Power Distribution Transformer

Gentlemen: Please forward any ideas.

13.8KV/480,3PH/3W,1000 KVA, Delta-Y, Resistance Grounded serving a Motor Control Center. No Neutral Conductor.

Measured Secondary Voltage Phase to Phase 480V, all Phases.
Measure Phase Currents 80 A, 80 A, 85 A.
Measure Secondary Voltage Phase to Ground at the MCC:
P1 - 420 V
P2 - 380 V
P3 - 77 V

Any ideas about whats going on with this transformer and distribution system ?

 

[dpc]

You may have a ground on the P3 leg. This will cause P3 voltage to ground to read low and the other two phases to approach line-to-line voltage.

I'd suggest you take it out of service and look for the ground.

dpc

 

[busbar]

There is failing insulation [to ground] primarily on the third phase, potentially anywhere from the serving transformer secondary winding out to a motor stator or possibly something else like a remote 480V solenoid.

The grounding resistor is most likely connected from the 1000kVA-transformer XO bushing to system ground bus. It’s probably running fairly hot (and must be rated for continuous duty). The idea is that the 480V system will continue to operate with a single phase-to-ground fault, UNTIL another fault develops on another phase. If the fault not located, the second ground fault them you may have two overcurrent devices operating simultaneously. Current during a solid fault should only measure a few amps maximum on a 1MVA 480V system.

Do the voltage readings to ground change periodically? If they move back to balanced for a time then the fault may be downstream of a magnetic starter that cycles through its operating process. The fault can be found by a methodic process of elimination, isolating sections of the 480V system.

The resistor is crucial for damping transient overvoltages caused by capacitive phase-to-ground current that exists naturally—and will limit their damage to insulation. Instead of constantly swapping meter leads for phase-to-ground readings, you can just monitor voltage across the resistor. With normally good insulation, voltage across the resistor should be close to zero. (Conversely, you would expect about 277V with a solid ground fault.) This is good justification for having consistent and well-identified phases throughout the system.


Check thread238-3370 thread238-6870 thread237-7403 thread238-6870 thread238-9005 thread238-10998

 

[ongchazira]

Is the 480V side of the transformer delta connected? If so then there is some ground path on your third leg. This will not create the problem unless another ground develops in the third phase at any point of your system. You have to take the transformer off line and check their insulation with respect to the ground.

 

[LightningBolt]

Thanks to dpc, busbar, and ongchazira for their correct responses.

We are a little embarrased by this incident as we know what the described measurements indicate. We just could not find the groundfault.

However, after further investigation we found that the grounding resistors and associated controls mounted in the Electrical Equipment Room were mislabeled. Essentially, we were checking out the wrong MCC because of this labeling problem.

Happily, with the newly discovered knowlege we were able to quickly located a groundfault on the P3-Leg of the correct MCC in an analyzer shelter HVAC load.

Elaboration on the lessons learned from this incident are unnecessary. I am sure you guys know there are many.

Regards

Lighting Bolt

 

[ESOXmaniac]

Hope - you locate the fault, before second fault appears. This illustrates the benefit of using ground fault detection and tapped grounding resistor with pulsing contactor to locate the fault without taking down the entire system.

You can make your own pulsing circuit w/500V/10A contactor + small timer control circuit to pulse the contactor on & off to switch a test resistor in parallel with the existing grounding resistor. W/O knowing the existing grounding resistor value it's hard to say what value of test resistor to use. Measure the current & voltage across the grounding resistor and calculate new value for the parallel test resistor.

f. ex if grounding resistor is 27.7 Ohm (10A @277V) a 100 Ohm 500W test resister would increase the fault current by ~3A.

The goal is to increase the fault current at the fault by lowering the grounding impedance enough to get a measurable current change that can be measured w/ amp probe. Usually 3 to 5 A is enough. Every time contactor closes fault current will increase. This can de detected on the faulted phase and traced downstream to the fault source w/o shutting down entire system.

 

[busbar]

However expensive and time consuming, hard-earned resolutions can be very educational. Fortunately, LB's mislabeled components did not cause any immediate hazard to personnel. Some installers have little appreciation for accurately identified electrical equipment and the long-term safety ramifications of errors and sloppy practices.

As far as pulsing systems for high-resistance grounding fault location are concerned, I've found them to be a colossal waste of time. The low-range big-jaw ammeters look good in the literature, but in my experience, the fault/leakage current divides and goes a hundred different directions. For me, much more effective is a widely visible set of “ground lights” that be seen while observing processes and equipment cycling. But then a 2500kVA unit sub feeding 135 motors can require some tedious searching, and make timely fault isolation difficult and sometimes ignored.

Supporting manufacturers’ information at:

http://www.schneider-electric.ca/

http://www/en/products/ground/F0980AB9901EPR1.pdf

http://www.geindustrial.com/products/manuals/GEI-72116.pdf

 

[RJF]

We have a number of 1000 to 2500 KVA 480 volt subs with high resistance groundings systems on them and have had good luck with the "low range, big jaw" ammeters in finding the ground faults. The ground currents do split up along the many return paths but we have almost always been able to find the fault using the pulsing function of the HGR system. Of course the HGR fault detector should be alarmed and the alarm should be logged when it starts and stops. That information with some understanding of the process will often lead us straight to the problem equipment. What I have found is that most of our electricians do not understand high resistance grounding. We now have a number of electricians, technicians, foreman and engineers who understand the system well enough to help find the faults.