5 Avoidable Mistakes in Electrical Earthing Design

This is a case study about when GreyMatters was asked to analyse a Rail Track Feeder Station in Europe.  This Feeder Station had already undergone a preliminary Electrical Earthing Design Study. The principle constructor was concerned that the previous analysis had uncovered some “strange numbers.” These were raising more questions than it had answered.

Following a review of the previous electrical earthing design study, the potential sources of error were identified as:

  • An inaccurate soil model (>10% rms error)
  • An incorrect fault level (100% error)
  • The contribution of the rebar mat had been ignored
  • The conductive Steel fence had been ignored
  • Touch voltage hazards had been identified, yet were left unmitigated
  • CDEGS MALT version had been used on a rail system which contain long parallelisms
  • The impact of the Rail Bonds had been ignored

Exploring the Problems

Mistake 1 - Soil Model

It’s often stated that Earthing is not a precise science. It relies on the interface between two components, 1) the man-made (electrical system) with 2) the natural element, e.g. the geology of the site.

The man-made can be well defined. Its behaviour can be easily predicted and calculated.  Whereas, nature has a way of throwing curved balls (unpredictability) which can make ‘making-sense’ of this interface a lot more challenging.

Some of the best minds in the business have tried to simplify this interface into elegant equation(s), some of which feature in the legacy Earthing standards, such as BS 7453, IEEE Std80, BS7430 but even these equations rely on assumptions to simplify the mathematics into an advanced, yet manageable suite of equations.

The key point is the Soil Model is the foundational cornerstone that ALL subsequent parts of the Earthing System Design are built upon and are 100% reliant on for their validity.  This means, elimination of error at this stage is of absolute capital importance.  In this case studied, the stated error % reported was greater than 10% - this is unacceptably high error as a starting point.

Mistake 2 - Fault Level

Fault levels should be accurately calculated by the supplying authority or asset owner.  These days, this is usually done by circuit theory based computation, but manual calculation is also practical.

For the case studied, a fault level which was twice the magnitude of the real situation was used and reported.  This was probably due to the incorrect data being provided or assumed.  The net effect was to immediately overstate any resulting EPR by 100%, which would have led to significantly more buried copper being installed than was necessary.

Mistake 3 - Impedance Contributions

Rail, rebar, and bonded equipment can all provide valuable contributions to the overall impedance of an Earthing electrode.  As well as controlling the surface voltages and providing possible parallel Earth-return paths back to source of the fault.

A challenge with manual calculation of multiple returns is that without the use of high-end FEA tools, such as CDEGS, the complexity when considering multiple layer soil structures can defeat even the most accomplished of mathematicians.  The various standards do not help much either and only provide simplistic calculations or indicative split or reduction factors for guidance, but applying these without considering the full context, is not a good idea.

Mistake 4 - Spotting the risk but not doing anything about it!

Since the mid 2000’s, there’s been a paradigm shift away from manually calculating final Earth resistances of earth mats to understanding and ‘controlling’ how a human being interacts with a system in fault.  The standards now set limits on step and touch voltages during fault conditions.

Touch voltages are fairly simple to explain - they are simply a voltage caused by a potential difference between one thing you can touch and another.

Mistakes Electrical earthing Design

Step voltages are simply the potential difference between your feet. By looking at the human’s physical ‘points-of-contact’ or interaction with an energise system and controlling this interface to stay BELOW permissible limits.

Permissible limits are defined by the IEC in a set of plot curves that map out the survivability of a typical population. In the UK, when considering earthing system design we use the IEC C1 curve, for which 95% of people will survive an electric shock below the curve - this implies 5% will not survive, so it’s important to understand the risk is probabilistic in nature.
permissible-touch-voltages-in-electrical-earthing-design

In the rail environment, a different standard applies in the EU, which uses different calculations to set the touch voltage limits.  Unfortunately, in this case the previous earthing design study identified the magnitude of these touch/step voltages but offered no advice or measures to lower them to within the permissible limits – so, effectively the design (as presented) remained in a state that could kill or injure anyone touching or in the immediate proximity of the site.

Mistake 5 - Software Selection

In this case, the previous report was carried out by others using CDEGS ‘MALT’ version.

Unsurprisingly, the creators of CDEGS, SES Technologies, recommend that the correct version/module is matched to the site specific context. The less capable versions, MALT and MALZ rely on a number of simplifying assumptions the can lead to a significant underestimation of EPR, sometimes greater than 50%.  Below are a couple of the key points from the CDEGS versions to be aware of:

CDEGS-in-electrical-earthing-design

  • MALT (CDEGS) assumes that the entire earth mat is at the same voltage, which is very dangerous to assume, particularly on larger sites with low soil resistivity.  
  • MALZ (CDEGS) does allow for voltage drops along conductors, but does not account for inductive or capacitive effects between conductors.
  • HIFREQ (CDEGS) - directly solves Maxwell’s equations to give you the full picture of the performance of your earthing system, including both above and below ground elements. We cover this in more detail on our blog, here [link].

