The Top 3 by E3

Series Episode Four: PVsyst Parameters

• E3 Consulting • Season 3 • Episode 7

Share episode

Send us a text

In this final episode in the PVsyst series, Daniel Tarico and Frances Plourde discuss the impact of system design on photovoltaic (PV) system modeling, focusing particularly on shading losses and module mismatch within the context of using the PVsyst software. 

Frances also covers light-induced degradation (LID), which happens when PV modules are first exposed to sunlight, particularly affecting crystalline silicon modules with certain dopants. LID is a one-time effect but can reduce module efficiency by 1-3%. This factor must be accounted for when designing and modeling PV systems, especially for long-term performance projections.

The conversation highlights the importance of considering the long-term effects of module degradation, mismatch losses, and shading when designing and financing PV systems. 


Daniel Tarico:

Hello, welcome to the Top 3 by E3, a monthly podcast about the intersection between engineering, energy, and project finance. She's a project manager here at E3. Welcome.

Francis Plourde:

Frances.

Francis Plourde:

Thanks, Dan, it's great to be here. As with the rest of our series, we're focusing our discussion on modeling performed using PVsyst.

Francis Plourde:

PVsyst allows for a lot of different parameters of system customizations, each of which can have massive impacts on model system production, which then can have great impacts on financial projections as well. The industry standard PV system modeling software, we believe it's really important for site owners and developers to know how these models are produced in order to understand their accuracy and applicability. Several of these customizations relate to loss assumptions in the model, which reflect how much generated power the model expects to be lost due to various system criteria.

Francis Plourde:

The modeling has become so important, which is why we've decided to put together this year's podcast To put this in perspective. There really isn't much we can do about the weather at a specific location. We discussed that earlier. What we do is measure the weather over many years and develop a weather data file that has been used to estimate the solar energy that will be coming into the PV plant. However, we do have some control over the energy production. When we design the plant, we select equipment features and we determine the layout. Both of those impact the production. All that information is input into the PV system model. Now, inevitably, there are going to be imperfections and inefficiencies that cause energy production losses, even when a system is at its very best. So, Frances, could you tell us how the system design can impact the generation of losses for what we are modeling?

Francis Plourde:

Certainly, losses can be introduced from almost every component of a PV system, so anything that can impact how photovoltaic energy is produced and transmitted can introduce losses or reductions of efficiency. Some of these losses are well known and can be incorporated into system design quite simply, such as inverter conversion efficiency, but others require more knowledge of equipment behavior and performance in less than ideal or real world deployment environments. In less-than-ideal or real-world deployment environments.

Francis Plourde:

So, the topics that we're going to discuss today are examples of causes of PV system loss which depend on a multitude of variables and that should be carefully considered when putting together a PVSys model.

Francis Plourde:

Okay, sounds good. I think we're ready to get going. The first type of loss we'd like to discuss, or I would like to discuss, is the shading loss. Now, pv modules This can happen even in cases where modules are partially right, so that shaded right shading can be caused by external objects, such as nearby buildings or trees, even tall grass. sunlight to produce power because they convert sunlight into electrical energy. I think most people know that and a module shading reduces system production and generation, for obvious reasons. Shading can also be caused by components of the PV system, specifically in the form of what we call inter-row shading. Now, if the rows of PV models are installed too closely together, rows can create shade on other rows, depending on the time of day and the angle of PV models are installed too closely together, rows can create shade on other rows, depending on the time of day and the angle of the sun. And how well the tracker is functioning. Can you explain how PVsyst accounts for shading losses?

Francis Plourde:

Certainly so.

Francis Plourde:

First off, it's important to clarify that PVSys specifies two different types of shading far shading and near shading.

Francis Plourde:

Far shading relates to objects that are caused by the horizon and the position of the sun, while near shading is caused by nearby objects.

Francis Plourde:

So far shading is modeled in PVsyst by defining the horizon line, so where the sun is in relation to the PV system, and the horizon line incorporates objects at a distance of approximately 10 times the size of the PV system area, so the area of the PV system multiplied by 10, that's the distance on every side of the system that the horizon line calculates shading from.

Francis Plourde:

So, for example, if you had very tall mountains nearby that cast shade on the system at different times of day, if you had very tall buildings that were kind of in the distance, if you think of the horizon skyline of a city, that is the shading that is contemplated by the far shading horizon line parameters in PVSys.

Francis Plourde:

So for near shading, this is near shading is what's caused by objects much closer than that, so objects within the nearby vicinity of the PV system or in the system itself. Pvsys allows users to calculate near shading losses through the use of what's known as a shading scene, and so this is where users of PVSyst can go in and design the area surrounding the system, putting in 3D objects which account for nearby trees, buildings, like you said other vegetation, other objects, including aspects and equipment residing in the PV system itself, and the shade scene incorporates things like elevation angle, the angle at which the modules are installed, and then the direction which the PV modules are facing, and this is where the user can also specify the system's row arrangement and inter-row spacing in order to correctly model inter-row shading concerns.

