Newsletter

×

Subscribe to our mailing list

+1(503) 655-5059
Home > Videos > Introduction to Grounding Calculations
An Introduction to Grounding Calculations and Why They Are Necessary

This webinar, given by Michael Antonishen, P.E. at TriAxis, a Division of DEA, provides a basic introduction to grounding safety calculations (IEEE 80 step & touch) that are generally performed for medium and high voltage AC power stations or similar facilities. Basic grounding concepts and calculation inputs will be introduced, and the need for this type of modeling and analysis will be discussed. Using software programs to model grounding systems and perform analysis is now standard for all but the simplest grounding designs, and commercially available programs have many uses. We will explain some of the most common uses of fully featured grounding programs and show example results for a simple grid design.

See the full transcript of the webinar below.

 
V714
Download free demo

 Watch our most popular videos

EasyPower OnSite Intro

Utility Short Circuit & Arc Flash

Arc Flash Hazard Reduction

Arc Flash Hazard Reduction

Changes in NFPA 70E 2015


 See all of our training videos

Full Transcript of the Video

Mike Antonishen: Hi, I'm Mike Antonishen. I'm an electrical engineer at TriAxis, a division of David Evans & Associates. Today we're going to be talking about system grounding calculations. The type we'll be discussing today are most commonly associated IEEE 80, which is the IEEE guide for safety and AC substation grounding. Power and electrical system grounding is a massive topic and can be very confusing. So today we're just going to go over the basics of what the calculations are, where and when to use them, with some examples from the real world sprinkled in. So, like I said, we'll cover grounding and grounding calculation basics. It's kind of the definitions, the why and the where. We'll get into an intro to calculations, inputs in modeling, then we'll cover some brief examples of the IEEE step and touch calculations and some software examples. After that we'll do a brief overview of other software uses. So, before we get started though, before we dive into things, some key definitions that we'll be (00:01:00) touching on these words many times.

So it makes sense to define now. Earth current or grid current, or ground current as it's sometimes called. For a single line to ground fault, ground grid understudy. Earth current is defined as the current flowing through our local grid and back to some remote source through earth. This current leaks to earth out of the local grounding facilities and leads to local ground potential rise. It is this ground potential rise that causes our safety issues. So ground potential rise is defined as the maximum electric potential that a grid and the surrounding soil may attain relative to remote earth. Under normal conditions, with no fault in your grid, it's close to zero. But during ground fault conditions it can be elevated. It's a pretty simple calculation to get maximum potential rise. It's defined as the earth current (00:02:00) times the resistance of the grid. Ground potential rise can be different at different points on the grid but the maximum calculation is quite simple.

Another key term is step voltage and touch voltage. Step voltage is defined as the difference in surface potential of voltage by a person bridging a distance of 1 meter with their feet or 3.28 with their feet. It's essentially the difference in voltage between your two feet as you're standing on ground and that difference in voltage can cause a current flow which could be dangerous. Touch voltage is the same thing but with what you're touching. It's the potential difference between where your feet are on the ground and what you're touching on some sort metallic structure or something like that. It can be evaluated at multiple distances depending on the situation you're in or how conservative you want to be. For a visual representation of ground potential rise, we have a large grid shown here. In this image, there's a topographical overlay of (00:03:00) GPR shown floating about the buried grounding conductors beneath. This is a large, complex, connected grid. The fault that is causing the GPR is likely located in the pink area on the right, the area with the largest GPR.

You'll notice that other connected portions farther away also have elevated GPR but to a lesser extent. This is due to the impedance of the grid conductors and the different paths the fault current takes on the grid and off the grid to get back to some remote source far away. The difference in GPR between any 2 points is the basis for step and touch potentials. We're likely to see some hazardous step and touch potentials on this grid where there are large changes in GPR within a short distance. So now jumping into why, why we do grounding analysis, why we do grounding calculations, why we ground. Generally speaking, we ground systems to inform equipment bonding to achieve predictable fault current return paths. (00:04:00) A predictable fault current return path allows for fast and effective system coordination and fault isolation. This protects both equipment and personal. Equipment bonding to ground or to ground grid prevents dangerous voltage from being present between 2 pieces of equipment or between equipment and earth. And a ground grid or ground mat is designed to create an equipotential plane.

