Fault Condition Voltage Rise

[sellakc@gmail.com]

Would love some input from the pros!

Scenario - Metal clad 13.8KV switchgear mounted on a concrete floor that has a properly installed 4/0 ground grid within the concrete floor. Grounds come up from the grid to the grounding bar in the switchgear. For argument's sake, let's say that the gear has 18 kA available fault current on the line side of the gear.

The gear is part of a local industrial plant distribution system, and is in an outage - this gear is opened, verified, LOTO, and grounded per 1910.269 section M and N.

In this condition, if somehow the gear was either backfed while grounded, or the line side switch intentionally switched back on (insert disgruntled employee here), how much voltage would be present from the concrete pad to a phase? Would there be significant voltage rise during the fault (which would be extremely short duration obviously with a fully-shorted and grounded system). Would this potential rise be if a person was touching a phase insulator for replacement at the exact time that power was applied to the system?

Consider this ala 1910.269 Section that requires T&D grounding and creation of an equalized potential zone (1910.269(3) as part of the grounding protocol.

 

[jim dungar]

Is this a real life situation or a test question? It sounds like it might be part of some type of investigation.

 

[sellakc@gmail.com]

Is this a real life situation or a test question? It sounds like it might be part of some type of investigation.

Actually neither. It was a question that came up during a training session at a data center.
I’m thinking that there should be little to any voltage development phase to ground during the fault, as the grounding cables are hooked to the grounding bar of the gear, and to the rebar in unit concrete.

 

[steve66]

It all depends on ohms law. If the resistance of the ground is much less than the resistance of the source, the voltage will be less.

Actually, a lot of this depends on if there is a ground conductor ran with the supply feeder, and what size is it? Or are we counting on the earth as the ground return path? The voltage will be much higher if we are counting on the earth.

 

[sellakc@gmail.com]

Totally agree. In this case, the site has a grounded wye secondary, hooked directly to the grid (no grounding impedance resistor). The concrete encased rebar was included in the ground grid before concrete was poured. So effectively, the worker is standing in concrete with an encased ground, and the temporary grounding cluster is shorting all 3 phases and hooking them to the ground bar.

Their question was the concrete impedance. Would it cause any significant voltage rise during the accidental re-energization? I’m thinking not - just wanted other insights.

I suppose if the site had resistive grounding, it would greatly affect the answer, as ground fault current would be limited to typically 10 or 100 amps, according to the type of resistance level of the grounding resistor.

 

[EC Dan]

Not sure why the concrete matters, virtually all the current will flow through the grid back to the source. The voltage at the point where the operator is touching a phase conductor and where the phase conductors are grounded depends on the resistance of the return path. Let's assume it's 0.01 ohms. We know the source impedance based on your voltage and available fault current. Adding the return path reduces the fault current a bit. The voltage at the fault (where the phases as grounded) will then be the product of that reduced fault current and the return path resistance, which doing some quick and dirty math could be >100 V.

 

[steve66]

I agree about the concrete. Almost all the fault current will flow back to the source through the ground conductor.

For example, assume the source xformer is 100' away, and there is a #1 ground wire, and we just happen to know there is 20 KA of fault current available. Also assume the ground jumper is 10' of #1 wire.

When the power is turned on, we could have 20KA * .0154 ohms. (I'm using the DC resistance of #1 wire from table 9, and dividing by 10 for 100' of cable, instead of 1000'). We know the AC resistance and surge impedance will be higher, but lets ignore that for now.

So 20KA * .0154 = 300 volts (approx.) So across our 10' jumper wire, we will only have 30 volts - (one tenth the distance = one tenth resistance and one tenth voltage drop.)

So our worker who is holding a phase conductor and standing on concrete is only exposed to 30 volts. Assuming he isn't standing on wet concrete with really crappy shoes on, I don't think he will feel anything.

Question #2: Is our disgruntled worker who turned on the power getting an arc flash blast?

 

[sellakc@gmail.com]

Excellent! Makes total sense. Thanks! And I’m sure this mystery worker is in for a not so good surprise! - and most of these sites have 4/0 cables bonding straight from the grid in the concrete to the bus bar in the panel. This means that the potential 30 volt buildup would even be less due to its lower impedance, right?

 

[JoeStillman]

There are methods to calculate such a voltage rise in IEEE-80. There is a module in SKM for this too. For a given soil-resistivity, available fault current, ground grid AWG and spacing, the step and touch potential are known.