Francis Plourde:

Okay, so to summarize, really the shading depends on two of the system design features. And secondly, how close are you to nearby trees and hills, maybe boundaries of the open field where you're building the array? These will block out part of the sun and the power reduction from that loss of sunlight, I guess, is kind of obvious. But I know that module shading can lead to mismatch losses as well, which is something that I don't think is so well understood by many people, and that mismatch loss amplifies the shading losses. Could you speak about that a bit?

Francis Plourde:

Certainly, you're definitely correct on that front. So mismatch losses can occur when there are differences in voltage between modules or between strings of modules, and this is often caused by shading not shaded, or if in a string of modules you have some modules that are shaded and some modules that aren't. This creates what we call voltage mismatch, because the voltage in the unshaded area is significantly higher than the voltage in the shading area and the output voltage of the individual module or the string of modules is limited to the lowest performing module or area of the modules. So this can overall reduce system production or lead to other electrical issues in the system due to that mismatched voltage. And mismatch losses, as we've said, can be caused by uneven shading, by different factors, but this can also include uneven soiling. So if you have areas of a module that have more soiling or, like dirt or dust, build up on certain portions of the module than others, or more build up on one module in a string compared to the others, that can also induce voltage mismatch losses.

Francis Plourde:

We've talked about vegetation growth. If certain areas in a string are more impacted by overgrown foliage or grass, that can cause voltage mismatch. And then, especially if you have differences in performance, so if you have modules that are underperforming compared to other modules in that string, that can also introduce voltage mismatch losses. So PVsyst models allow for incorporation of several different types of this mismatch losses. So PVsyst models allow for the incorporation of several different types of this mismatch loss.

Francis Plourde:

As we've discussed before, mismatch losses due to shading are calculated by creating a shade scene of the system and incorporating far and near shading considerations, and then mismatch due to differences in module performance can be specified as a constant by user input. So the standard PBCIST assumption is one to two percent loss, depending on Then, the user can specify how to incorporate module degradation into mismatch calculations known module performance, but again it changes based on the modules that you're using. because mismatch due to module performance tends to increase as those modules degrade and tend to perform worse over time. So PVsyst allows you to define what it calls string mismatch, which then can be used to quantify losses due to differences between strings.

Francis Plourde:

Okay, Well, it sounds like if I was to interpret that, that module quality variance or something like that is something that will also contribute to mismatch loss beyond what you get shade. F rom the shade assume that means a system designer, by selecting higher quality modules, will be effectively designing a system as lower mismatch loss, either at the very beginning or throughout the life of the project, if they degrade more slowly. That to me seems like an important consideration. Now, are there other module-related loss factors which PVsyst can model?

Francis Plourde:

There certainly are, and one of the most interesting ones to discuss is what we call light-induced degradation, or LID. And LID is a distinct concept from what we consider standard module degradation, which we assume modules are going to degrade or lose their production capacity by about half a percent per year from their initial nameplate efficiency. This is a process that occurs over the life of the module. Lid is something that occurs instantaneously when the PV modules are first exposed to light. So the first time that they're installed in their deployment environment, when they're very first exposed to sunlight, is when LID is going to occur.

Francis Plourde:

LID is a very specific phenomenon that only occurs in crystalline silicon PV modules. Very specific phenomenon that only occurs in crystalline silicon PV modules. But on top of that, it only occurs in crystalline silicon PV modules that are produced with what we call P-type dopants. So boron, for example, is a P-type dopant that's used to create the semiconductor material that's used in certain PV modules. Anytime anybody is using P-type doped silicon PV modules, people should be aware that those types of modules are going to be susceptible to LID and need to design their PV-Sys models accordingly.

Francis Plourde:

Okay, Beyond that, you know this is one of the more important degradation factors and another point where the equipment selection will have a long-term impact on the model performance. Could you go into a little more detail about how that works?

Francis Plourde:

And this is where we get into some very interesting semiconductor physics related to PV module.

Francis Plourde:

So the losses that are caused by LID are due to excess oxygen atoms that get caught in the silicon PV wafers during manufacturing, and so these additional oxygen molecules can combine with the boron that's used to dope the silicon wafers, and that doping process is what creates the semiconductor material that allows for PV modules to operate. Semiconductor material allows for PV modules to operate. But when these oxygen molecules combine with the boron molecules, this creates energy levels within the structure of the semiconductor material, and these energy levels will essentially trap some of the electricity or the electrons and holes that are produced by the PV module. So this reduces the amount of power that the module is actually outputting, so it reduces module efficiency. This process occurs incredibly quickly, again when the module is initially exposed to light, so it can usually be modeled as an immediate decrease in production capacity rather than a compounding source of degradation, like we consider other module degradation to be. So the amount of efficiency loss due to LID is usually assumed to be one to three percent, depending on the specific module that you're using.