During a fault or other conditions causing GPR, the grid is meant to bring all of the ground and all of the equipment around you to the same potential so that the voltage between any 2 step or touch points is meant to be safe during faulted conditions. Low resistance ground grids, more copper in the ground or a ground grid that's well designed can help keep local ground potential rises lower which can help with step and touch safety. But it also helps provide low resistance paths for voltage surges such as lightning and transience. In summary, all of this rounding (00:05:00) design and connection of metallic objects contributes to the design of a system that is safe for personal and helps minimize the damage and interruption during system faults or other abnormal operating conditions. So, where do we ground? In this presentation, grounding is done in all, grounding is thought of and designed into all electrical designs. However, in this presentation we'll be kind of discussing it in context of substations, generation facilities, or locations in MV, HV distribution distributions. Such as maybe a solar farm, or a wind utility scale, wind. The basic concept is anywhere that fault current returns to a remote source via earth. And so the easiest way to make that make any sense is to give an example.

And so, if you have a single line to ground fault, we're worried about single line to ground faults generally because single line to ground faults are where we see ground current flowing. (00:06:00) Where current is flowing between conductors to earth and back to some source. If you have a single line to ground fault on the low side of a substation, the low side being with a, delta to low side grounded transformer. That fault is going to exist mostly on the, the phase conductors and then to the ground and then the grid conductors and the return, if it's within that same substation it's going to be local to your grid. So, most of the fault current is going to return directly on your ground grid conductors and back to that Y ground of the transformer. In this case, this is not a remote fault. Because everything is local and flowing generally on conductors you're not going to see much ground potential rise from that. So, this is the sort of situation where we're not generally worried about this. There's not current flowing out of our grid back to some (00:07:00) remote source and so we don't see a lot of ground potential rise. We just have circulating current and maybe some localized effects depending on the magnitude of the circulating current. The situation we do worry about is when you something like a single line to ground fault on the high side of that same substation. The Y can be delta Y or Y grounded on the high side. But the main point is that the fault source for a fault on the high side of this substation would be something that's fed by maybe another generation source farther away, or a substation far away, and what you end up with is fault current that flows into your local grid and then needs to get back to it's source which is somewhere outside of our yard.

So, it's going to leak out of our ground grid conductors to earth, then back through the earth back to it's source. It can take paths through the earth or on transmission shield wires or distribution neutrals, but (00:08:00) generally a portion of it's going to leak out and that portion that leaks out causes ground potential rise. So that's where we worried about, that's we're worried about it, that's where we do the calculations. So, what are the grounding calculations that we perform? IEEE 80, which is the basis for most of these calculations provides guidelines and formulas to determine the allowable step and touch values. This is generally done for 50 kilogram, conservative, 110 lbs human. These allowable values that are determined using IEEE 80 are based on human body resistance, top layer soil resistivity, fault current magnitude, fault duration, X over R ratio, and they're calculated for each grid design based on these changing factors cause they are almost never the same. The IEEE 80 standard outline's calculation methodologies that (00:09:00) help determine the size of grid conductors required for the amount and duration of fault current, helps determine the grid area, the number of ground rods, the conductor spacing that could be used for a least cost design to prevent least dangerous step and touch voltages. And finally, you can use the calculated resistance values to compare against actual grid resistance measured in the field for verification of design, or for additional items like ground potential rise calc for specification of telecom HV isolation equipment.

So, software tools that are used to aid in the calculation of this design. The most basic that is available and has been available for a long time if the IEEE 80 spreadsheet. This is a spreadsheet that can take a user through all of the calculation steps necessary to design a grid. Spreadsheet is useful for a full design but only in limited situations. And this is because it requires uniform soil, uniform conductor spacing generally good for small and equipotential grounding (00:10:00) systems. The equipotential assumption if you will is a, is one where you assume the grid is small and because of that the resistance between any 2 points is so low that it can be neglected. So, the IEEE spreadsheet is a good tool but it's only generally really good and accurate for the smallest simplest designs in an actual uniform soil. On the other hand, there are software tools available that can handle not only uniform soil but multi-layer, or multi-zone models. Most of them come with a calculation module to help determine soil model from raw resistivity data with minimal error which is helpful when you're doing multilayer soil models. In addition to that these software tools can model large and non-equipotential grounding systems. So, for larger ground systems or conductor materials that aren't copper you have to start worrying about the resistance of the (00:11:00) grid and these can incorporate that into their calculations.

