Modeling & Analysis Techniques for Solar PV Grounding Systems

- Welcome, everyone. My name is Jim Chastain. This is the EasyPower Thursday technical webinar series, and our topic today is Modeling & Analysis for Solar PV Grounding Systems. And as we are in the habit of doing, before I introduce our speaker, we'd like to run a couple poll questions by you, get a little bit of feedback on the topic as we get started. Very much appreciate your participation. So, the first question is, have you been involved in a large utility scale PV project previously? It's strictly for the edification of the speaker so he gets some context for the audience's interest in the topic. It looks like we have a quorum, thank you. Here's how folks have weighed in on the topic. It's like we have a pretty good spread of candidates, thank you. Second question is, do you perform your own grounding studies? We do record the presentation and it will be posted on the website, and you will receive a notice via email when that video is up. Should be early next week. All right, thank you, looks like we have a quorum. This how folks have weighed in on this one. Outstanding. And then finally, how familiar are you with the IEEE Std 80 guide for safety in AC substation Grounding? Little trickier inquiry. And this is probably the most important question as far as the feedback because it'll help David understand how much in-depth to go on the basics. All right, thank you for participating, here's how folks have responded to that question. Excellent. At this time I'd like to hand the baton over to David Lewis, the EasyPower grounding and power systems engineer and a certified professional engineer. - Thanks, Jim for the introduction. Thank you so much for calling in to this webinar, and definitely appreciate the time and any questions that you send over. Again, thanks for your participation in those poll questions, 'cause I think we'll spend a little bit of time making sure everyone's familiar with the concepts of IEEE 80. So, today in this webinar we're going to be talking about the modeling and analysis techniques for solar photovoltaic systems. And really the purpose of this presentation is to highlight different methodologies that engineers are using for evaluating these large photovoltaic sites. And we're primarily going to focus on the personnel safety criteria that's derived from the IEEE Std 80 criteria. So, my goals is that people that attending this webinar will be able to describe utility scale photovoltaic systems, their configurations, and be able to identify different challenges that are encountered when evaluating these very large systems. Then also being able to identify when that data is going to make a significant change to your analysis and results. So, we'll kind of take a pros and cons approach for different analysis techniques as we step into later portions of this webinar. In order to meet those goals we're going to talk about the purpose of grounding, make sure everyone kind of has a understanding of what we're trying to achieve, since I think about 40% of the attendees are newer to the subject of the grounding analysis. So, I want to make sure everyone has a base level of understanding of what we're trying to achieve here before going on to photovoltaic power plant design options and how those design options play into the grounding analysis challenges. We're also going to take a couple minutes to switch over to the software so that we can show in XGSLab how these different criteria or different components come together in the software. And then we'll show a summary of different modeling techniques that engineers are using. So, I want to start by giving an understanding of why we're doing these grounding analyses and what's the purpose of putting grounding systems into the earth. And first and foremost, it's really about personnel and public safety, and there are other aspects to a grounding system that will help your power system operate during normal and abnormal conditions and helps prevent equipment damage during lightning events and faults. But really we're going to focus on personnel and public safety. And when we talk about that, I'm focusing on the touch and step voltage hazards that could occur. And a couple terms to be familiar with is, whilst refer to as the GPR or the ground potential rise, this is the voltage of your grounding system during a fault event. So, most people will assume that a grounding system has zero potential, but when we're doing these analyses, we're looking at this system when it has a significant voltage or significant voltage compared to a remote earth or someplace far away. Then another aspect to be familiar with is the earth surface potential. So we, when we're evaluating these systems, we're trying to reduce the voltage gradients or reduce the voltage difference between an individual's hand and their feet, or one foot and another. And whether or not the GPR and the earth surface potentials are adequate or sufficiently low, it's we're going to be pointing towards the IEEE Std 80 to help establish a criteria for what's a survival threshold of voltage. So, certainly recommend looking into the IEEE Std 80. It is due for an update I believe in 2023, so there should be new content coming soon, but really that's what's going to drive the compliance criteria for this grounding system analysis. And when I was talking about the touch and step voltages, I want to make sure everyone's clear on exactly what this looks like. So, a touch voltage hazard is essentially if you have a fault on your power system, your equipment, all of your grounded bonded equipment can be elevated to a ground potential rides of a 1,000 volts. And if somebody's in contact with this piece of equipment, their hand is going to be at a 1,000 volts, but the soil voltage, there's going to be some voltage drop associated with that soil and therefore there can be a difference in potential. So here their hand's in contact with a 1,000 volts, their feet are in contact with 800 volts. So, this is what I would refer to as a 200 volt touch voltage. You can have a similar type of concept occurring with