I'm often asked about Reduction factors or 'Split factor' when looking at Earthing Designs. So what do these factors refer to? Why are they necessary? And when should you apply them?
What is a Split Factor (or Reduction)?
In Earthing terms, the Split Factor can be defined as the 'split' or proportion of fault energy that chooses to use soil conduction as its preferred route back to the source, compared with how much of the energy takes alternative route(s). Alternative routes could be the cable sheath arrangements, CPCs or any other conductive structures intentionally or unintentionally bonded.
These split/reduction factors are a crude approach originally conceived to simplify manual calcs and overcome less capable versions of Circuit Theory based software tools. Why? Before the arrival of more recent electromagnetic (EM) software. Most EPR (Earth Potential Rise) calculations were either done manually (groan) or by using lesser capable Circuit Theory based software tools. However, it soon became apparent that while calculating the EPR of a simple earth mat is relatively straight forward, the situation becomes far more complex when one introduces the impedance contribution from multiple connected objects that sit in, or on, real-life soils, i.e. multilayered soils (geology).
Take, for example, an electrified rail network with all its interconnected overhead line equipment (OLE). Or, the interconnected route of overhead towers (sometimes known as pylons or robots in parts of Africa). Imagine all the interactions that would need to be examined/calculated along the route, and then factor in 3 or 4 different resistivity layers of soil at varying depths. Complicated, right?!
The capability to model ALL these electrical interactions including the above-ground OLE simply has not been widely available until very recently. This more recent technology (EM software) has been around for years but the take-up due to prohibitive licensing cost has meant that lesser capable versions have enjoyed more market penetration than perhaps they deserve.
So, what does a Split Factor look like?
Here's an example of a Split factor in action:
- 25 kV input Voltage
- Single-to-Ground Fault is 10 kA, cleared in 200 ms
- The Earth Electrode resistance is 1 Ω (Re)
- For the purposes of simplicity, let's say our target EPR for system design, is 1 kV
WITHOUT a Split factor/reduction factor, the EPR (Earth Potential Rise) is EPR = 10 kV, right? It should be fairly simple - just apply Ohm's-law... V = I x R, to check. EPR (V) = Fault (I) x Electrode (Re)
This 10 kV is an order of magnitude greater than the target EPR of 1 kV. So, would need a lot of additional hardware to be installed in the dirt in order to lower the Resistance (Re) to achieve the 1 kV target. In addition, the relationship between the amount of copper needed to lower the Re in the soil is not linear. It actually takes exponentially more copper in the ground to lower Re until 0.1 Ω is reached. And that is likely to be expensive = cost lots of effort = deep pockets required.
For the Earthing Designer facing this scenario. He/she would have to calculate energising the Earth Electrode with 100% of the Fault energy (10 kA in this case). This leads to a worst case EPR of 10 kV. Trying to sell the idea of lowering the Resistance by a factor of 10 to the person who holds the purse-strings (budget owner) after they've seen what the final Earthing Design looks like at worst case levels is going to be a big ask!!
Clearly, the conductive above-ground paths are going to have a big influence on what remaining fault energy goes to ground; but the actual percentage will depend on many things, not least, the unique soil resistivity and geostructure locally, the geometry of the local electrode, the current density, as well as the overall impedance of the alternative path(s) to name but a few.
So, this is where the Split Factor makes its entrance. The split is a simplification to make these calcs more palatable.
Example - Split applied
Compare the above case, but now consider a split of say 80/20. That is, 80% of the fault energy takes the conductive path(s) and 20% takes goes to ground (soil conduction).
Now. For a 10 kA fault level in the example, that means the Earth Electrode has only to be energised in the calculations with 2 kA (20%). The remaining 8 kA can effectively be ignored and assumed to take the above-ground or alternative path(s).
By using this crude but simplistic Split factor approach, the resulting EPR is cut to 2 kV in one stroke, from the 10 kV previously.
This now means, the Earthing Designer has a much more attractive proposition to present to the civils project team now!! Apparently, less to install, less to finance, a much easier sell to the project manager. But wait!... that's an 'assumption', right?! How is that good engineering practice?
When to use a Split Factor
Manually calculating a split factor can be exceptionally complicated and test the mathematical skills of the most gifted. So, for the most part, it's easier / sensible/reasonable... take your pick. For people not armed with the sharpest tools, to approximate how much of the fault energy will choose to head down the more conductive path(s). Versus, how much will go-to ground via soil conduction through the Earthing Electrode. This is what the Split or Reduction Factor is about - a split is a generalisation, a simplification, an assumption or in some cases, an outright guess.
Even today, those that hang on to less capable versions are constrained to very simplistic buried-grid only representations of earthing arrangements. Most legacy modeling software tools only have the ability to build super-simplistic representations of earth grids in the dirt. Other connected structures or objects above ground are conveniently ignored. At best maybe there's a lumped impedance representing connected objects but this lacks the physical geometries, material properties and some of the other key electrical characteristics to bring them to life (so to speak). And herein lies the problem.
When NOT to use a Split Factor
Split factors and reduction factors are crude assumptions. If you do not have access to the latest Engineering Electromagnetic software tools, such as CDEGS-HIFREQ, you may be forced to use Reduction and Split factors and their associated assumptions to make practical best guesses at representing what's happening in real-life. However, if you do have access to the latest top-end software, unshackled by the limitations of legacy tools; then Reduction and Split factors can be consigned to a thing of the past!
Models built in today's electromagnetic world are far more complex and reliably replicate both the in-ground elements (e.g. buried earth electrode arrangements). As well as the above-ground elements such as the overhead line equipment (OLE, OHLs, Transformers, pipelines, bridge structures, etc). ALL the electrical interactions are accounted for 'in the model', i.e. in the virtual space.
The EM versions of models still need specialist users. To ensure all the input properties and electrical configuration correlates to real-life outputs. Still, the upside is that one more technical assumption can be eliminated, which is good engineering in anyone's book. All the paths are accounted for, no second guessing, no ambiguity, no inherent built-in error. It's all as close to real-life as is practically possible today.
It's also true. No matter how amazing an advancement in practice can be. Legacy thinking still takes a while to let go of old habits that are not always so relevant today.
Challenge the convention, and get the Safe Earth System you deserve ... call or chat with us today.
This post is written by Ian Griffiths, Principal Engineer at GreyMatters, an Earthing & Lightning Consultant of 27 years, one of the top 1% accredited CDEGS consultants and professional advisor to international utility companies, data centre and infrastructure developers.