Unfortunately, the graphics engine is the same for all versions so the output plots and reports look the same, but the results may not be!  This is especially true for the larger system and/or where there are significant distance(s) of parallel elements… for example, a rail network and associated systems.

What follows are the steps taken to correct...

Getting the foundation right

First, GreyMatters attended site to perform a series of soil resistivity (Wenner) soundings using high accuracy equipment capable of injecting 10 times the normal signal current to that of other accepted instrumentation.  By using a far more potent signal it is possible to dominate other potential sources of in-ground noise, to ensure accurate interpretation of the returning signal is assured.  In addition, a higher signal energy is capable of reliably sounding to deeper depths whatever the geology to ensure the subsequent Earthing calculations are using the optimum data-set.  This is particularly important for the electrically larger electrode system.

soil-resistivity-in-electrical-earthing-design

You can find out more about this in our blog series on soil resistivity testing.

Using high signal injection equipment, GreyMatters was able to develop a more accurate soil model by halving the error from ~12% down to less than 5% using CDEGS RESAP.  Having the key foundational piece of data down to less than 5% meant the subsequent design and calculations are all based on a solid, reliable, accurate data which is key when human lives/safety are at stake.

Representing the Real World

Finite Element Analysis (FEA) simulation techniques rely on building meaningful representation(s) of the real physical world in a virtual environment.  This can involve above ground elements as well as the buried. Indeed, for this case an accurate geometric model was built by GreyMatters research engineers including several kilometres of track, the OLE (overhead line equipment) structures, cabling, switchgear, transformers and the earth mat so that the impact from the track bonds, the fence and rebar could be examined against their positive contribution to the EPR of the site.

Accounting for the impedance contributions and other parallel paths not only better represents ‘real life’ but avoids over-engineering the electrode system by utilising the full mix of accepted elements.

An often ignored part of any earthing system design is the civils, e.g. more specifically, the foundations. It’s not often recognised that concrete is porous, so when wet has a low resistivity.  As a result, rebar can be used to great effect to reduce earth resistance without significant additional cost. This by itself had the most significant impact on the touch voltage hazards. Additional surface voltage control, and fence bonds were also introduced.

When modelling systems with above ground components it is critical to use the HIFREQ module, as MALT and MALZ ignore all above ground structures, while HIFREQ  calculates the interactions between ALL the elements (Maxwell’s equations).

GreyMatters’ final design reduced the site touch voltages from more than 3500 V to below the permissible safety threshold of 645 V across the entire site, as well as reducing the EPR by a factor of 10, from more than 9000 V to less than 900 V without introducing significant additional copper.

Summary

In this case study we highlight how 5 of the 7 “deadly sins” of Electrical Earthing Design materialised during a real life scenario which without correction could have led to an unsafe design being delivered.

Key Points

  • Rubbish in = Rubbish out. Robust, accurate, correct Data must be used at the outset to develop accurate model
  • High-end soil resistivity testing equipment capable of dominating any in-ground noise. Understanding the soil model is critical to an accurate earth design, and inaccuracy here is dangerous.
  • Ignoring the contribution parts of the above ground network can play when modelling the impact of rail bonds. It’s important not to assume something is going to improve the situation – proof is required that any additional bonding actually makes the site safer.
  • Safe environment designed without introducing additional costs. By modelling existing structures, it was possible to use the existing equipment to reduce the hazards on site without requiring additional construction which more than paid for the cost of the study work.

Further reading: The 10 Commandments of Electrical Earthing Design

If you believe your current Electrical Earthing System Design needs a rethink from some new GreyMatter then get in touch and we can discuss your challenge.

"Working with GreyMatters has totally changed my expectations of a professional engineering consultancy firm … for the better! Technically, they were on the money! More importantly however ... they kept me updated throughout the entire project, which was so refreshing, and therefore quickly put my mind at ease that things were being taken care of." (C. Turgis)
  • subsea 7
  • Atkins
  • Siemens
  • Hydro66
  • Western Power Distribution
  • National Grid
  • BAM Nuttall
  • Rotary Building Services
  • M&W Group
  • Valero
  • Technip
  • Royal Airforce
  • PM Group
  • MOD
  • ISG
  • Ineos
  • Exelis
  • Essar
  • Deerns
  • Network Rail
  • AWE