Software tools can also generally achieve a more economic design which which is kind of shorthand for less copper and less excavation required. They can also achieve more complex designs and an example of complex designs include uneven spacing you, you can perform calculations the local mitigation of local effects as opposed to mitigating a local effect by beefing up the whole grid you can just, you know, beef up one area and rerun the simulation saving yourself on conductor everywhere else. You can do large grids like we’ve talked about, you can do things like multiple grids with circulating current between them and then you can also do some pretty cool calculations with transferred potentials. Transferred potentials being those where you maybe you have a fence running adjacent to your site and you want to know whether or not you want to ground that in a certain way or connect it to your site and so (00:12:00) you can model that and, and, and you'll see the potential's developed around the fence and on the fence as a result of that and determine if that's also safe for step and touch during your fault condition. So, the main reasons for software use as opposed two IEEE spreadsheet would be if a 2 layer or a multilayer soil model is necessary. Some studies that have been run on this and it's been indicated that the uniform soil model is actually quite rare, it only occurs in a very small percentage of locations.

So, it's not always a good assumption and you might end up with a much more conservative design if you're forced to use that ex form soil model. In addition to that software tools have the capability to model uneven grid conductor or rod spacing. This often ends up in a more economical design because you don't need the density of conductors everywhere, you need it in certain areas. It changes (00:13:00) the, requirement for this will change depending on soil characteristic so it's not always the same. With software, you can also a model localize effects like we've talked about. You can solve them with local solutions, software is much more capable of doing large grids with at which, which, which includes mutual and self-impedances so no equipotential assumption. And then finally you can, you can use the software tools to model the presence of nearby metallic structures and their effect and the grid effect on them such as fences and buried pipes. So, like we've said before large grid, circulating current situations where multiple grids must be analyzed and they interact with each other.

While we're on the topic of software tools I'll note that there are several simplified software products on the market for grounding analysis. But, many of them only offer simple analysis with a built in equipotential or the super conducting assumption as it's sometimes called (00:14:00) These simplified programs neglect the self-impedance of conductors, they neglect the actual impedance of the conductors and while they are a step up from the IEEE 80 spreadsheet they are still only able to be applied appropriately in limited situations. So, relatively small copper with a single energization point. For fully featured grounding analysis software tools SCS CDEGS may be the most well known in the North American market, my company, TriAxis division of DEA is a CDEGS license holder and I do use CDEGS. But it should also be mentioned that EasyPower has recently partnered with a company called SINT, it's an Italian company that serves worldwide markets, which makes and sells the XGSLab software product. XGSLab is also a fully featured software product that competes with CDEGS. Their main product offerings and modules are quite similar. We'll talk more about XGSLab and other software capabilities later on. All of the screenshots in this presentation are from XGSLab and we'll (00:15:00) be touching on it as we go through. So, before we jump into some examples to look at and some visuals, we should talk through and make sure we understand all of the calculation inputs, at least briefly.

The 3 main inputs are the soil resistivity, the model that comes from the soil's resistivity, measurements of preliminary grid design, and the fault current magnitude and duration. The most important, arguably the most important input would be the soil model, and this is because it's kind of the basis for the whole design. You can have exactly the same grid design, and exactly the same fault current but your soil model is going to determine whether not that grid design is safe in, with that fault current and the soil model can vary so much that that is really important to have a specific soil model for each site. So generally, the soil model is gathered with 1 or 4 point testing, (00:16:00) which I've shown in the equations to the right with the, with the diagram, or the Schlumberger method. Once you've measured the resistance you calculate the resistivity of each layer. There could easily be an entire webinar on this so I won't dive in deep into soil resistivity measurements and soil models right now. But these methods are outlined in IEEE 81 which is the guide for measuring earth resistivity, ground impedance, and earth surface potentials of a grounding system.

