Calculating Ground Potential Rise in an indoor 12.5kV substation 4

[jawadr]

I am trying to calculate ground potential rise to make sure it is not over 5000V. The substation is in the basement of a high rise with concrete pad. Incoming is 12.5kV and stepped down to 600V. As per the coordination study, it has 13.4kA of fault current available. There are 6 groundings rods in a 49feet x 20 feet rectangular configuration with an average resistance of each grounding rod being 1.4 ohms with respect to utility ground. Groundings rods are connected with a copper wire, and there is no mesh under the concrete pad. I have been trying to figure out the ground potential rise calculations and step and touch voltage calculations without any of those expensive programs. Has anyone used a quick and easy method of calculating this? Any help would be appreciated.

 

[arango]

There's no need to use expensive programs, you can do this by using ATP and Microsoft Excel, following the procedure proposed in the IEEE 80, the only thing else that you need is the soil resistivity. I use this very often and it has turned out to be very accurate.

 

[jawadr]

I do not have access to IEEE Std. 80 yet. Is there any way to do the calculations though without soil resistivity. The concrete has been poured so we cannot get any more soil readings. I was reading through the code that mentions we cannot have more than 5000V in ground potential rise. And i have the formula to check for tolerance values for step and touch voltages. Do i just have to get someone to take step and touch voltage readings to make sure they are under the tolerance values?

 

[goodchoi]

GPR = Sf * If * Rg
(Sf : fault current division factor = Ig/If,
If : ground fault current, Ig : ground(or grid) current,
Rg : ground resistance)

For example,
If = 13.4 kA, Rg(6rods) ~ 1.4 ohm / 6 = 0.23 ohm,
Sf = 1.0 (extremely conservative assumption)
GPR ~ 13.4 kA * 0.23 ohm = 3.12 kV < 5 kV

However, I suspect your fault current is too high.

Zsource= VLN / Ifault = 12.5 kV / 13.4 kA=0.93 ohm

I think this is too small....

 

[jawadr]

Thanks goodchoi,
Gives me a starting point to work on. In your experience, what is a good Sf value to consider for moist soil (0.8)?

 

[cuky2000]

Why you main consern is the GPR?

Probably this is OK. Most installation like the one described do not require calculations.

Beware the step and touch potentials are more relevant from the safety stand point. Since the concrete resistivity could be estimated as low as 30 ohm-m, this have significant impact reducing the allowable potentials.

 

[jawadr]

I thought I had the step and touch potential figured out but as it seems, I can only calculate the tolerance values for touch and step potentials. From my research, concrete cured for 200 hours can give upto 250 ohm-m. From my formulas, i get the following tolerance values Estep=410V and Etouch=226V.

I am however confused on how to calculate what kinds of step and touch voltages there would be in the room to make sure they fall under the tolerance values? Any formulas i seem to find require soil resistivity that i dont have as the concrete is already poured.

 

[jghrist]

The reinforced concrete floor will be essentially an equipotential surface with no significant step- and touch-potentials. There is no hand calculation method that I know of to calculate step- and touch-potentials on a concrete slab.

I think the only step- and touch-potentials that may be of concern will be outside the building with someone touching the building while standing next to it (touch-potential) or standing with their feet apart next to the building (step-potential). Again, no simple way to calculate. I would put a ground ring around the building which should reduce touch-potentials enough.

 

[tin2779]

From what ever little understanding of grounding I have :

Equipotential surface can handle any possible step potential issue but not touch potential issue.

Concrete is still has some conductivity and when a fault occurs and if some body is touching the enclosure, he will get a shock depending upon the current level, standing close to switchgear.

I guess I just said in a little different way what "Jghrist" said.

But Jghrist, This "Grounding Ring" around the building, what this will comprise of: Is is the same way, three feet out side the building lke substation, One road at each corner ?

Thanks

 

[cuky2000]

Jawadr,

I believe you are not near to resolve the problem yet. The calculated allowable step and touch potential need to be compared with the actual potentials. This is not an easy task to be calculated manually.
Even thought a sophisticated computer program is available, the accuracy of the calc is as good as the model used. Several representative data need to be consider such as soil resistivity model (single or multi layers), grounding grid configuration including rebar, concrete resistivity, allowance for SC current growth, current division factor, etc.
Beware that the concrete is a hygroscopic material expected to have low resistivity closer to moisture source and extremely high resistivity in dry condition. It is expected lower resistivity closer to the external building wall. In the center of the building the resistivity could be very high.
To estimate the current division factor is primarily a function of the number of incoming and outgoing grounding conductors.

 

[jghrist]

If the switchgear is bolted to the floor, then it will be at the same potential as the floor, at least where it is bolted. If there is a 12.5 kV ground fault to the switchgear, practically all of the return current will flow either through the switchgear bonding conductor to the grounding electrode or through the 12.5 kV system neutral. Very little will flow through the concrete floor, so there will be very little voltage gradient in the floor. Because of this, I would not consider any step- or touch-potential within the building.

You would have to do a complicated analysis to prove this, but how many of you calculate the touch potentials within substation control buildings? There is equipment in there that is bonded to the ground grid, just like substation steel. I've modelled control building floors in substations as isolated grids, representing the rebar not connected to the grid. I haven't modelled the concrete as a separate resistivity volume, but concrete is usually lower resistivity than the upper soil layer (which would lower voltage gradients), so ignoring it is conservative. Modelling the control building this way results in low gradients within the building area.

By ground ring, I mean a conductor around the perimeter of the building. Like the ground conductor around a substation, but NEC requires a ground ring to be buried 30" instead of the normal 18" in substations.

 

[rcoloradoemestica]

go to

http://www.usda.gov/rus/electric/bulletins.htm

and download 1724E-300 on chapter XI it shows how to calculate the ground grid

Ricardo