Touch and step voltages with shorcircuit current at low voltage side 6

[lume7006]

Hello everybody,

When you have a large industrial plant extended over a large area there are typically several electrical rooms, each one has its own grounding grid, all of the buildings have a ground ring and the main substation has another grounding grid and all of these elements are joined together.

IEEE std 80, states that in order to size the conductor to be buried the shorcircuit current on the secondary side of the main transformer (at the main substation) should be considered but, to determine GPR, touch and step voltages only primary side shorcircuit current of the main transformer has to be considered.

However, when you have a large industrial system and all the grounding grids, rings are joined together, we surely have low voltage transformers (460-480 V) wye grounded which shortcircuit current has a very high magnitude.
Then, my questions are:
Will these low voltage side shortcircuit currents have to be considered to size the conductor to be buried?
Do we need to determine step and touch voltages in low voltage electrical rooms with the corresponding shortcircuit current?

I ask this, because, doing the evaluations with low voltage side shortcircuit currents the size of the conductor, step and touch voltages have higher magnitudes and I do not know if my previous description is taken into account to design the grounding grids of industrial systems or is not.

Any help or comment will be appreciated!

 

[memos1006]

I think the analysis should not have been done in that way!

 

[rcw retired EE]

Grounding and bonding systems must be sized to handle the maximum fault current in a plant. In many industrial facilities, the maximum current is from the 480V system with ground fault currents 65kA or higher. The bonding jumper for the X0 low voltage neutral bushing to ground has to handle that current, as does the equipment grounding and bonding conductors that bond the switchgear, the MCC's, motors and other equipment enclosures that might become energized. If the design follows NEC code, the equipment bonding conductors will be adequate to carry the ground fault current.

If the design meets the NEC, the earth will not be used for fault current return. Fault current returns to the source on the equipment grounding conductor or bonding conductor that runs with each circuit.

Since it doesn't flow in the earth, the 480V ground fault current will not contribute to step or touch potential.

But assume that a bonding jumper breaks and 480V fault current is flowing in the earth. Assume the grid has 0.5 ohms impedance. 65kA x 0.5 ohms = 37.5 kV for a GPR. That high a GPR can generate dangerous step & touch.

This is where the logic breaks down. How can a solidly grounded system with a maximum of 277V to ground create a GPR of 37.5 kV? Where does that extra 37.2 kV come from?

Obviously, that is not possible. The maximum touch voltage a 480/277V system can generate is 277 volts.

The moral of the above misleading dissertation is to demonstrate how we can forget about fundamentals and arrive at incorrect conclusions.

Design each part of the grounding system to safely handle the expected currents at that location. Design the grid to control step and touch voltages based on the high voltage current that returns to the remote substation through the earth.

 

[UtilityGy]

Bravo RCWilson,

You had me for a sec. while I was reading this post and I said to my self How you can come to 37.5 kV but you infact were alluding to the misnomer.

I actually myself learned grounding from this post. You and Jghrist had contribution towards it. I recall reading your posts written four years. I salute all people like your self who spend so much of their time answering questions.

 

[radug]

rcwilson,

I am not in NEC world, but there are some new projects for CSP plants that can arise in my company for the USA. For large motors, both in MV and LV, we have usually used the grounding grid for fault current return.

You said "If the design meets the NEC, the earth will not be used for fault current return". So, is it not a common practice in large industrial plants or generating stations to use the grounding grid as fault return? Where does the NEC require that?

Thanks.

 

[rcw retired EE]

NEC Article 250.4 "General Requirements for Grounding and Bonding - (A) Grounded Systems - (5) Effective Ground-Fault Path - ...The earth shall not be considered as an effective ground-fault current path." A low impedance ground fault return path is required. The earth alone does not meet the criteria.

I believe most inspectors would consider a ground wire going into the earth as using the earth as a fault return path, even if it was welded to the grid.

The green or yellow/green wire that is run with circuit conductors or the electrical conduit or cable tray is the usual fault return path. The ground grid is another fault return path, but it has a higher impedance due to the separation from the phase conductors.

Our designs bond all MV and larger LV motors to the grid in addition to bonding the frame through the green wire and/or the raceway (conduit & tray). The connection to grid is to control possible touch voltage issues at the motor during a winding fault.

 

[lume7006]

Hello everybody,

I also want to thank RCWilson for his valuable contribution.

Your explanation was quite clear!

I understood that I had a misconception about grounding, and it clarified what IEEE std 80 states: "only HV side of main substation will contribute to GPR".

Best Regards,

 

[alehman]

OK, I'll be the one to throw some cold water on this thread (hopefully not while it's energized).

Since it doesn't flow in the earth, the 480V ground fault current will not contribute to step or touch potential.

I disagree with this statement. Say you have a fault to the frame at some motor with an equipment grounding conductor, but no local earth connection. Assuming all of the fault current returns via the EGC, then it seems to me that the frame of the motor will be at some fraction of the phase voltage, Vtouch = Vphase * Zground/(Zphase + Zground), relative to ground.

Alan
“The engineer's first problem in any design situation is to discover what the problem really is.” Unk.

 

[rcw retired EE]

Alan, You are correct. There could be a potential difference between the equipment case and the person's feet creating a hzardous touch potential if the local bonding/grounding was not present.

I was thinkng more of how grounding programs analyze ground grid designs to calculate step and touch potentials. The programs usually don't account for current flow in above ground grounding conductors.

A good engineer will look at the bonding of all components that are likely to become energized to mitigate those touch hazards.