So, the other important input obviously, is the grid design. You generally start with a preliminary grid design. You can make it with any sort of CAD or free CAD program. You can import DXF files generally, and that's a good way to start and then you iterate to get to your safe design once you've got all of your inputs in. Most software tools also come with a conductor sizing tool, IEEE (00:17:00) and IEC both have standard conductor sizing equations which can be applied, and the outputs of that conductor sizing equation is generally the cross-sectional area of conductors required for your design. 4/0 is generally the most common grid conductor size. 2/0 is also used and some larger ones are also used, generally you're not gonna need larger than 4/0 unless the fault current exceeds 50,000 to 60,000 amps, but magnitude, duration, and conductor material all factor into this decision so, it's generally calculated each time depending on what you're using. it's a pretty quick calculation. Finally, fault current magnitude and distribution are very important. Generally, you use the highest single line to ground fault, the most zero sequence current that's flowing that returns to a remote source. So, generally speaking it is on the high side of the substation. You use either the primary of the backup clearing time.

There's a robust discussion on primary VS. (00:18:00) backup clearing time that we also won't get into today, it's touched briefly on in IEEE 80. And in addition to the fault current magnitude and duration you can apply things called split factors. And these split factors are things that account for the fact that it even when you have all of your fault current returning to a remote source, not all of it is necessarily going through the ground. Some of it might be going through your transmission shield wires or distribution neutrals. If it's leaving the substation on a, or generation facility or facility on a conductor it's not leaking the ground and it's not causing ground potential rise. So, you account for that by reducing the amount that you call earth current in your design. There are several factors that affect what flows through the ground versus what might flow back through a transmission shield wire or a distribution neutral. Some of the factors are distance to remote sources, fault contributions from each source, soil layers in between your source and your fault, tower resistance, and the actual neutral (00:19:00) path resistance. So, that's again, a pretty complex topic. You'll calculate that too and it helps reduce the amount of current flowing actual to earth. It helps you model the most realistically the situation.

So, jumping right into examples now, we can see an example substation layout, grid layout. So, this is just showing the bellow grade conductors for a substation. We're not modeling anything above grade at the moment. The blue is the grid's. Consider that maybe 4/0 copper or something like that. And the green are the foundations and you can see that the foundations in this case are bonded to the grid in certain locations. Then the foundations, the green in the foundations represents the actual rebar inside the foundation. Foundations aren't generally used as a main grounding conductor but they, in a robust grid like this can be connected to the grid. And so, this is sort of an example of what you might upload from a DXF file into a grounding program. (00:20:00) The next input, the next thing that you would do in a, in a grounding analysis program is you would determine your soil model that you apply to the preliminary grid that you imported. You know, what's your grid sitting in? What we're showing here is a multi-layer soil model and it's based on soil resistivity measurements. The black crosses represent the max measured values of soil resistance for each spacing and the blue is a curve fit on top of those. The soil model, finally, in red is one that has been calculated based on the measured values.

The programs iterates until it finds the model with the least error. There is an RMS error value on the far right-hand side that shows 3.261% which is a pretty low value. It indicates how well the model that is on the screen matches the data that we've measured. In this case it looks like we have a 4-layer model with the parameters called out on (00:21:00) the right. This visual shows the ground potential rise with the proposed grid design. This is during the design fault that we've applied, in the soil model that has applied, with the preliminary grid design that is shown below. You can see the grid and some ground rods that are going down in the Z direction from the grid. You can see, this is a nice visual because you see that the grid area, the elevated area in pink on the GPR chart, it's brought together up onto an equipotential plane. So, even though the voltage is about 8000 volts on this equipotential plane, there are minimal potential differences on the plane. So, if somebody's standing on this, touching something, or if somebody's taking a step on here it's apparent that it, they're most likely safe. This isn't the diagram to determine safety but it shows what the grid does, what (00:22:00) we're build, and what happens during a fault. It's a pretty cool visual. And you can also see that as the farther you get away from the grid there's no current leaking into the ground where there's no conductors and you can see that the ground potential as you go, the scale of potential as you go away from the grid kind of falls off like a placemat on a table or something like that or a table cloth on a table.