a step voltage, but this is a scenario where somebody isn't in contact with any equipment necessarily, they're simply walking across the earth's surface and there's a voltage gradient in the earth's surface such as 900 volts for one foot and 800 volts on another. And that results in what I'll refer to as a 100 volt step voltage. So, that's kind of the basics of what to consider for a grounding analysis. And I do want to mention that there is a standard that is the IEEE 2778, which is very related to this grounding analysis. It's really the IEEE standard for photovoltaic systems. What's provided in that is more of a how to or different techniques methodologies for performing a study for a grounding system. And then IEEE Std 80 is the criteria that holds the voltage thresholds for touch and step voltages. So, they are going to be a highly related standards, and I do believe IEEE 2778 is also receiving comments, so plan to be updated probably relatively soon. Don't have a specific date or year of that yet though. Continuing on, we'll kind of go into a general overview of a photovoltaic system design so that everyone can understand just what these systems look like and what we're talking about when we're actually modeling a grounding system for a photovoltaic in our software. So, photovoltaic systems, they're generally just interconnected panels and this allows us to convert DC power into an AC power. So, we become a utility scale photovoltaic power plant once we cross that five megawatt threshold. And at this time five megawatts is a relatively small system. There are several 20, 40, 100 megawatts, some as large as 300 to 500 megawatts being installed in the U.S. And these systems are electrically relatively similar. You're going to have multiple feeders or collector strings that are going to take the power that's generated from these panels. They're going to go through an inverter, probably a step-up transformer, or maybe this step-up transformer will convert that to their interconnecting distribution utility line. Or maybe you have a wye, delta-wye transformer with a collector substation associated with this photovoltaic power plant that might step up that energy that's generated at this power plant to higher voltage levels like 69, 138, 345, et cetera. But these systems can take several different physical forms and this is what's going to drive our grounding system analysis in some ways. So we can have a system that we'll refer to as a congregated system where all of our components are within one field. So, all of these panels are in one region and they interconnect with, we'll say a collector station. Or we could have a dispersed site where either due to the land ownership issues, the physical geometry or topology of the environment such as a more hilled environment, you're going to have some separation from one area to another. And this is going to derive some of how we are designing our system in that it could determine whether or not we might have a collector substation where we kind of have this image on the right where it's a station or substation associated with the photovoltaic power plant that is allowing us to interconnect with the utility. Or we could have something a little simpler where it's just a step-up transformer that feeds into a switchyard or a substation that's away from our photovoltaic system. And internally to our PV power plant, we could have overheads, overhead circuits with neutrals, and that would be more consistent with those dispersed systems. Or we could have interconnections through our feeders using our cable with concentric neutrals. How we're actually holding our panels up is also going to be dependent on your system. You can have a stationary panel system where the solar panels are set at a specific angle and they're not changing that angle. You could have a single access tracker that's kind of illustrated here where as the sun crosses over the panels, the panels, their orientation is changed to maximize the amount of energy that we're able to convert from that solar radiation. So, that is a gain on an efficiency on these solar panels. So we can even do a dual access tracker where we can change not just the angle in one direction, but in multiple angles so that we can maximize the energy from the sun even more. And all of these are going to have a little bit different implication as far as how we are bonding and grounding these various panels. So, often photovoltaic panels are going to have some sort of bonding to your system, a grounded and bonded portion to your system. You could either have a continuous conductor that might be run with your panels. You could have a panel track that is a self-grounding and bonding panel track, and you need to be paying attention to how these connections are made when you're designing your system and evaluating the grounding system performance. And just to go on to that, you can kind of see on the illustration here there's, or the image, there's a photo of an actual metallic bond that's made between the supports of the panels to a support of the panel. And then down here we can have a point where our AC grounding system is interconnected with our DC grounding system. So, there might be a DC equipment ground that's run to all of our panels, but these often are bringing in to the same grounding system in the end. So, this is going to give us a continuous path for any voltages that might transfer from our AC system to our DC system. So, something to be aware of. In addition to that, your trackers, how they're supported is going to have some physical parameters. So, this is going to dictate how well voltage may conduct on these different components. So, you can have a tracker that's more of a circular pipe that can rotate. You may have a stationary rack that's built up more like this photo on the right, and I've heard, I've not personally worked on, but I've heard there are some manufacturers that will actually specify insulated and isolated supports so that there's not a bonded connection from the panels to each of these posts. So, that would be one area that's going to drive how we model our grounding