The Physics of Elecrtrical Failure: Ungrounded Systems 3

[lengould]

My underatanding has always been that none of this matters provided the system is installed and operated to minimize (peak phase-to-ground voltage over time causing insulation breakdown and future failures x production time lost to the immediate effects of a single fault). If in a particular situation that is accomplished by a high resistance ground indicator with a good tracking policy, or a solid ground with a fast diagnostic/repair crew really makes no difference. The economics should dictate, so the question becomes "what are the economics"?


Pechez les vaches.

 

[peebee]

I mostly agree with you, lengould: the right questions to ask are "what are the functional needs/desires/economics".

But keep in mind there's code requirements for solid grounding on some systems too. . . .

 

[jbartos]

Suggestion: Reference:
1. Donald Beeman "Industrial Power Systems Handbook," First Edition, McGraw-Hill Book Company, Inc., New York, 1955.
Chapter 6 by L. J. Carpenter, and L. G. Levoy, Jr.
"System Grounding"
It reads smoothly as a bestseller.

 

[whinton]

There is a solution to the transient over voltage spikes and phase voltage instability in an ungrounded power system. Phaseback was designed for just such a problem. We have them in our facility with all 35 of our 2000 kva power transformers protected and haven't had the instability generally associated with ungrounded delta power systems. The link is:

http://www.applied-energy.us

I hope this helps
Bill

 

[mc5w]

Some electrical utilities have measured parasitic AC voltage as high as 2,000 volts on an ungrounded 480 volts system. The usual culprit for overhead services was capacitance coupling from higher voltage lines. There are some other culprits. What makes the parasitic voltage go away is to connect three 100,000 ohm or 1 megohm resistors from each phase to ground. These resistors have to have a hefty power rating to drain off the parasitic voltage without overheating. 1 megohm resistors are a rather old trick. !00,000 ohm resistors would need a 40 watt or higher rating to drain off parasitic voltages without overheating.

I have seen a ground monitor that used say 500 va or 1 kva control transformers and 130 volt 200 watt light bulbs. Ground detection light bulbs in this size range will absorb all parasitic overvoltages.