So, based on that ground potential rise that we just saw on the last slide, the program can do calculations for each 1 meter step in there and each 1 meter, 2 meter, 3 meter touch within there and it can do safety calculations, safety calculations for step and touch. The step and the touch allowable voltage are calculated per the IEEE 80 equations. Starting on the top right you can see that, these are kind of small but, top right is an example of the safety calculations for both step and touch in (00:24:00) that grid without any surface covering applied. So, standing on bare, native soil. And you can see that the USSP on the top right is the step potential. USTP is the touch potential. The green areas represent where both are below the IEEE 80 calculated danger condition and the yellow indicates where your step is safe but your touch is not safe. And so if you, if you look on the top left, this is an example of where we've applied a certain amount of gravel surface covering. And you can see that the application of gravel surface covering has taken this grid design without adding anymore copper, from an unsafe condition around the edges to a safe condition through the whole grid design. The bottom 2 are a little bit fancier looking graphs.

They also show essentially (00:24:00) the same thing, that you have a safe step, a safe step and touch everywhere with that gravel. Another example to show something slightly more complicated, is for a PV/Wind plant. Up in the top left corner you can see where the fault is applied at, probably the project substation, and then throughout the model you can see the, you know, the pink, and the blue, and the green conductors represent different grounding conductors connecting to different things throughout the site. And this is quite a large site. So, in this case we're likely to see ground potential rise differences throughout the site due to the impedance of the conductor's. You can do some pretty neat things in these grounding software packages. This is an illustration of the leakage current distribution for a fault at the substation. And so, you can see that the red is minimal leakage current out to some remote earth. Then, as you get closer to (00:25:00) the substation which is now, on the right side of the page where the fault has been place you can see that there's a lot more current leaking out of those conductors. So, say that the remote source is somewhere far away.

All of the current is trying to follow the lowest resistance path back and so most of it is flowing directly out, but some of it is taking a path along the other conductors and then leaking out over there. Likewise, with the current leakage distribution you've got, this one is showing me potential distribution, and it shows what potential each conductor has been raised to because of the fault. And so, again, we see the highest potential in the substation on the far-right side there. And as you get away from the substation, due to the impedance of the conductors, the potential rise on each conductor goes down where on the left side of the site your conductor potential is somewhere (00:26:00) between 94 and 155 volts and on the far-right side we're some are up in the range of 880 to 945. And so what that looks like on our elevation chart now is, this is essentially showing the same thing as the last one but just in an easier visual way to look at, is that you can see that the substation grid has been raised up to a very level and everything that is still connected has been raised up but not to the same extent. And then finally, the end result for this one, and I believe this is on native soil as well, we show that with that fault, with this grid design as modeled we're all green which means that all of the step and touch hazards have been handled.

If you get to this point and you've still got some yellow areas where you want to mitigate touch potentials you might want to consider adding ground loops or (00:27:00) increasing the area of the grid, trying to get the grid to a lower resistance. There are a lot of different solutions you can apply, but this represents kind of the end iteration of a design. you might not get to this point as quickly. It might require some engineering opportunities. So, software capabilities. We'll step you through software capabilities in terms of XGSLab offerings, but we'll also talk about others. First, we'll talk about the 2 base level modules where the base level module is called XGSLab GSA. And this GSA module is similar to the CDEGS Malt module as well as some offerings from ETAP and SKM. There are a couple other simplified grounding modules out there. All of these modules have what's called the equipotential assumption that we've talked before where there's no conductor impedance which means the ground potential rise of all of the conductors on the grid is exactly the same and they all (00:28:00) leak the same amount because there is no impedance between them. This sort of assumption tends to underestimate the GPR in the case of low resistivity soil or large grids so that's what causes it to have limited capability.