system, as we'll see when we step into modeling later in this webinar. Other pertinent physical details to be aware of is what the posts are. It could be an I-beam, it could be a drilled pier, and in some sites they may be installed in areas that have ultimately high resistivity. Or if it's installed in relatively low resistivity, then there's concerns about corrosion occurring. And if there are those concerns, they may use a post that is coded so that it is protected from that corrosion. So, that is going to be a parameter that's going to drastically change how we might approach our grounding system analysis here. So, all that information that we've just kind of talked about is really just laying the background of the different photovoltaic system designs and some thoughts that you'll want to think about when we're actually doing a system analysis. So, for a grounding system, typically a grounding system's going to consist of some sort of horizontal conductors as shown on this photo on the left, with some ground rods. And these can, photovoltaic systems be placed around the inverters and maybe at the step-up transformer, and possibly you'll have a larger facility where it has its own collector substation, maybe trenched ground that's run out with your cables or your overt lines, and that makes the connections to each of the inverters. And of course, all of these are likely to have some sort of perimeter fence that's going to play into another aspect of personnel safety, because if you have a large ground potential rise, there's an opportunity for that voltage to be transferred to this fence either directly through metallic connections or through the soil itself. So, one of the first issues that we run into with performing a grounding system analysis is characterizing the soil. Because these sites can be very large you can have variation of the soil characteristics as you go from one corner to another. And that variation is typically going to occur along the upper layers. Often the deeper soil layers are going to have a more uniform resistivity because of the environments often, that stratifications often occurred from the same source material and they're undergoing the same climatic environment. So they're going to have as relatively uniform soil resistivity at deeper layers. But those top layers are going to see human interaction, more recent climate events that are going to change how that soil resistivity looks. So, multiple measurements is a typical process in characterizing that upper soil resistivity values and seeing how it might change across the site. For those that do grounding system analysis, it's as you might know from IEEE Std 81, you also need to understand the soil characteristics at deeper layers, because even though your grounding system may not directly connect to those deeper layers, that has a direct effect on your system voltage, your ground potential rise that occurs, as well as your system impedance. So a challenge with these very large system is getting a sufficiently long traverse to be able to characterize those deeper layers that drive your system impedance. Another thing that's a bit of a challenge with photovoltaic systems is that you're unlikely to apply any sort of surfacing material to these sites, in contrast to substation designs where if you have a substation like we had shown at the beginning, you are likely to have some sort of crushed rock gravel surfacing and that acts as an insulating layer for those touch to step voltage and helps reduce the current that could go through somebody. That's likely to not be installed in these photovoltaic systems because it's just very costly and it needs ongoing maintenance to maintain that. So, all of those are going to be issues that will pertain to photovoltaic systems in particular. Now, I'll switch over to the software in a minute, but we've talked about kind of the physical characteristics like the piers or the tracks, how your photovoltaic panels are supported and how you're making these connections between your step-up transformers to your main collector substation. And that's going to play into how we're going to energize our models and what we're going to use for the conductors that represent these systems. Additionally, there's tools that don't have advanced soil modeling capabilities. So, being able to capture this stratification may be a challenge in some tools that really are not suitable for these large photovoltaic systems. That might be fine for a small ground rod or small transmission tower for instance, but not really suitable for larger substations photovoltaic systems. And of course, there's going to be different computational capabilities. So, I'll go ahead and switch over to the software real quick and kind of show how this analysis process might look for doing a grounding system performance. So, here is the XGSLab software when you first start it. Here's our start page, and you can enter in some project information, but I'm just going to go and set my reference standards to IEEE 80 2013 and set my frequency for my power system. And right now I'm not going to enter in a sub model, we'll just use a uniform sub model and I want to create a grounding system for my photovoltaics. And I'm just going to point to my CAD model, and just for this purpose of this webinar I am importing a table of the layers. So, we already have a CAD layout for our system and what XGSLab does is you can point to various layers and assign different properties to those layers as you import it into my system. So here we have to our conductor that's being imported and putting that at a depth of two feet. Then we have our tracker that we're importing, and that's a negative two feet. So in a grounding software the negative's going to be up in the air. So, I'm importing this and then lifting those conductors so that they're represented above the earth's surface. And you can kind of see all of these characteristics. So, if I go to, let's see the PV track, I can go over to my conductors here, scroll down, and then if I show my conductor, this is something that a user could be entering into the software to say