Substation Fence Energization

[RSChinn]

Are there any national codes or standards that say substation fences cannot be energized? We have some fences near transmission lines that have an induced voltage of between 6 and 12 volts when measured from a remote earth point.

 

[Kiribanda]

RSChinn,
As far as touch voltages and step voltages are concerned this 6V is nothing. What I will do is to lay some insulating
gavel up to 1m away from the fence. Can it be a transferred potential from somewhere else?

 

[racookpe1978]

ALL such metallic objects around substations and switchyards (including gates and low-to-the-ground objects like bridge decks and vehicle guardrails and handrails) have been continuously grounded at every facility I've worked near.

 

[Skogsgurra]

A voltage in the order of 10 V in normal operation is much too much. That can easily increase hundredfold during an earth fault and can be lethal.

Therefore, there are limits as to how much touch potential there may be (and step potential). Has any calculations or measurements been carried out? Most countries do not allow operation of plants where the touch potential is too high.

When you say "remote earth point" - how remote? And is it measured with an equivalent "human body" load or with a DMM with 10 megohm input impedance?



Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[stevenal]

As far as codes, I recommend IEEE 80. I'm not aware of any limits on potentials to remote earth, unless those potentials can be transferred into the station to become a touch potential. The usual limits apply to local touch and step potentials.

 

[fx007]

For substation fence induced voltage calc, there is an excellent manual and software by SES CDEGS. The manual also specifies limits as per IEEE and other standards. Find it, read it. Buy software to calculate.
In your case, initial voltage of 6-10V referred to remote point is meaningless, since the real dangerous overvoltage will happen under short circuit conditions of the nearby power line or at the substation. Now, how many times someone will touch the fence during the short circuit nearby, is something IEC did a study, so you might use this study to demonstrate that IEC is saying that the probability of occurrence of short-circuit event is very low, and someone touching the fence at the exact same moment which last 1-5s at most is even lower. That is if authority accept IEC.
Much as I know, no other software can help you calculate the induced voltages under this condition. Doing it manually is possible, but maybe too complicated?
Provision of 1m of gravel beyond the fence line and proper grounding/earthing of the fence and fence doors are a must, using IEEE or IEC. That is also the minimum of standard, everything else is extra, including detailed calculation with CDEGS.
I hope this will help.
Regards

 

[krisys]

If the fenced premise is supplied with electricity, the fence shall be earthed. All the fence gates shall be bonded to the fence.

The fence earthing is a subjective issue and cannot be generalised. Depending on what facility is fenced? Is it an industrail facility, A comercial establishment or a vegitable farm in the remote area or a junk yard? What is the transmission voltage etc.

If the fence is having a stray voltage, install a couple of earth rods, form an earth grid and connect the fence to the earth grid at no. of points. This should be good enough to hold the fence at earth potential.

 

[eamonkerrigan]

If this is in the US, it would fall under OSHA criteria which limits the voltage of any above grade appertenance to 50V or less. You are within this limit, but I never felt very comfortable with that 50V limit myslef. If you're measuring a constant potential, I'd question whether the fence was bonded correctly, if at all. I don't think installing stone is necessary unless you do a fault study and determine that the touch voltage under fault conditions is above the safety limits. This is a steady state issue you are desribing, so IEEE 80 doesn't really apply.

Skogsgurra raises some good questions. How was this measured? Has anyone complained about getting shocked? I think more data is needed. I have used the software referenced above (SES CDEGS software) to calculate similar voltages under steady state conditions. Since you reference a transmission line, I'm assuming you have a long metallic fence running parallel with some transmission lines, which is leading to induction issues? I'm just guessing here, but that problem can be addressed fairly easily with good bonding. Without any more information it's hard to make any recommendations, but I wouldn't just throw some rock around the fence without getting a better understading of the problem. My first instinct is that it's not an issue with not having enough insulation, it's an issue of not enough bonding.

 

[jghrist]

I assume that the substation fences are grounded to the station ground grid, meaning that the 6-10 volts also exists between every grounded object in the substation. Unless the "remote earth" is brought into the station by communication wires or some other means, the voltage to remote earth is meaningless from a safety standpoint. Even if it is, 10 volts is too low to worry about. The real concern with voltage to remote earth is transferred potential during ground faults.

 

[PHovnanian]

the probability of occurrence of short-circuit event is very low, and someone touching the fence at the exact same moment which last 1-5s at most is even lower.

I'm not so sure this is valid. As two unrelated events, maybe. But I can construct some scenarios where work inside a substation goes terribly wrong and the instinctive reaction of the crew is to run for the gate.

Actually, I've been on site when something like this happened (everyone survived uninjured).