I'm not sure what you mean by limitations. IEEE Std 80-2000 discusses non-homogenous soild at some length in Section 13. Equations for calculating step and mesh voltages are given only for homogenous soil, but reference is made to computer algorithms for calculation with multilayer soils.
As jbartos has indicated, there are numerous limitations. Perhaps we can help if you are more specific in what kind of limitations you are referencing.
Limitations i am referring to are those such as grid geometery, grid size. Annex D lists a variety of grid shapes that can be accomodated. What i really want to know is in what situations would the equations for calculating maximum step and touch voltages and grid resistance fail.
Homogenous soil.
Square, rectangular, symmetrical T-shaped, trangular, or L-shaped grids.
Uniform spacing of parallel grid conductors.
While I don't see them in Appendix D, Thapar's paper [B144] that is used as a reference for the calculation of n in eq. D.6 has the following limitations:
0.25m <= h <= 2.5m (depth of grid)
d < 0.25·h (diameter of conductor)
D > 2.5m (spacing between parallel conductors)
I am wondering why these limitations werent given. Thanks for point them out.
What about grid size?? The guide also states that irregularity factors were added/adjusted to accomadte ground rods and produced favourable results. From you experience to what extent these corrections can be used?
Comment: There are limitations given by availability of rod sizes, e.g. diameter, length, material copper or galvanized steel, etc.
It is necessary keep in mind that some high soil resistivities will cause requirements for ground grid density, grounding rod diameter and length, etc. unreasonable. Then, alternate measures need to be assessed, e.g. soil treatment to reduce the soil resistivity.
I'm not sure why the limitations were not given; they should be. IEEE Std 80 seems to be de-emphasizing manual calculations now that commercial software programs are available that run on PCs and are well within the budgets of utilities. The programs are very expensive for individual engineers or small firms, however.
Before we got the SES grounding software, I made extensive use of the simplified calculations with the irregularity factors. Although I haven't made many comparative calculations between SES and the simplified calculations, generally I think that my simplified calculations resulted in conservatively designed grids.
I can say that since getting the software, I have not found any situation where homogenous soil is a good assumption.
I don't think there is a limit on grid size. On large grids, being overly conservative is more expensive.
Restricting the design to evenly spaced parallel conductors forces the use of an uneconomical grid design because it is much more efficient to use closer spacing near the perimeter of the grid, especially in soils with a low resistivity upper layer.
Comment: Some utilities are still using old approaches to the substation grounding, which do not adhere to IEEE Std 80. Apparently, doing it old way can save some amount of copper and labor. However, the safety has to be watched more carefully.