this is a steel conductor that's four inches wide, it has a thickness of 0.2 inches, so that way I can model accurately how the voltage is going to drop across that conductor. And when I import this from a CAD file, I have my system layout represented here. So, we can see if I zoom in this would be something like the ground loop around my inverter. Here we have represented the connections from a dedicated ground to my first post, and then this is my tracker and then each of the posts that go into the earth that support that tracker in the air. So, I'll just inject a fault so that we can look at the voltage that occurred across the system. So, this will take about 10 seconds, but while that's running, since I've just imported this into the software, what it's doing is it's doing a check of all of my conductors being interconnected, making sure there's not drafting errors when you import, such as conductors that are separated and kind of islanded. If there are, that's something that the software can evaluate. You just need to either turn off that check or set the electrode parameters so that the software knows you intentionally don't want these two things directly connected. And then while that's running, there's also a thread count here that I can augment based off of my processor to allow me to do with this study faster or slower. With my analysis what I want to show is its voltage potential. So, one challenge I think that many engineers run into when they're doing a grounding system analysis for photovoltaic is that if they're using a simple tool, that tool is likely to use an eco potential plane assumption. And what that means is essentially all of the conductors in my model are assumed to have no voltage drop, they're superconductors, and the system's really calculating the impedance based off of the area By my grounding system, not so much how the voltage is going to propagate through this system. So, with a tool like XGSA_FD I can see where I've injected my fault current, we have about 4,400 volts as my ground potential rise. But if I look over at this corner over here, I can see there's about a 400 volt difference between one corner of my system to another. And that's going to be an issue if I don't consider that in my calculations. So, really what we want to make sure is that tools are accounting for the self-impedance and mutual impedance so that they consider the voltage drop when you're looking at large photovoltaics. And even this I would consider not a large photovoltaic, this is what I would say a medium size or small, most photovoltaics now are going to have many times this dimension. So, you'll see this issue more prevalently on even larger systems. And what that plays into is our compliance on our system. So, right now I'm just going to plot my touch and step voltages, but first I need to set my compliance criteria. Right now what I'm doing is setting the compliance criteria based off of IEEE Std 80. So my permissible touch voltage in the event of a fault, somebody in contact with any of our ground and bonding equipment, according to IEEE 80 they should be able to survive about 268 volts. And then the step voltage, my permissible limit's about 377 volts. So I'm applying IEEE Std 80. And if I go back to my analysis here, I can look at my touch and step voltage plots. And there's another thing that's very important, or very important detail to capture when you're looking at photovoltaic systems is what your reference is. So if I'm plotting my touch voltage plot for instance, what I've done here is I'm plotting the reference voltage or my touch voltage based off of something. So, a touch voltage again is the difference between my equipment and the earth's surface potential. So, I can tell software like XGSLab what my reference is, so I can say my reference is this point right here, this 131 that you could see where I've injected my fault. Now, I can change that to be my elements instead. And these elements I might set to a reach of let's say 15 feet. And if I re-plot this, what we'll see is a very different touch voltage plot, and that's because this software consider the touch voltage based off of a search radius. So, now we have a difference in what this plot looks like, because I'm calculating my touch voltage based off of my conductors. So if I'm over here on this corner, I want to know the voltage difference between my feet standing at this point to this conductor, as opposed to my previous plot, which was showing the voltage based off of the earth's surface here, the voltage at my feet here compared to my reference point. So by default most software is going to use this type of reference and that's another detail to be attentive to. So, if your software, if you're using XGSLab, I want to make sure that you are aware of the elements reference so that you're choosing the correct touch and step voltage reference. And this is going to play out as far as a safe areas plot goes too. So if we plot our safe areas within all elements, we can see the yellow areas are representing where we're exceeding those touch voltage criteria. Green areas is where we're within that touch voltage compliance criteria. So, we do see some areas of green kind of in the central point where the voltage difference between our metal and the earth surface is relatively close. Same thing around where we've injected that fault, but if I change to my reference point again, doing the safety plot, we no longer see these safe areas show up in the central area. So we get a little more accurate view by looking at it from the all elements perspective where we can see in reality there are actual, there's within IEEE 80 compliant voltages in these green areas that are not shown if I'm using a simple reference. But a thing to know if you're using a tool that doesn't account for that voltage drop is that you can actually artificially see positive results. Meaning you could see green areas and areas that are are really non-compliant, or you could see artificially negative results, meaning it could look like it's failing, but in reality when you account for that voltage drop, your system looks