issue with 480 volt ungrounded system

Status
Not open for further replies.
G

glcc elec

Member
Location
ludington mi
I have a problem with a 480 volt ungrounded system at the foundry were i work we keep having a voltage issue that keeps showing up and going away when you check phase to phase between any two you read 480 volt when you check A phase to ground you have about 600 volts, B phase to ground the voltage eceeds the voltage on my 1000 volt fluke meter, C phaseto ground 1009 volt. last night we had it happen again a it stayed that way for about 45 min then just went away any ideas were to start looking for the problem and could it be atransformer causing this
 
glcc elec said:
i do not believe ther is a neutral on this system it is just a 3 wire system
Click to expand...
I missed the ungrounded part sorry. How is the system fed. Is it a utility transformer or customer trany? Hopefully someone else will jump in as my experience is lacking here but it sounds like a bad or eradict trany. Not sure that is possible.
 
As I recall, those readings are not unusual using a Fluke on an ungrounded system.
With some resistive loads to ground, the voltages stabilize.
I'm sure one of our members will advise, but I have seen resistors installed for that purpose but I don't recall values.
I recall when using an old Triplett meter that voltages quickly stabilize.
 
glcc elec said:
I have a problem with a 480 volt ungrounded system at the foundry were i work we keep having a voltage issue that keeps showing up and going away when you check phase to phase between any two you read 480 volt when you check A phase to ground you have about 600 volts, B phase to ground the voltage eceeds the voltage on my 1000 volt fluke meter, C phaseto ground 1009 volt. last night we had it happen again a it stayed that way for about 45 min then just went away any ideas were to start looking for the problem and could it be atransformer causing this
Click to expand...

I have gotten some unbalanced phase voltages off a transformer with an ungrounded secondary and no load on the transformer but what you are describing is off the charts.
 
I have seen this on delta primary feeding a transformer with a delta secondary, and the MGN was lost down the road up on a pole, if the transformer has a primary winding starting to leak to case it can cause the MGN to rise in voltage and can be dangerous as there is no fault current path back to the zig-zag grounding transformers used by the utility to derive a neutral that is bonded to the MGN, I would get the utility involved but expect them to not find anything if it's not happening when they come out,
it took me several times to get it to happen when they were there, the case above happened after a lightning strike that both cause a connection on the MGN to start failing and a primary winding in the transformer bank to intermittently start leaking, they first found the bad MGN connection but after fixing this, the next time the winding went into a case fault it tripped the primary fuse so they had to replace the transformer tank that was bad.

Check your system grounding to electrodes and very important if you have a city supplied water system with metal piping that the bond to it is good, as this can limit the voltage from the primary if such a fault occurs, as a fault like this can cause the system grounding to go as high as the primary if there is no fault path and can be very dangerous.

Also many DVM's will read high frequency signals as voltage error's so make sure your not getting HF being imposed on your 480 volt system from a HF source (such as VFD's), try using an analog meter when it happens.
 
I was thinking a possible transformer issue also so i told my boss to call in acompany to choeck them so we could atleast either confirm or eliminate the transformers but all the company had them do was put in apower monitering analiser as far as a vfd there is only one on that system an i isolated that and we still had the problem I guess another question would be could it be a light fixture causing this issue some lights were un hooked the first time we had this issue and the problem still came back so they were hooked back up some unhooked some different lights today and said tese ones must be the problem
 
130205-1459 EST

glcc elec:

Suppose a corner of the delta was grounded, then the maximum voltage from either of the other two legs to ground is 480 V RMS. If there are unbalanced resistive loads from the 480 bus to ground, then the voltages from each leg to ground will be between 0 and 480.

Thus, you may have a serious problem.

First, put a screwdriver in the earth some where, best outside the building a ways. Connect a long insulated test lead to this ground test point. You can use connected extension cords or just reel off some small wire, #24 would be fine. This connects to one terminal of your voltmeter. The other voltmeter input should be a clip lead to another screwdriver. With your meter and the second screwdriver go around and test various building conductive parts, machines, and other earth points inside and outside the building.

I don' know what to expect in your case. I only have a few tenths of a volt in my house and backyard. In a CNC application I saw a couple volts difference between two machines, and it was impossible to do RS232 communication to one of the machines. Ground rods or not made no difference, and I would not have expected the ground rods to be a solution. I have seen in the past a comment that in some plant a ground wire on an RS232 cable would draw a spark when both ends were grounded and broken.

This type of measurement would be my starting point.

.
 
glcc elec said:
I was thinking a possible transformer issue also so i told my boss to call in acompany to choeck them so we could atleast either confirm or eliminate the transformers but all the company had them do was put in apower monitering analiser as far as a vfd there is only one on that system an i isolated that and we still had the problem I guess another question would be could it be a light fixture causing this issue some lights were un hooked the first time we had this issue and the problem still came back so they were hooked back up some unhooked some different lights today and said tese ones must be the problem
Click to expand...

A defective ballast shorting to ground. That may explain the erratic voltages, as the windings on the ballast expand and contract the resistance of the fault changes, hence changing the voltage readings.
 