Francis Plourde:

Okay, so well. I'll trust you on the solid-state physics, but I do believe that, again coming back to the idea of the system designer selecting good quality modules, it's really important to consider how these discrepancies in quality can lead to large impacts in system performance over a long period of time, and this is important, especially with finances targeting 30 years of operations. Could we widen our scope a bit and think about that? How can PV systems be used to evaluate PV system performance over the long term?

Francis Plourde:

Yeah, you're absolutely correct that it's incredibly important to focus on the long-term performance of these systems and how you can apply PV system models to that.

Francis Plourde:

So it's essential to be able to model expected PV system performance in the long term and not just at the time of installation. As an example related to our discussion within this podcast, mismatch losses can increase over time as the modules degrade, as we've discussed before, since modules installed at the same PV system may degrade at different rates. So maybe you have different types of PV modules that are installed in the same system, or you have different batches of modules that are produced at different times that are installed in the same system. Where you have different batches of modules that are produced at different times that are installed, so they might not all degrade at the same rate. Pvsys allows users to determine expected system loss over a defined period of time due to the combined impacts of standard module degradation and mismatch degradation. We will be talking more about how the results of PVSys models are interpreted to use in production and financing estimations in another podcast in this series, but it's definitely important to be able to interpret PVSys models to know how these sites are going to perform long term.

Francis Plourde:

Okay, so we've talked about the shading, something which depends on the siting layout. You know, we've talked about the shading, something which depends on the siting layout, and we've talked about module mismatch, which depends on the shading, but also the module quality, and maybe differences in degradation rate within the solar field. Seems like both of these can be modeled by PVSys. Now, what about the inverters? These are another item on the major equipment list. It's controlled by the designer, you know, selected by the developer, the designer.

Francis Plourde:

Absolutely so. Just a quick recap. So inverters are the site component that converts the DC energy that's produced by the modules into AC energy, that is, input into the grid, put into the grid, and these are essential for PV system operation because otherwise you wouldn't be able to convert the electricity that's produced by PV modules into a form that's usable. But the inverter efficiency is something that we definitely need to look at when we're looking at system production. Inverter efficiency across the board industry standard is usually pretty similar. We're usually looking at about 98% conversion efficiency, depending on which inverter that you're looking at, but 98% is typical near full power and inverters don't degrade like modules do over time, so you don't see that half a percent decrease in output capacity per year like we see in PV modules.

Francis Plourde:

Inverters tend to be replaced or rebuilt at least once or twice during the project lifecycle, depending on different issues that come up or excessive maintenance or things like that that may need to be performed.

Francis Plourde:

When we're looking at inverters and their impact on energy production, we typically focus on what we call uptime or availability, which is the amount of time that the inverter is operating at full capacity, at the capacity that you expect to see, based on the given conditions and how it is actually performing at that time. So faulted inverters will not produce any energy. So uptime in that case is the most critical parameter that regards actual energy production. We'll reserve that discussion for another time, since it can be quite lengthy. But in this case the thing to keep in mind is that you want to select reliable inverters in the design phase and then also to budget for proper maintenance in your financial model, because regular inverter inspections and maintenance and repair can, even though it's an upfront cost every year for that additional operation and maintenance cost, that can save you a significant more amount of money down the road if those inverters need to be replaced.

Francis Plourde:

Okay, well, that's good to know. Well, thank you, Frances. I think this is a good place for us to wrap up this discussion. I've covered quite a bit of detail here, especially regarding those PVsyst loss characteristics. Now to summarize, I'd say that shading from on-site equipment surroundings must be taken into account during design and production modeling, and it might lead a designer to actually modify their design if the shading's excessive.

Francis Plourde:

And it's important to know that the shading loss is amplified by module mismatch, some of which is inherent in the modules due to their natural variation in quality and the quality controls, but some of which is caused by their aging in the field as well. The PVsyst software allows us to model all of this, as well as the initial light-induced degradation, which is a one-time effect, whereas the general degradation of modules is an ongoing effect for the life of the project. As for inverters, I think we've agreed we'll defer that to another time because that's kind of a long and separate discussion. Again, thank you for listening. If you have any questions for me, for Francis or the rest of the E3 team, or if you have a suggestion for a future podcast topic, please feel free to reach out to us via email at e3co@e3co. com and we look forward to hearing from you. Thank you.