compliant. So, that's just some of the things to be aware of when you're using a software tool that you need to understand a little bit of what's going on underneath the hood, and as well as what options there are. So, it's with XGSLab we want to, again point out the reference point versus all elements. So, going back to the presentation here, we have what we already talked about, this eco potential assumption that a photovoltaic system really is not appropriate to analyze with equal potential assumptions, because it's just not going to consider this voltage drop. Another thing to be aware of is your reference. So, we already talked about this, how that reference is going to make a big difference in your plots. And I didn't mention this, but I set my reach distance for my calculation here to 15 feet when I was calculating with all elements, meaning that I'm looking within 15 feet of each of my bonded pieces of equipment. I did that because I'm considering that an individual that's standing may be able to reach their arm out three feet, but that panel could extend out from the tracker 12 feet, for instance. And in that situation their total reach distance would be about that 15 foot reach distance. So, this is another area that you can kind of control what your maximum voltages are that you're looking. So, instead I can set it if it were just a three foot reach, we could recalculate this and have a more controlled analysis to look just within three feet of any of my conductors, so that way we can get a maximum touch voltage based off of this search radius. So, going back to the presentation, so there's the how that software tool actually analyzes your system whether it's concerning that voltage drop or not. But there's also another aspect to consider with photovoltaics is how your system fault is occurring. So this might be a little challenging to follow for people that are just getting familiar with the concepts of IEEE 80, but basically it's when you're doing a substation grounding system analysis, a power system fault occurs, essentially your power system's operating normally and then a fault occurs from one of your phases to a structure and that current going through the earth back to its source. This is the current that's going to produce our ground potential rise. And that's going to create those touches type voltage hazards. There's going to be additional paths for current to flow, such as the transmission shield wires, or the distribution neutrals, or your cable neutrals. But the current that's going through the earth is what's producing that ground potential rise. And the concept is that this current going through the earth is essentially going to a source that's remote distance away. It's sufficiently far that these systems are not directly interacting. And for a low site fault, let's say at our smaller substitution instance, if we have a local fault on the lower site, that current would go down to our grid and take a metallic path back to the source or back to that neutral winding. So, we're not going to see the ground potential rise occur for that local fault. If this fault were say happened down here, then again we could have this ground potential rise. But with the general concept is that current taking all metallic paths is not going to produce a ground potential rise. The issue is that with a large solar site, it's really not this straightforward, because the physical dimensions that these systems can cover is multiple acres and therefore, we'll say a "local fault" can be miles away from where your source is located, and therefore you can get some ground potential rise again. And to illustrate how this works is we have the nets module and just creating a simple system. I've stepped through kind of a comparison of the amount of current that will return on these metallic paths, versus the amount of current that will go through the earth. And what I'm stepping through is the difference between an overhead cable, kind of like typical distribution tower configuration, versus a cable with a concentric neutral. And what we'll see is that there's going to be a mutual coupling effect that helps reduce the current going through the earth when we have this cable. So, on the top of this chart, the top row we have a overhead line that's a one mile length and what we'll find is there's a total fall current of about 9,000 amps in the event of a fault. During that event about 89% of the currents going to go through the earth through our grounding system. And part of this is due to these solar photovoltaic systems have a very large area that they'll often encompass. So, they often have a very low impedance. So, you'll see quite a bit of current actually going through the earth at these systems. But if I change the configuration, let's say we have about 4,000 miles of overhead line and that transfers to an underground cable. When we get closer to our panels, well then we start to see a reduction in the amount of current that goes through the earth producing our ground potential rise. And as we kind of step into this a little bit farther down the the chart, what we'll see is if we have a system that's just fully cable, just a one mile cable, we will see an increase in the amount of fault current that's available based off of the fault current availability and the impedance. But we do see a rather drastic reduction in the amount of current that will go through the earth and produce our touch to step voltages. So, your power system configuration is going to change the amount of current that's going to go through the earth that produces your ground potential rise, that as well as produces your touch to step voltages. And what I want to do is kind of step through some different examples. So, here are three different examples of different methods for modeling that fault current. So, in a typical analysis like what we just did in the software, we just placed a fault current, 10,000 amps being injected into our system. And what we can see with a simple injection is the green areas representing where we've met touch voltages, yellow areas where we're exceeding our touch voltage criteria, and