There are benefits to it too though. In grids where you don't need the extra computational power and you're sure you can use the equipotential assumption you get slightly easier data entry and fewer computer resources are required to calculate this. The GSA module, or this level of module is also useful. It's definitely a step up over the IEEE spread sheet because you get the numerical and graphical output useful for investigation of GPR, leakage current, surface potential, surface potential, and touch and step voltage distribution. Stepping into the next level, what we've kind of termed as starting to be a fully featured software is what XGSLab terms their GSA_FD (00:29:00) product. The FD stands for frequency domain. This is, as I said, the beginning of the fully featured sort of thing. It's similar to the CDEGS Malz or multi-ground-Z module. And the main change between this and the last one is that the equipotential assumption is now gone. We are still only modeling below grade facilities but now we include, now we include the self-impedance of the actual impedance of the conductors. And that also includes, maybe conductor codings or something like that, and the conductor size.

Another key thing is that XGSLab adds capacitive coupling and mutual impedance in addition to self-impedance which has the effect of providing more accurate calculations. These products can also be used more accurately in high frequency applications where as the GSA or the module below is typically not as accurate at a (00:30:00) higher frequency. Studies have shown generally that the mutual coupling, the mutual impedance is important because studies have shown that neglecting it can introduce about an error of about 20% into some models. GSA_FD does have this but it's important to know that some of the competitor products don't take mutual coupling into account until stepping up into another level in the software module complexity and cost. This is a slide kind of showing the importance of including the self and mutual impedances. And showing what you get when you make that equipotential assumption. So, on the top left you can see this is the same grid modeled 3 times and you can see with the equipotential condition the ground potential rise for this grid is exactly the same everywhere in the grid where there's a conductor. And then, if you step to the far right, you can see that where we are modeling only self (00:31:00) impedances, the left side of the grid which is likely where the fault is has a higher, and then as you get farther away from the fault point the impedance of the conductors has the effect of reducing the ground potential rise farther away.

And then finally the bottom shows where you've included both self and mutual impedances. This does have an effect and in this case, you've actually underestimated the GPR when you haven't considered self and then self and mutual impedances. So, they are important to, to include. So, more kind of full features of the, of that second level model is that, again, we can do frequencies that aren't just power frequency. Generally, the range is up to 10 MHz with some accuracy. You can use it for DC calculations, for cathodic protection, you can use it for calculation of magnetic fields due to grounding systems or cable, magnetic fields (00:32:00) caused by your grounding system's cable. Finally, you can use it for electromagnetic interference calculations maybe caused by your facilities or other. Generally, these other facilities we're talking with end up being buried pipe lines or other buried electrodes nearby. But the fact that you can, that you can calculate the induced current and potential due to the resistive and capacitive coupling is helpful in many situations.

So, stepping up to the next type of module, XGSLab has their XGSA_FD module, which adds an X to the front of the last one. This starts to be comparable to the CDEGS HIFREQ module and it extends, it essentially extends everything from the GSA_FD or the MATZ application to overhead systems. So, now, (00:33:00) in addition to modeling the influence of your underground systems on things like pipelines or whatever, you can model the interaction of overhead and/or underground power lines on installations such as pipelines, railways, communication lines, you can calculate the effects of lightning and also the fault current distribution due to a lightning strike. So you gain a lot more capabilities with the above ground and you start getting one step closer to modeling, or actually you are at the point where you can model reality and really do some, some, some almost academic level studies and answer very specific questions on your grounding system when you have to worry about above ground facilities. And then finally, kind of the next step in the top of the line with, which you only need to use sometimes is the XGSA_TD and this is just taking everything that's developed and every single other module into the time domain. And (00:34:00) the useful application of this is to look at the response in the time domain of your conductor network when it's energized with a current or a voltage transient, usually this is most associated with lightning, and so you can look at the lightning strike dissipation into the soil, the electric and magnetic field generation, determine damage due to equipment to in surge arrester, determine what might be dangerous to people.

It's very helpful, and so the XGSA_TD module uses outputs and information from the other modules in the frequency domain to help calculate those things. And then finally, a quick package comparison, we've mentioned other software products, and what compares with the different modules of XGSA and CDEGS, and the basic is that there are a few full packages or fully featured softwares, XGS and CDEGS are 2 of them and (00:35:00) you can see their comparison here. So, thank you very much.