600V 3phase 3 Wire (Delta) Main Service Woes 1

[woodsd012]

I have a 3 phase 3 wire 600V service to a commercial building. The tenant experiences flickering or dimming of the lights, and her UPS's are constantly switching between battery and normal. This is consistent with a sag. Measuring on the line side of the main switch, we get:
A-B, A-C, B-C all approx 600V, which is expected.
However, referencing ground we get A-G at 540V, B-G at 587V and C-G at 59V. I would have expected 347V from phase to ground. Furthermore, why is C-phase inconsistent with the others? Our initial thoughts were that perhaps a fuse was blown and that phase was backfeeding. We have not confirmed that a fuse has not blown (with the exception of the fact that 3 phase motors are all still working, but that could be attributed to single-phasing), but is there something else I should be looking for? I have a funny feeling that something is wrong with the ground at the main board. Other than that, I think that I may have to start looking to the street and the local utility for problems with their transformers.

 

[redtrumpet]

Sounds like a single line-to-ground fault on phase C to me. That is why Vc-g is only 59 V. The voltage to ground of the other two phases is elevated by a factor of 3^0.5 (1.73). Your fuses won't blow on this fault because as of now, only the capacitive charging current is flowing to ground. However, you are overstressing the insulation of equipment and cables because your line-to-ground voltages are now 1.73 times higher than normal. This increases the likelihood of a second line-to-ground fault, which will lead to large fault current flowing and a probable blowing of a fuse.

Find the ground fault - and think of adding ground detector lights to your main service.

 

[busbar]

Check thread238-9005 thread238-3370 thread237-7403 thread238-6870 for discussions of this. There, folks talked about 480V but the same rules generally apply. Ungrounded delta [or ungrounded wye] systems can have the problems you describe.

Does the serving 600V transformer secondary (possibly utility-owned) have a solid neutral (wye-point)-to-ground connection? If not, your phase-to-ground voltage readings are at the mercy of the entire 600-volt-system insulation.

Are any lighting ballasts (or other loads) served at 347V? If yes, then you have an immediate fire hazard from a “loose neutral.”

Are the UPSs served at 600V or 208V? For 208V, then in almost all circumstances the 208Y/120V wye-point (usually an X0 bushing) must be solidly grounded; i.e., building steel/electrode.

More information about the system at hand will help yield higher-quality advice.

 

[jbartos]

Suggestion: The check for a deficient fuse is mainly that the fuse opens the circuit, i.e. there is a zero voltage of phase C to ground. Another cause could be a faulty connection, loose connection, faulty switching device, etc. in the phase C power distribution installation, which has similar effect as the blown fuse.

 

[busbar]

Dr Bartos, aren’t the phase-to-ground voltage readings (assuming they’re not ø-n) indicative of insulation degradation of cø insulation? The poster stated that ø-ø readings were all around 600V. Isn’t that the same situation as is described in Beeman chapter 6 for an ungrounded-delta system? That doesn’t seem like a fuse or connection problem unless it is served as solidly-grounded 4-wire wye (600Y/346V). If it were, it’s likely that the original poster would have found other problems like lighting ballasts in flames. A discontinuity in the neutral conductor upstream of where the voltages were measured would manifest the described problem, but the poster specifically describes the system as 3-wire delta. [In the US, I believe lighting circuits in commercial buildings are limited to 300V to ground, but the poster doesn’t give a location.]

 

[jbartos]

<

http://www.eng-tips.com/images/new.gif>

busbar (Electrical) Sep 12, 2001
Suggestion to the previous posting marked by ///\\\:
.... aren?t the phase-to-ground voltage readings (assuming they?re not ø-n) indicative of insulation degradation of cø insulation?
///Yes, it may be the case. If the delta connection has a grounded corner, which is not so clear in the original posting, then c to ground voltage may be very small, almost zero. I mentioned &quot;etc.&quot; in my above posting, since there are more possibilities left.\\ The poster stated that ø-ø readings were all around 600V. Isn?t that the same situation as is described in Beeman chapter 6 for an ungrounded-delta system?
///Yes, except the grounded delta corner, which normally also has all phase-to-phase voltages 600V.\\ That doesn?t seem like a fuse or connection problem unless it is served as solidly-grounded 4-wire wye (600Y/346V).
///I interpret a partially open circuit as one with the higher impedance in line (in series), across which the missing voltage drops.\\ If it were, it?s likely that the original poster would have found other problems like lighting ballasts in flames. A discontinuity in the neutral conductor upstream of where the voltages were measured would manifest the described problem, but the poster specifically describes the system as 3-wire delta. [In the US, I believe lighting circuits in commercial buildings are limited to 300V to ground,
///Yes.\\ but the poster doesn?t give a location.]

 

[woodsd012]

The location is in Toronto Canada. It is definitely a 600V Delta system. As far as lighting goes, it is all at 120V. As far as symptoms go, nothing has failed catastrophically yet. We will be checking C phase thoroughly for a ground fault, and I will let you know of the results.

 

[busbar]

woodsd012, your follow-up comments make sense. With a 600V ungrounded-delta system and all of the transformer, cable, motor and switchgear insulation in good shape, you should expect to see three ø-ground readings of around 346V. A solid ø-ground fault will, of course, give the faulted phase 0V to ground and the full ø-ø voltage on the other two ø-ground measurements. You have a case of ‘weak insulation’ somewhere on Cø, accounting for your readings. The location of the poor insulation will probably be surprisingly minor, but it is an invaluable way to learn about the problems unique to ungrounded-delta distribution. Ideally the system can run with one phase faulted forever, but when the next fault occurs on a different phase, then multiple fuses or breakers will operate that may, at first, seem unrelated.

If the erratic UPSs are operated at 208 or 120V, then the 600&#8722;208Y/120V transformer secondary X0-connection/ground conductor may have a problem like poor continuity or possibly more than one connection to ‘local’ building steel or similar ground electrode. Both can cause the circuit neutral voltage-to-ground to be significantly above ‘zero volts,’ and make the UPS think it’s seeing a low (or high) line-voltage condition.

 

[woodsd012]

Well, as it seems, the problem is out of the building. Rather than starting to shut everything down and start searching for the ground fault, I had hydro in to verify the integrity of their service. They have confirmed that the problem is upstream of this building at an auto shop, operating a large stamping press. We had originally thought that this might be the case, but were trying to exercise other options before getting hydro in. Thanks for all of the help guys!

 

[Wirenut]

USA....
USA....
USA.... Wirenut

 

[busbar]

woodsd012, thanks for the update. Leakage/very-low-current faults can be difficult to trace, especially when your customer is sharing a system with someone else’s electrical gear that you may not have ready access to. The fault current can be less than 1 ampere, making the process of elimination tedious. The only practical way—short of isolating loads one at a time while monitoring ø-to-ground voltage—is to use a low-range big-jawed split-core current transformer to read each circuit, with the CT simultaneously placed around all three individual circuit’s phase conductors.

It may not have been apparent, but the hydro troubleshooter probably appreciated all your efforts. This will make it easier the next time you have to deal with such a situation. And, you have experience with a special electrical problem that a lot of electricians do not understand.

Have you been able to tell if the UPS problems have also ceased?

 

[edjh2]

Try fault locating equipment that is permanently installed and notifies immediately of a fault and its location!