then some red areas where we're exceeding that touch voltage criteria. What we can do with this software is we can actually inject current at one point of our model and then do an inverse or dual energization back at wherever our source is located. And what we see is a little bit, well, we'll say a very minor difference in the voltage values using the same green, yellow red scale, looks very similar. And then if we were to have this same disperse system, but in this instance we're actually modeling the overhead line and the neutral so that we can have the current carried out to the faulted point and then account for that mutual coupling and the conductive coupling that could occur along that distribution neutral or that collector's circuit neutral. And we'll see a little bit of improvement here for this particular system. So, with this kind of dispersed system going through these different energization methods, we don't see a significant improvement with the entire system model where we're accounting for the overhead lines versus just our simple injection. But there is some improvement when we account for those current that's going to take the alternative paths back. If we have a different kind of configuration though, let's say this congregated system that's going to have a higher likelihood of a cabling system throughout with concentric neutral around that, we'll now compare this system using the different energization methods. So again, we can look at a simple injection method where we're just putting in 10,000 amps here. We have the green areas located kind of in the central area where we're elevating the soil voltage such that it's within compliance of our, within the compliance criteria. Then we'll have a dual energization approach where we're again injecting 10,000 amps and returning that back in our source. Or we can consider a system where we're actually modeling the cable with that concentric neutral and we can see because of that mutual coupling that's occurring, our green areas drastically expand. So, in this particular environment we're going to see a much greater improvement by taking these extra steps and modeling this system more fully and modeling that power system configuration. And I do want to take a moment to point out for our users. So in the screen conductor's library, this is where you're going to be able to set your cable parameters. So, here we have cable with the face conductor in the center, then we have the concentric neutral around that, and this is going to be the screen conductor that you can model in the software that you can inject the current down the base conductor and have that mutually coupled to that concentric neutral to be able to model this. So, there's a whole different slew of ways that we can approach modeling the different components of our system. We have obviously our dedicated ground conductors, which we showed earlier where we have a trench ground or just a loop around or converters. There's other components that I've been modeling, and that's what I'll call the auxiliary components. Those are going to be the posts that support our panels, the tracker arrays, all of those additional metallic paths that current may go down that can act as an auxiliary grounding system. It's not really the dedicated grounding conductor, but they will conduct electricity through the earth. Another thing to consider is that we've just talked about is your grounding conductors auxiliary systems with your power system modeled into the environment. Now, because of the amount of how large these systems are, you can actually exceed most computers' memory when you're trying to do a very detailed model of this entire environment. So there's different approaches that engineers can take in order to try to account for all of these different components to help them develop a more accurate analysis. So, what we're going to step into the next portion of this is looking at a detailed regional models or reducing some of those auxiliary components, maybe using lumped impedances to consider these other paths a current can take. And I will make sure to get through this last portion of the webinar so that we can leave some time for questions. Certainly trying to get through a lot of content relatively quickly. And here we have what I'll call my comparison of a system where we've just modeled the ground conductor. So, the dedicated ground conductor is represented kind of leading out from our step-up drop and substation. And then we can see around that conductor is an area of red, essentially indicating that we're exceeding both touch and step voltage criteria. What we can do is modeling the posts that support the panels and the track connection between them and all those grounded conductors. We can essentially reduce the difference between the earth's surface potential compared to the grounded and bonded equipment that we're looking at. And this is going to be a more accurate view of the system performance, because now we're considering how current's going to flow through all of our different components of our system that are bonded to our dedicated grounding system. And what this is doing is it's one step helps reduce our system impedance, so we're reducing our overall ground potential rise, but it's also by all of these connections elevating the soil's voltage at this point. And that's why you kind of see in the central area where we start to see more of these touch voltages being compliant as compared to the corners. Now, the challenge I mentioned is with very large systems it may be very challenging to get a computer that has sufficient memory in order to store all of these different components for analysis. So, here on the left is we'll call it grounded and auxiliary model. We've modeled about half the posts of this entire environment, and in this method we've been able to model pretty much the entire space of this system, and we're just using the simple dualization to keep things apples to apples. But what we can see is a lot of this space we are exceeding the touch voltage criteria. If we change how we're modeling this, we can switch from modeling every