Earthfaults, Delta Connections, and man safety

[Rodmcm]

If you look at many generation plant connections from MV to HV then the MV (say 11kV) bus is often unearthed until the generator synchronises.

The popular idea is that there is no earth return from the MV to the HV and hence, no fault current, no step and touch potential issues, and little man safety risk. The usual protection is a broken delta or neg seq voltage to detect an earth fault.

This makes me nervous, as inherently I do not like any unearthed systems at high voltages. It is easy to create a managed earth with an zig zag transformer or similar.

So, looking at it from a pure safety point of view, is there a risk? Is there any possibility for earth return current, and how would you calculate it or assess the risk?

Also, I would be interested in views on broken delta protection. A protection engineer I know thinks they should be avoided at all costs!

Interested in the thoughts from the collective wisdom of this forum

 

[ScottyUK]

Thninking aloud and assuming this relates to a large generator with a conventional isolated phase bus between the machine and the GSU transformer with a 3-gang single-pole air blast or SF[sub]6[/sub] generator breaker, consider the the credible failure modes:

[ul]
[li]Phase-earth fault within generator[/li]
[li]Phase-earth fault within GSU transformer[/li]
[li]Phase-earth fault within IPB[/li]
[li]Phase-earth fault within breaker[/li]
[/ul]

[ul]
[li]Phase-phase-earth fault within generator[/li]
[li]Phase-phase-earth fault within GSU transformer[/li]
[li]Phase-phase-earth fault within IPB[/li]
[li]Phase-phase-earth fault within breaker[/li]
[/ul]
[ul]
[li]Phase-phase fault within generator[/li]
[li]Phase-phase fault within GSU transformer[/li]
[/ul]
[ul]
[li]3-phase-earth fault within generator[/li]
[li]3-phase-earth fault within GSU transformer[/li]
[/ul]
[ul]
[li]Inter-turn fault within generator[/li]
[li]Inter-turn fault within GSU transformer[/li]
[li]Inter-winding fault within GSU transformer[/li]
[/ul]

Of those only the first four would fit the criteria you're worried about where there is voltage displacement but no short circuit. The remainder would be detected using conventional techniques.

Current will flow in the phase-earth fault due to capacitance because the earth fault will upset the normal near-balanced earthed-Y connection of the capacitances. The current can be estimated reasonably easily, but is unlikely to be significant in terms of equipment damage. Obviously the insulation in the non-faulted phases will be exposed to [√]3 x its normal operating voltage but this can be dealt with through appropriate design and selection.

Without significant current flow you won't get step or transfer voltages because these arise because of the passage of fault current through the resistance of earth, and the fault current isn't there to produce them. If the fault involves a second phase or all phases then the curent involved may be very large and step, touch and transfer potentials may exist until the fault is cleared as happens elsewhere in similar situations.


Of course if you are considering embedded machines on a distributor's network then some or all of my assumptions may be wide of the mark.

 

[Rodmcm]

Thanks for that. Yes the question has arisen because of embedded generation, and an open 11kV line feeding a transformer up a hill.
The problem is that nearly everyone you talk to knows of the isobus concept and unearthed connection, but there are many instances (hydro in particular) where the transformer is at the top of a hill and the generator at the bottom of the gully, connected by overhead lines

 

[waross]

I prefer the wye:delta scheme to a zig-zag scheme. The wye:delta uses off the shelf transformers, often small distribution transformers.
The zig-zag will probably be a special order.
With a wye:delta scheme, your fault current limiting resistor may be from the star point of the wye to ground and the delta closed, or the wye point may be solidly grounded and a resistance may be inserted in the broken delta to limit the ground fault current.

Bill
--------------------
"Why not the best?"
Jimmy Carter