As I recall a ground fault on one phase of an ungrounded system that is making and breaking rapidly, will cause high voltages on the the system.
 
gar said:
130205-1459 EST

glcc elec:

Suppose a corner of the delta was grounded, then the maximum voltage from either of the other two legs to ground is 480 V RMS. If there are unbalanced resistive loads from the 480 bus to ground, then the voltages from each leg to ground will be between 0 and 480.

Thus, you may have a serious problem.

First, put a screwdriver in the earth some where, best outside the building a ways. Connect a long insulated test lead to this ground test point. You can use connected extension cords or just reel off some small wire, #24 would be fine. This connects to one terminal of your voltmeter. The other voltmeter input should be a clip lead to another screwdriver. With your meter and the second screwdriver go around and test various building conductive parts, machines, and other earth points inside and outside the building.

I don' know what to expect in your case. I only have a few tenths of a volt in my house and backyard. In a CNC application I saw a couple volts difference between two machines, and it was impossible to do RS232 communication to one of the machines. Ground rods or not made no difference, and I would not have expected the ground rods to be a solution. I have seen in the past a comment that in some plant a ground wire on an RS232 cable would draw a spark when both ends were grounded and broken.

This type of measurement would be my starting point.

.
Click to expand...

I believe the OP is talking about an ungrounded delta, although he did say he has lights on this system? are the lights 277 volts or 480?, I have not seen many 277/480 Y's ungrounded?

Even an ungrounded delta that has a single phase to ground fault will only give you 480 volts like you said, and as Don has pointed out a bad or loose connection, can cause inductive spikes especially when power factor correction caps are used, but that should show up in problems with motor loads and or flickering lights and other loads.

A recorder monitoring the system but instead of connecting the ground to the system grounding connect it to a small rod installed like Gar has pointed out, might show up a utility problem when it happens the next time.

Since a ungrounded delta does not use the MGN for current like in a grounded system, if the MGN does have a bad connection it can cause strange voltages to appear on service grounding system if there is a connection from the MGN to the premise grounding network, as I said before this can be a shock hazard as well as cause insulation damage to motor windings as well as conductors if it happens for prolonged periods, but because it is of a very intermittent nature it can be very hard to track down and will require the recording of several different places in the system to confirm what is happening, one such place that monitoring should be done is between the system grounding and remote Earth, as if you find that when it happens you have any elevated voltage on the system grounding then you know to look at the utility side for problems.
 
I have never worked with delta 3 wire so If you have a ungrounded 3wire delta 277/480 where do you get your nuetral from for your lighting and receptical circuits a transformer from inside the building?
 
liquidtite said:
I have never worked with delta 3 wire so If you have a ungrounded 3wire delta 277/480 where do you get your neutral from for your lighting and receptacle circuits a transformer from inside the building?
Click to expand...

First a delta does not have a neutral and is for 480 volt 3-phase loads only, the reason behind the use of ungrounded deltas is mainly for continuation of power in the event of a line to ground fault, if a ground fault detectors installed as required in the NEC, it will sound an alarm but no OCPD will open until a second fault happens, this allows for the safe shutdown or continued operation till an outage can be scheduled so production can continue then the repairs can be made.

Mainly used when the loss of power can be dangerous or can cause damage to the machinery or production is too costly to shut down, there are other reasons but these are the main ones, another benefit is that the utility can loose a phase but the delta will continue to provide all three phases as an open delta, a Y will not do this, so industrial plants that have a high 3-phase motor loads will commonly have a ungrounded delta service for these loads and can have separate services or transformers to serve the 120/208 loads.

While a Y can be ungrounded it is just not common for it, and a Y is the only way to achieve 277 volts, but an ungrounded Y will still provide 277 volts from the X0, it would just not have a reference to ground.
 
augie47 said:
As I recall, those readings are not unusual using a Fluke on an ungrounded system.
With some resistive loads to ground, the voltages stabilize.
I'm sure one of our members will advise, but I have seen resistors installed for that purpose but I don't recall values.
I recall when using an old Triplett meter that voltages quickly stabilize.
Click to expand...

Ditto.
 
130206-1516 EST

hurk:

I understood it was an ungrounded delta. My suggested ground voltage measurements were to try and see if there was substantial ground current, or other voltage sources that could be spotted by the measurements. And as you said it needs to be emphasized that there could be shock hazards present.


liquidtite:

120 and any other desired voltage is obtained from a transformer with a 480 V primary from one phase of the 480 supply.

Every machine tool control cabinet in a plant will have a main 480 fused main disconnect and from this a multitude of fuses or circuit breakers feeding various things like motors, drives, and in particular a transformer from 480 to 120 to supply the 120 devices and control circuitry. In the old relay days the 120 secondary was floating and a pair of pilot lights were used to indicate a ground fault. Today with programmable controllers it is common to ground one side of the 120.


To produce voltages greater than the nominal 480 you will need resonant circuits to ground, or inductive circuits being opened, or possibly some other type of generator. But another generator (voltage source like power company neutral problems) is what may be present here.

If the problems are high frequency noise, then a filter at the meter input may help sort this out.

.
 
Status
Not open for further replies.
Top