other row to half of the posts are modeled, but we're modeling each of the rows. But again, we see relatively similar results. A lot of this area is still exceeding our touch voltage compliance. When we look at this final model, we have eliminated these other regions that we were previously modeling, but we're doing a higher detail on our region that we're looking at. And what we can see is that we're starting to see more touch voltages that are becoming within compliance. And what I'll press is that this is for this particular system is going to be the more accurate representation of the real-world performance, because we're accounting for all of the components within this region that we're looking and we're not seeing the effects or the what will likely be beneficial effects of these extra portions. And little tricks in order to consider these other portions. So just looking at this model to this model, we still have high detail in the region that we're looking in this top right corner, but we might model some of the auxiliary components that we are affected by in the closer area, or maybe we might use something that's a lumped impedance. And you can see each one of these little impedance symbols represents a forced impedance to remote earth that is representing the impedance of each one of these systems. And this comes up relatively close in comparison to our actual model where we have all of these components, it's a little bit better performance. Part of that can be artificial, because we are making this connection and indicating that that is to remote earth and that's just a parallel impedance to our system and it doesn't produce the ground potential rise that you're going to see from me is that might affect this system performance. So that's something to consider in that modeling. A fun thing that you see though for these very large systems. If I look at a system that's I lumped impedance just around my regions where I'm focused versus lumped impedances throughout my system. I didn't make a huge improvement on my system performance comparing these two. And that's an indication that for this particular model on this particular system, your what's happening out in this environment isn't going to play a large role in a fault that occurs in one component of your system. So, understanding where this region, where this cutoff is, is going to depend on your specific systems geometry and solar resistivity, as well as your electrical properties, how you're energizing this system. But that's something that you can look at through sensitivity testing, either with the scheduler tool. Over in the files there's a scheduler that you can run different models, or just exploring how these system operates in this different environment. So, it's really going to be a site specific. Another thing to consider is the perimeter fence. Now, there's different recommendations from different engineering teams on what the best methodology is to help reduce touch voltages that can occur along this perimeter fence. So, when you're modeling your system in high detail, what you'll see is that it's going to change the voltage that's conducted through the soil to that fence. So, when you're looking at your model, what you want to do is try to account for that voltage gradient and how that's going to transfer out to your perimeter fence. So, on the top model we're not doing that, we just have the dedicated ground, and because of that what we see is there's touch voltages exceeding compliance limits in some portions, kind of along these corners where the yellow's indicated. But when we do a more detailed model where we have all of those auxiliary posts, we see a little bit different story. And in fact, we see non-compliance in this quarter when previously that looked like it was compliant. So, you need to account for the effects that these auxiliary components are going to, what they're going to do as far as how that's going to conduct voltage in the soil and around that perimeter fence. Obviously there's different methodologies for handling that perimeter fence in IEEE 2778, I believe is talking about how you can actually transfer voltages if you run a direct metallic bond from your grounding system to your perimeter fence. And in that way you could actually elevate the hazard potentials that are occurring along that perimeter fence because you're elevating the voltage of that perimeter to all of your equipment's potential. That may or may not be a a good idea, depending on again your system geometry and your soil environment. If you are running a direct bond to your outer grounding system, it can help in reducing your grounding system impedance. But again, photovoltaic systems generally are very large and will have a lower system impedance overall, but it's something to consider. So, kind of in conclusion of this webinar, there's a lot of physical and electrical design considerations that are going to dramatically affect your photovoltaic system and the corresponding grounding system performance. And engineers really need to have a good understanding of the tools that they're using, especially considering if they are evaluating self-impedance and mutual impedance rather than ignoring that voltage drop and how it's making that touch voltage reference to make sure that what they're analyzing is providing accurate results. That's shown through this kind of sensitivity testing. We had relatively good results by doing a more detailed regional model, and that is a pretty ideal methodology in order to compensate for lacking in memory for your computer. With XGSLab software you should be able to model pretty much anything. It's mostly going to be your hardware that will limit how much of your system you can model. And again, really recommend sensitivity testing in order to understand how your specific system performs, because the case studies that we walk through in this webinar are showing a specific soil resistivity, specific fall current, a specific geometry, and those are all going to play into how your system performs. So, we've left a few minutes for questions here and I'll go ahead and open up the question