that way, additional faults will cause same problems that grounded systems can have. The idea is to indicate there is a fault, but not have immediate shut down and time to make an orderly shutdown for certain critical processes. If you ignore the fault indicator a second fault may cause sudden shutdown anyway when an overcurrent device responds to that fault.

 

[mbrooke]

Safer to an extent. A ground fault in the system doesn't energize non current carrying components, it just creates a grounded reference point and now the system is grounded. If you ignore fault indicators and continue to run that way, additional faults will cause same problems that grounded systems can have. The idea is to indicate there is a fault, but not have immediate shut down and time to make an orderly shutdown for certain critical processes. If you ignore the fault indicator a second fault may cause sudden shutdown anyway when an overcurrent device responds to that fault.

I disagree, without an EGC a ground fault on an ungrounded system may kill if the system is large enough due to capacitance.

 

[__dan]

The first fault already exists, the only question is where. Single motor size up to 100 hp. If there are any motor drives connected directly to the floating supply, likely the drive front end (filters) are internally also Y connected and grounded.

Drives can be floated at the drive front end grounding screw (usually) but no one I know pays attention to this. The drives could be expected to be found with the grounding screw installed.

First fault is assumed. What happens to the second fault

The high impedance of the open EGC, (or through the Earth), likely may keep the circuit open, but create the voltage raise at the more remote structure. They are all remote regarding the EGC between them, but they are also connected haphazardly through touching and air, gas, water lines.

The actual fault current return path will either be enough to trip the breaker or raise the voltage of all the connected exposed remote metal (that is normally expected to be grounded). The remote EGC system and everything connected to it, because of the high impedance Earth return path, it can go to 277. If the breaker trips instead, my guess would be that the drives were bigger than the fault current through them.

 

[Russs57]

Couldn’t we run around the plant with an a amprobe and see which of the three phase conductors don’t sum to zero?

Asking because I have limited experience in non standard systems.

 

[kwired]

Couldn’t we run around the plant with an a amprobe and see which of the three phase conductors don’t sum to zero?

Asking because I have limited experience in non standard systems.

Maybe. If there is only one fault and it is (was) an ungrounded system there will be no "fault current" to cause a current imbalance on the supply conductors. This procedure will be more effective on a grounded system as that fault current will seek "ground".

 

[__dan]

Couldn’t we run around the plant with an a amprobe and see which of the three phase conductors don’t sum to zero?

Asking because I have limited experience in non standard systems.

The system discussed here, there is no guarantee the metal enclosure is safe to touch to open and Amprobe.

You could bang 3 ft of pipe in the ground at an assumed remote earth, tie 500 ft of #12 thhn to that and run around with a voltmeter to see if you can measure a voltage difference from the remote Earth probe to *any* metal, elec equipment, gates, metal siding, I beams ...

Next would be making a list of all the drives on the floating supply, gathering their factory instructions, and opening them looking for the bonding screw (maybe after checking to see if the enclosure is not at a voltage level above Earth).

Not a time to get adventurous. Get management to sign off on the next procedure.

 

[Wire-Smith]

Couldn’t we run around the plant with an a amprobe and see which of the three phase conductors don’t sum to zero?

Asking because I have limited experience in non standard systems.

every facility even underfunded ones with ungrounded or resistance grounded systems that i have been too have a ground fault locator (you can buy used ones cheap) it's a toner you follow where the tone gets louder and you can find the fault easily.

if there's only one unintentional ground connection (some might call a ground fault) on an ungrounded system there will not be current flow on the ground. if there's 2 different phase conductors unintentionally grounded, there will likely be opening of ocpd if bonding is proper/adequate.

 

[Bretzel]

We found the ground fault to be in one section of the plant; an area of about 4 acres. Within this area the feeders form a loop. It appears to have been configured in such a manner to allow for shutting down blocks of service while maintaining service to the rest of the area. Nothing the POCO doesn't do. However, some of the switch gear is circuit breakers and fused switches rather than simple switches. I understand the thinking for the blocks and for the protection but I don't think the circuit breakers or fuses would act any as automatic overload protection should the entire loop be energized, as it was. Keeping the loop from being fully energized is now a procedural issue after repairs are made.

