CDEGS

[kookypedia]

Guys,
I have 26 KA ground fault to be taken care by ground grid. I have put so much copper and rods but it doesn't change that much. Should I put ground well or anything else?
Any help would be appreciated,

 

[Power0020]

Just put few conductors in the middle of regions/squares/zones with high touch voltages.

Usually, major touch voltages appear around corners and outer regions (fringe effect)

 

[kookypedia]

True, but I have a concrete foundation in those area.
May be I should put more rods around those areas.

 

[HamburgerHelper]

You should consider expanding your grounding grid outside of your substation if you can. The L shape of your grounding grid is going to force you to use a lot of copper conductor to bring down the step and touch potentials around the edges and corners. I have worked on a substation that had a section that were fenced off from personal due to the odd shape of the substation and whoever not wanting to spend money on a robust grounding grid.

 

[HamburgerHelper]

Are you sure you are modeling the concrete correctly? Concrete usually is fairly conductive. With some mixes very conductive and it has rebar,too.

 

[jghrist]

Do you have incoming or outgoing lines with shield wires or neutrals? If so, have you used FCDIST or SPLITS to determine how much current actually flows through the earth?

You could also model pier foundations as grounding electrodes (UFER) if you provide a means to connect the rebar to the ground grid. Perhaps clipping the rebar to anchor bolts which connect to grounded steel structures. I agree with HamburgerHelper that you need to model reinforced concrete slab foundations. They will be almost equipotential surfaces without high touch-voltages. The area covered by concrete slabs could be ignored instead of modeled; it won't have the highest touch-voltage.

 

[jghrist]

You could model a concrete slab foundation as an isolated object of a closely spaced grid of wire (the reinforcing) without the complications of the surrounding concrete. This would be equivalent to assuming the concrete has the same resistivity as the upper layer of soil. You could get complicated and model the concrete as a finite volume of soil, but this won't make enough difference to be worth the effort.

 

[kookypedia]

Thank you all, I didn't know that I have to model the concrete!

 

[kookypedia]

jghrist,
If concrete has same resistivity as top soil layer, why we even model it?

 

[jghrist]

If concrete has same resistivity as top soil layer, why we even model it?

That's why I said

You could model a concrete slab foundation as an isolated object of a closely spaced grid of wire (the reinforcing) without the complications of the surrounding concrete.

 

[Power0020]

Modeling concrete can be a bit tricky, in dry conditions, concrete has a very high resistivity, but in wet conditions, it is almost a good conductor.

Some specs ask for epoxy coated reinforcement bars, making the model almost useless as the bars will have virtually no contact with soil when bonded to the main grid.

what about soil replacement?

Did you check the actual earth return current? you need to run a fault distribution study on CDEGS and fault contribution study (ETAP/DigSilent) to get the actual grid return current going to grid through neutrals for remote (out of substation) faults. This may not be easy but can help reduce the design current greatly.

 

[waross]

I suggest that you check this Wiki article to get some idea about the grounding properties of concrete. The article has a couple of links that you may want to pursue.

https://en.wikipedia.org/wiki/Ufer_ground

Excerpt:
During World War II, the U.S. Army required a grounding system for bomb storage vaults near Tucson and Flagstaff, Arizona. Conventional grounding systems did not work well in this location since the desert terrain had no water table and very little rainfall. The extremely dry soil conditions would have required hundreds of feet of copper rods to be inserted into the ground in order to create a low enough impedance ground to protect the buildings from lightning strikes.

In 1942, Herbert G. Ufer was a consultant working for the U.S. Army. Ufer was given the task of finding a lower cost and more practical alternative to traditional copper rod grounds for these dry locations. Ufer discovered that concrete had better conductivity than most types of soil. Ufer then developed a grounding scheme based on encasing the grounding conductors in concrete. This method proved to be very effective, and was implemented throughout the Arizona test site.

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

 

[kookypedia]

Thank you all for very informative post. I want to share the CDEGS help desk email in this regard too,

The concrete foundations reinforcement bars are generally connected to the grounding grid and may improve the grounding performance of the grid. The concrete itself, being a very hygroscopic material, tends to attract moisture from the neighboring soil and will eventually acquire to same resistivity as the soil it is in contact with. Consequently, you often can model the rebar only, and omit the concrete volumes.

Rebar meshing is often very dense and modelling it exactly could easily generate a very large number of conductor segments (the software will automatically subdivide conductors at intersections). This could lead to long computation time. You can model a coarser ‘skeleton’ of the reinforcing steel with a reduced number of conductors, it should not significantly affect the grounding performance of the added rebar, and if anything, it is conservative. After you have completed the design phase, you can always launch one last complete run to confirm the design while you attend to other matters.

Finally, if you want to model the rebar as steel, you can define a new conductor type and assign it to the concerned conductors. To do so, go Define|Conductor Types and define a new conductor type with Computed Impedance Specification. Steel has a relative resistivity of 10-13 (relative to annealed copper) and a relative permeability (relative to free space) that can have a wide range, we usually take 250-300. In the screenshot below, I defined a typical rebar material.

Then you can assign this conductor type to any conductor in you model. First select it, then right click over it and go to Characteristics to set the Conductor Type to the new Steel rebar type.

 

[waross]

From CDEGS
The concrete foundations reinforcement bars are generally connected to the grounding grid and may improve the grounding performance of the grid. The concrete itself, being a very hygroscopic material, tends to attract moisture from the neighboring soil and will eventually acquire to same resistivity as the soil it is in contact with. Consequently, you often can model the rebar only, and omit the concrete volumes.

Ufer grounds from Wikipedia
Concrete is naturally basic (has high pH). Ufer observed this meant that it had a ready supply of ions and so provides a better electrical ground than almost any type of soil. Ufer also found that the soil around the concrete became "doped", and its subsequent rise in pH caused the overall impedance of the soil itself to be reduced.[3] The concrete enclosure also increases the surface area of the connection between the grounding conductor and the surrounding soil, which also helps to reduce the overall impedance of the connection.
For what it's worth.

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