bar and give people a chance to enter those in. So, the first question I saw is, "How does the IEEE 2778 guide for solar power plant grounding differ from IEEE 80, or do they need to be used together?" Really they need to be used together is how I would respond to that. The IEEE 80 is what's going to drive the compliance criteria. There are practices that would be typical for a substation but would be atypical for a photovoltaic system that are really where 2778's going to help you. Really the concepts of IEEE 80 are applicable for all grounding system analysis. It's really IEEE 80 gives some practical guidance that is more tuned towards a substation application. But conceptually the concepts in there are broadly applicable. How you apply those concepts is more on a case by case, site by site basis. Another question that came in is, "When characterizing soil for a photovoltaic system, do we take those soil resistivity measurements and make an average model to represent them across the entire site?" Well, my response would be, it depends. And I think the major point of this webinar is really to indicate that when you're doing these photovoltaic systems, there's a lot of, it depends when you're making engineering decisions. So understanding these concepts is going to help drive that. But some systems there's going to be dramatic soil resistivity changes, and in those you might break your system out into different regional analyses with different soil models to account for that. And with XGSLab there are different soil modeling capabilities like multi-zone soil modeling that can allow you to account for voltage or that solar resistivity across the earth surface. Multilayer is going to be more common methodology that you can account for that stratification. My typical process is doing multilayer with a regional analysis. So yes. Another question that came up was that we are, I'll just read it. It says, "I heard that XGSLab software factors in the voltage drop. Are we talking about the voltage drop along grounding conductors of the grounding grid, or along the PV cables?" Really it is whatever you model in the system. So the XGSLab software is really more of a Maxwell's equations solver. So it's a methodology so that you can evaluate how fields are going to propagate in various mediums from metallic systems, some sort of insulated prop, insulated materials, the soil. So, whatever you model in your system, you should be able to account for in XGSLab. And really I want to say that GSA_FD and XSG_FD are the tools that are accounting for self-impedance and mutual impedance. If you're using smaller systems, GSA is a tool that is going to be a little bit faster and, but it omits that self-impedance and mutual impedance. Again, this session is going to be recorded, so if you wanted to revisit this, please feel free to do that. We've covered a lot of ground and there's a lot of concepts, especially for the audience mix that might need to review. So please feel free to send me an email or reach out to EasyPower in general. We'd be happy to help If you have any follow-up questions. I do want to try and get a couple more questions before I'll cut this off. There's multiple questions that I'll make sure to respond to. So, if I don't respond to you verbally here, I'll make sure to send you an email with your, with the answers to your questions. One of the questions that came up was in the fault current split portion here is just re-elaborating on how the ground potential rise is going to occur, versus the, I guess its relationship to its source. So, what we have illustrated on the left is by power system operating normally, and then we'll say a fault happens. And so the current is going to try to return to its source, wherever it may be. In reality our power system's going to have multiple sources, but back at wherever your source is, you're going to see a ground potential rise at those locations. If you have a fault locally, that current has a very low impedance path via the metallic structure back to its source. So the delta is allowing you to feed this fall on the wye side. So this current is going to be a single underground fall for instance, that is going through the grounding conductors and then back up to the neutral. And because it's not actually going through the earth, you don't develop a voltage in that soil. So you don't develop those touch to step voltages. Another question, I think this might be the last one and then I'll go ahead and end the webinar since we're already a few questions, a few minutes over is, "How are you obtaining accurate solar resistivity samples, and is there a recommended practice for densities and depths?" So the IEEE Std 81 is the guide for performing solar resistivity measurements that's associated with doing grounding system analyses. And the depth that your measurements it, the typical practice is that you want to have, let's say an A spacing that is equivalent to the dimension of your system. For photovoltaic systems that's often not very feasible because you don't really have the ability to take a three mile long traverse if your system has a dimension, maximum dimension of one mile. So typical practices are that you might take multiple measurements at each one of your areas of focus. So for a site like this that's dispersed, we'll have shorter solar resistivity measurements, maybe a 100 foot A spacing at each one of these points. And then we might a 1,000 foot traverse, one or two of those in order to characterize deeper soil layers. Whatever we can physically perform based off of the site's geometry and what site permissions allow, as well as your test equipment's capabilities. So you typically need a stronger, a more powerful or more sensitive solar resistivity meter in order to really accurately measure these. So again, thank you everybody for attending this webinar, covered a lot of grounds, so please do feel free to email questions as they come. You can email me at david.lewis@easypower.com, and there's also support@easypower.com, and support@xgslab.com if you're one of our current users. Thank you everybody for jumping into this webinar.