The loop itself is not the cause of the ground fault. However, in the process of investigating the loop , we found the appearance of two separate faults on the same leg within the loop. In our inspection process we shut down one block where the overhead lines were ancient, had rotting crossarms and were in close proximity to a metal structure. We found no evidence of the lines having arched against the structure but the lines had enough slack to do so in a strong enough wind. We removed these lines altogether. We tested for faults before restoring power to the area and found none. We closed one switch and the fault returned. We opened this switch and closed the switch feeding the other end of the now former loop. A fault appeared on that end as well. We have been through the 480 side of the rest on the system in this area. There is one building fed underground through 60 year old pipe on the north end of the loop. That feeder is my prime suspect for the fault at this time. We'll are planning a shut down in that area in the coming weeks. If the building service isn't the problem then we have to start looking into the secondary side of everything in the area.

The other end is all newer work; underground in 3" conduits. We lines we removed were the first section of original lines where this new work comes out of the ground at one end of the loop. This newer work has three underground vaults along it's path. We are planning on a shutdown which will allow us to open the vaults and see what is going on in them. My first concern is that there are three line sets at the beginning but only one is currently in service. Two of the line sets are 3 wire 500 mcm with a bare ground wire that appears to be 3/0. The third set of 3 runs of 750 mcm, 15Kv cable with a 3/0 ground cable. The service starts on one set of the 500 mcm cable but ends up on the 15kV cable at the other end! I am assuming, which may be completely false, there is a good reason for this such as a cable went bad and someone switched it in a hasty repair. I have bandaids on turnicates on plaster casts. How do you eat an elephant? One bite at a time.

It looks like someone was thinking about bringing the primary all the way through the plant (this point is about as far as it gets from the substation) instead of having a substation at the POCO entry point. The work looks to be about 18 years old given the production dates on the wire and conduit. It is some of the best workmanship on property but there are zero records of who did it or why. It appears there was some work done about 10 years ago by the same company which contracts with the POCO for line work when the POCO crews are two busy.

I have convinced management to install a fourth wire on all of the 480 distribution and to replace the antique substation entirely. We are working with the POCO to supply us with a 480 volt temporary feed while the substation gear is removed, the PCB contaminated concrete floor is demoed, the structural brick repairs are completed, a new roof is installed, and the POCO brings in a new permanent underground primary service for us.

I did follow one contributor's recommendation and picked up a refurbished 3M fault detector.

I'm going to have to ask for an increase in my vodka budget.

 

[ptonsparky]

:thumbsup: Sounds like you have it under control and headed in the right direction.

Thank you for filling us in on your progress.

 

[ptonsparky]

One last question...does the EE wear suspenders or did you manage this on you own?

 

[__dan]

Thank you so much for the feedback and update(s). I'm sure a lot of guys are learning a lot here.

Certainly it was long past time for a survey. It's hard to find someone with a good mind for the technical problems you have. Just gaining an understanding is a big improvement. Most guys will look at what you have and tell you the breaker will trip if there's a problem. With the floating system, the breaker does not trip when you expect it to.

The site contiguous EGC, really it's there to limit voltage differences between metal parts and structures during fault and normal use conditions. Probably the single best improvement you could make regarding avoidance of future catastrophe. An investment in the avoidance of far bigger future costs.

Just a working premise, but I would look at the drives connected to the floating Y system. I am curious. It's a good bet some or most were installed without concern for floating the drive front end, at the internal input grounding screw.

 

[drktmplr12]

OP delivered.

Nice. :thumbsup::thumbsup::thumbsup:

 

[Bretzel]

Faults Located!!

Faults Located!!

It has been a long, long battle but we have located two major ground fault and one minor one, all on the same leg of the system. One fault was in an abandoned building. We simply cut power to the building and will deal with it if and when the facility is required in the future. Another fault was in an underground PVC conduit near a metal tower which rises some 60' above average terrain and appears to be ungrounded as it is sitting on concrete piers and has no ground rods. The final one is somewhere in an overhead line heading out to the far end of the property. With these faults cleared or simply off-line we have more normal voltage readings. Phase to phase is 480 volts anywhere on the system. On this 480 volt, ungrounded delta system we should be seeing 480 between each phase and ground. We are seeing 229, 226, and 220 to ground.

Any thoughts on this situation?

Thank you in advance.

 

[LarryFine]

Phase to phase is 480 volts anywhere on the system. On this 480 volt, ungrounded delta system we should be seeing 480 between each phase and ground. We are seeing 229, 226, and 220 to ground.

Any thoughts on this situation?

Perfectly normal readings for an ungrounded system.

You're not seeing 277/277/277 because the meter loads each line slightly as you test it. A low-impedance tester should show much less each time.

That's my theory. :roll: