00:00:00Introduction to PVSys Reports Series
00:01:05Understanding a PVSys Report Structure
00:04:42P50 Values and Probabilistic Energy Modeling
00:08:14Evaluating System Performance Against Predictions
00:12:17Key Takeaways and Conclusions
MENGENAI EPISOD INI
In the third episode of the PVsyst series, Daniel Tarico and Frances Wilberg-Plourd from E3 discuss the interpretation of PVsyst reports and how they relate to PV system energy estimation, specifically focusing on how PVsyst calculates energy production using statistical probabilities.
Key points:
- PVsyst Report Overview: The report begins with system inputs such as equipment, location, orientation, and shading factors, and then presents energy estimates, including monthly generation based on climate data. The final report includes an energy waterfall diagram showing energy production and losses on both the DC and AC sides of the system.
- P50 and Probabilistic Estimates: The P50 value represents the long-term average expected energy production. It's based on probabilistic modeling, considering weather variability. The system's actual performance will, in some years, exceed or fall short of the P50 estimate, with a 50/50 chance of being above or below the predicted value.
- Other Probabilistic Levels (P75, P99): Other probabilities, like P75 and P99, indicate more conservative estimates. The P99 represents energy production that the system is expected to exceed in 99% of years. These are useful for ensuring energy and revenue guarantees.
- Degradation and Performance Tracking: Module degradation typically leads to a half-percent annual decrease in energy production, impacting long-term projections. If system production drops faster than expected, it might indicate issues requiring attention.
The episode emphasizes the importance of accurate and up-to-date data in PVsyst reports to ensure reliable energy and financial forecasts.
Bahasa Inggeris
Amerika Syarikat
Tunjukkan NOTA 🔗
TRANSKRIP 🔗
WEBVTT
00:00:09.169 --> 00:00:16.356
Welcome to the Top 3 by E3, a monthly podcast about the intersection between engineering, energy and project finance.
00:00:16.356 --> 00:00:21.010
I'm Daniel Tarico, director of Renewables at E3, and I'll be your host today.
00:00:21.010 --> 00:00:27.423
Today, I'd like to reintroduce Francis Wilberg-Plourd, project manager, here at E3.
00:00:27.423 --> 00:00:39.844
This is the third in a series of podcasts where Francis and I have been discussing several topics related to PV system energy estimation specifically to the PVsyst software.
00:00:39.844 --> 00:00:52.213
We'll be discussing how to interpret the output of a PVSys model, how to read a PVsyst report and how PVsyst uses statistical probabilities to estimate long-term PVSys performance.
00:00:52.213 --> 00:00:53.594
Welcome, Frances.
00:00:54.981 --> 00:00:55.563
Thanks, Dan.
00:00:55.563 --> 00:00:57.729
It's great to be here discussing this with you today.
00:00:57.729 --> 00:01:04.674
I think it's so important for site owners, developers and financiers to understand how to interpret the results of a PVsyst model.
00:01:05.900 --> 00:01:06.561
Absolutely.
00:01:06.561 --> 00:01:14.725
What I have seen is the report from PVsyst is often the only aspect of the PVSYS model that some stakeholders will ever see.
00:01:14.725 --> 00:01:20.051
There's a lot of information in the report and it can take some time and experience to know how to interpret it.
00:01:20.051 --> 00:01:30.612
Speaking from a personal experience, I know when I first started looking at the full PVsyst reports they seemed like a lot of numbers on a page and a lot of the terminology and parameters that I didn't really understand.
00:01:30.612 --> 00:01:41.212
It took a while to get to where I can now go through the report, understand where there might be issues and focus on the details that may be important for the specific project.
00:01:41.212 --> 00:01:49.546
Reflecting on that and with our audience in mind, can you give a brief description of a PVsyst report and what information it contains?
00:01:50.808 --> 00:01:52.171
Absolutely so.
00:01:52.171 --> 00:01:58.531
Before we begin, it's important to note that the projects will often have PVSys models for each iteration of system design.
00:01:58.531 --> 00:02:11.792
So you'll have one PVsyst model at the initial conception of the project and then, as the project is better defined, things like location and equipment type and equipment availability get hammered out.
00:02:11.792 --> 00:02:24.450
Then you'll have different iterations of the report and different results of the report, and so it's important to confirm that whatever PVsyst report you're looking at is the most recent and accurate report available for that project.
00:02:24.450 --> 00:02:37.383
I've seen cases of energy and financial predictions made from PVsyst reports which were created prior to final design using different types of PV modules or inverters than what was actually installed on site.
00:02:37.383 --> 00:02:42.585
So, to start with, make sure you have the most accurate and recently available PVSys report.
00:02:43.447 --> 00:02:44.188
Yeah that makes sense.
00:02:45.091 --> 00:02:45.592
Absolutely.
00:02:45.592 --> 00:02:52.929
The first page of a PVsyst report is going to contain a summary of the design and input parameters that are used to create that model.
00:02:52.929 --> 00:03:06.871
This will include the system equipment, so the modules and inverters, module orientation, which way they're tilted and what direction they're facing, any shading issues and climate data sources used.
00:03:06.871 --> 00:03:12.451
The report will next detail whatever loss factors are incorporated into the PVSys model.
00:03:12.451 --> 00:03:21.149
So this can include losses due to the system itself, things like omic losses and thermal losses, or externally caused losses such as shading.
00:03:21.149 --> 00:03:32.545
Some of this we'll be discussing in a later podcast and, as I said, these can include losses from all different aspects of the system.
00:03:32.545 --> 00:03:37.944
So the report then details the energy produced by the system for each month of the year beginning in January, and this is known as seasonality.
00:03:38.085 --> 00:03:50.775
So it's looking at how much energy the system is capable of producing in each month of the year, depending on the climate conditions and other system conditions present at the site during the different times of the year.
00:03:50.775 --> 00:04:06.007
For financiers building monthly cash flow models, it's incredibly important to know the monthly estimated generation, as it is the annual estimated generation, so you know specifically what you're looking to make, what you're looking to produce in each month of the year.
00:04:07.028 --> 00:04:07.229
Yeah.
00:04:08.216 --> 00:04:25.536
The final page of the report contains what we call an energy waterfall diagram, which explains each component of energy production or loss from each system parameter on both the AC and the DC sides of the system production seasides of the system production.
00:04:25.536 --> 00:04:32.882
So the final result of the report is usually labeled energy injected into the grid, and this represents the energy that is finally transmitted out of the system into the electric grid where it can be consumed.
00:04:33.684 --> 00:04:35.276
And the project can get paid for that energy.
00:04:35.276 --> 00:04:44.367
So, to summarize, the report itself starts with a list of inputs, let's say the equipment, the layout, the location, the meteorological data set, etc.
00:04:44.367 --> 00:04:52.589
And at the end you have a monthly and annual estimations of the energy that would be delivered to the revenue meter.
00:04:53.774 --> 00:04:55.060
Yeah, that's a very good summary.
00:04:55.855 --> 00:05:01.127
Okay, now I would like to talk about the energy production estimates that the report contains.
00:05:01.127 --> 00:05:10.776
As we've explained, the PVsyst report presents the estimated energy production of the system in its first year of operation, let's say, with emphasis on first year.
00:05:10.776 --> 00:05:16.297
Can you give some detail on how the model calculates these estimated energy production values?
00:05:17.319 --> 00:05:18.663
Definitely so.
00:05:18.663 --> 00:05:28.903
The estimated energy production of the PVsyst system in the first year of operation, which is the final point on the energy waterfall of a PV system report, is known as the P50 value.
00:05:29.584 --> 00:05:34.341
Right, yeah, the P50 is really important in project contracts and financing.
00:05:35.684 --> 00:05:43.267
So, simply put, the P50 value is the output that the model calculates as essentially the long-term average system energy production.
00:05:43.267 --> 00:05:54.127
As essentially the long-term average system energy production, it's used most for modeling and financial prediction purposes because it represents the best statistical average for expected PV system production, given the system parameters.
00:05:54.127 --> 00:06:09.189
So since we know that weather conditions are never going to be the same from year to year, it's impossible to produce an estimation for energy production per PV system for a specific year and expect that it's going to match that production exactly.
00:06:09.189 --> 00:06:19.257
So instead we use a probabilistic approach to determine the relative likelihood of certain weather conditions and the associated impacts on energy production at that system.
00:06:19.257 --> 00:06:37.365
So during 50% of years the system is expected to produce less than the P50 estimate, and during 50% of years the system is expected to produce more than the P50 estimate, so that averages out to the P50 production estimate over the lifespan of the system.
00:06:38.485 --> 00:06:43.166
Right, okay, so let me interpret that and how the project is actually going to function.
00:06:43.166 --> 00:06:59.692
We know the weather is variable year to year, month to month, and the P50 is really about the weather more than anything else and it seems like in one year we have basically a 50-50 chance of doing better or worse than the P50 generation, really based on the weather.
00:06:59.692 --> 00:07:00.793
Is that correct?
00:07:05.375 --> 00:07:05.656
Yes, exactly so.
00:07:05.656 --> 00:07:09.343
The P50 is technically called an exceedance probability that is calculated from the weather and the system parameters.
00:07:10.605 --> 00:07:13.699
Okay, now I know from looking at a lot of PVsyst reports.
00:07:13.699 --> 00:07:17.937
There are other levels of probabilities that are considered in these models as well.
00:07:17.937 --> 00:07:20.545
Right, Could you discuss those briefly?
00:07:22.071 --> 00:07:22.572
There are.
00:07:22.572 --> 00:07:32.730
There are definitely some other levels of p-values, which you sometimes see in financial models for PV systems such as the P75 and the P99.
00:07:32.730 --> 00:07:36.963
But the concept is the same and again, it's based on weather variability.
00:07:36.963 --> 00:07:50.569
So the number after the p indicates the percentage of years in which the production of the PV system will exceed that value based on the amount of sunlight or irradiance incident on the modules.
00:07:51.312 --> 00:08:01.480
So the higher the number is after the letter P, the smaller the estimated energy value is, meaning the probability of exceeding that estimated energy value is higher.
00:08:01.480 --> 00:08:07.437
The output parameters of PVSYS models can be adjusted to export these p-values.
00:08:07.437 --> 00:08:19.315
So aside from the P50, the P99 is the most commonly used p-value that we look at because the P-99 is the energy production quantity that the system would be statistically expected to exceed.
00:08:19.315 --> 00:08:24.579
In 99% of years, the system would be expected to produce less than the P-99.
00:08:43.150 --> 00:08:53.905
So if we were a bit more technical, when you say the P50 number produced is using a probabilistic model, what are the probabilities that you are considering in that model?
00:08:54.549 --> 00:09:03.424
The factors included in probabilistic models used to calculate p-values are primarily related to the climate data used to determine the energy system production.
00:09:03.424 --> 00:09:06.806
So environmental conditions and climate data accuracy are prone to much more variability than other aspects of PV system production.
00:09:06.806 --> 00:09:10.923
So environmental conditions and climate data accuracy are prone to much more variability than other aspects of PD system design.
00:09:11.322 --> 00:09:21.951
You're not changing the modules or the inverters you're using year after year, or significantly changing other areas of loss or gain for the system.
00:09:21.951 --> 00:09:28.644
But the weather is going to be the most commonly adjusted input into system performance.
00:09:28.644 --> 00:09:43.110
So these uncertainties can be caused by variances in data collection equipment and analysis, or different changes in climate, data modeling systems and data sets used, and then other changes in the local environment.
00:09:43.110 --> 00:09:50.558
If you're seeing different weather patterns emerge or things like that, that can all change how the system behaves.
00:09:51.280 --> 00:09:52.443
Yeah, I see.
00:09:52.443 --> 00:09:54.549
So I guess there's two things there.
00:09:54.549 --> 00:09:57.317
One is the predicted performance and then I'm guessing.
00:09:57.317 --> 00:10:07.524
Another important question is how can this information be used to determine if an operating PV system is meeting its production expectations or is performing as designed?
00:10:09.211 --> 00:10:20.283
So, as we've been discussing the PV, the P50 and other P values are probabilistic estimations of power production which means that some individual years may have lowered production than the P50.
00:10:20.283 --> 00:10:29.572
This is when the P99 should be consulted, since the P99 effectively represents the floor of expected energy production for the system.
00:10:29.572 --> 00:10:36.195
If the energy production of the system in any one year is lower than the P value under consideration.
00:10:36.195 --> 00:10:38.860
The first aspect you should check is the environmental data.
00:10:39.322 --> 00:10:39.582
Right.
00:10:40.009 --> 00:10:47.791
Was this a cloudier year than normal for that location, or are there new shading issues on site that were not contemplated by the PVSys model?
00:10:47.791 --> 00:10:56.919
So if you had a new building that popped up right next to your PV system that is casting shade on the system that you weren't expecting, that's definitely something to consider.
00:10:57.600 --> 00:10:57.861
Right.
00:10:58.509 --> 00:11:05.599
So, after you've looked at all of those impacts to the system, other aspects of system under performance should be considered, such as equipment failure.
00:11:06.630 --> 00:11:08.916
Yeah, I mean we often see this in performance reports.
00:11:08.916 --> 00:11:13.197
You have generation versus budgeted and the budget is usually based on a P50.
00:11:13.197 --> 00:11:24.697
And we'll often see the actual generation versus weather adjusted, which takes account of the actual measured installation versus the P50 installation.
00:11:24.697 --> 00:11:31.937
We can also look at long-term production system degradation, which is honestly very important for the owner in their revenue projections.
00:11:31.937 --> 00:11:36.394
What are your thoughts there, at least as regards the PV sys performance model?
00:11:37.918 --> 00:11:57.082
So the industry standard rule of thumb is to assume a half a percent loss in DC power generation capacity per year of operation and this comes from the standard assumption for PV module annual degradation which assumes a half a percent loss in efficiency and power production per year at the module level.
00:11:57.082 --> 00:12:09.892
So this can be used as a rule of thumb estimation to determine estimated energy production for subsequent years by derating the P50 value for the first year's energy production by half a percent per year.
00:12:09.892 --> 00:12:18.422
The exact amount of annual derating of energy production estimation may vary depending on the type of equipment you're using or the environment you're using them in.
00:12:18.422 --> 00:12:23.522
Most of them that I've seen are in the half a percent per year ballpark range.
00:12:24.710 --> 00:12:28.440
Okay, yeah, well, I see this a lot.
00:12:28.440 --> 00:12:34.639
What I see is often between 0.4% and 0.7%, depending on the module warranted maximum degradation.
00:12:34.639 --> 00:12:43.375
It's also important to note that these module degradation loss are on the DC side, where the P50 estimates the energy production on the AC side of the system.
00:12:43.375 --> 00:12:54.072
And honestly, I guess at this point we'd be getting into a long discussion on system availability and other system degradation parameters, but that's a whole other subject we're not going to discuss today.
00:12:55.135 --> 00:13:01.895
Yeah, exactly, we know that DC and AC losses don't exactly correlate one-to-one due to various design reasons.
00:13:01.895 --> 00:13:09.441
But you can assume that they're going to be pretty close because the electrical energy going into the AC part of the system comes from the DC side.
00:13:09.461 --> 00:13:09.923
Yeah.
00:13:10.149 --> 00:13:16.744
So this half a percent per year value is useful as a general assumption for annual system degradation.
00:13:16.744 --> 00:13:27.284
So if the system production decreases more intensely or more quickly than this, it's a good snapshot of potentially systemic issues that require remediation.
00:13:27.970 --> 00:13:29.173
Right, that makes sense.
00:13:29.173 --> 00:13:42.981
Okay, now if we were to summarize this technical discussion, what do you think are the most important points for someone who's relying on PV values to either put together a financial model or a performance contract with their service provider?
00:13:44.009 --> 00:14:03.975
So my top three points to summarize our conversation today would be number one, to know that PVSys reports contain details on system inputs and expected production, and that understanding how to interpret PVSys reports is essential to understanding PV system utilization and operation.
00:14:03.975 --> 00:14:19.778
Number two, that P values represent probabilistic energy production estimates and can be used as baselines for energy and revenue guarantees, and that P values are very useful for quick checks of system performance against expected prediction values.
00:14:19.778 --> 00:14:34.352
And number three, pvsyst reports and P values are only as accurate as their input information, so make sure that the reports and the values that you're using to make energy production estimations are as accurate and up-to-date as possible.
00:14:34.874 --> 00:14:35.855
Yeah, that makes sense.
00:14:35.855 --> 00:14:38.383
Well, thank you for that discussion, Frances.
00:14:38.383 --> 00:14:45.491
This was very helpful to understand how PVSYST reports can be interpreted and used in establishing contracts.
00:14:45.491 --> 00:14:59.179
If you have any questions for me or 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, at e3cocom.
00:14:59.179 --> 00:15:02.475
We look forward to hearing from you and have a great day.
00:00:09.169 --> 00:00:16.356
Welcome to the Top 3 by E3, a monthly podcast about the intersection between engineering, energy and project finance.
00:00:16.356 --> 00:00:21.010
I'm Daniel Tarico, director of Renewables at E3, and I'll be your host today.
00:00:21.010 --> 00:00:27.423
Today, I'd like to reintroduce Francis Wilberg-Plourd, project manager, here at E3.
00:00:27.423 --> 00:00:39.844
This is the third in a series of podcasts where Francis and I have been discussing several topics related to PV system energy estimation specifically to the PVsyst software.
00:00:39.844 --> 00:00:52.213
We'll be discussing how to interpret the output of a PVSys model, how to read a PVsyst report and how PVsyst uses statistical probabilities to estimate long-term PVSys performance.
00:00:52.213 --> 00:00:53.594
Welcome, Frances.
00:00:54.981 --> 00:00:55.563
Thanks, Dan.
00:00:55.563 --> 00:00:57.729
It's great to be here discussing this with you today.
00:00:57.729 --> 00:01:04.674
I think it's so important for site owners, developers and financiers to understand how to interpret the results of a PVsyst model.
00:01:05.900 --> 00:01:06.561
Absolutely.
00:01:06.561 --> 00:01:14.725
What I have seen is the report from PVsyst is often the only aspect of the PVSYS model that some stakeholders will ever see.
00:01:14.725 --> 00:01:20.051
There's a lot of information in the report and it can take some time and experience to know how to interpret it.
00:01:20.051 --> 00:01:30.612
Speaking from a personal experience, I know when I first started looking at the full PVsyst reports they seemed like a lot of numbers on a page and a lot of the terminology and parameters that I didn't really understand.
00:01:30.612 --> 00:01:41.212
It took a while to get to where I can now go through the report, understand where there might be issues and focus on the details that may be important for the specific project.
00:01:41.212 --> 00:01:49.546
Reflecting on that and with our audience in mind, can you give a brief description of a PVsyst report and what information it contains?
00:01:50.808 --> 00:01:52.171
Absolutely so.
00:01:52.171 --> 00:01:58.531
Before we begin, it's important to note that the projects will often have PVSys models for each iteration of system design.
00:01:58.531 --> 00:02:11.792
So you'll have one PVsyst model at the initial conception of the project and then, as the project is better defined, things like location and equipment type and equipment availability get hammered out.
00:02:11.792 --> 00:02:24.450
Then you'll have different iterations of the report and different results of the report, and so it's important to confirm that whatever PVsyst report you're looking at is the most recent and accurate report available for that project.
00:02:24.450 --> 00:02:37.383
I've seen cases of energy and financial predictions made from PVsyst reports which were created prior to final design using different types of PV modules or inverters than what was actually installed on site.
00:02:37.383 --> 00:02:42.585
So, to start with, make sure you have the most accurate and recently available PVSys report.
00:02:43.447 --> 00:02:44.188
Yeah that makes sense.
00:02:45.091 --> 00:02:45.592
Absolutely.
00:02:45.592 --> 00:02:52.929
The first page of a PVsyst report is going to contain a summary of the design and input parameters that are used to create that model.
00:02:52.929 --> 00:03:06.871
This will include the system equipment, so the modules and inverters, module orientation, which way they're tilted and what direction they're facing, any shading issues and climate data sources used.
00:03:06.871 --> 00:03:12.451
The report will next detail whatever loss factors are incorporated into the PVSys model.
00:03:12.451 --> 00:03:21.149
So this can include losses due to the system itself, things like omic losses and thermal losses, or externally caused losses such as shading.
00:03:21.149 --> 00:03:32.545
Some of this we'll be discussing in a later podcast and, as I said, these can include losses from all different aspects of the system.
00:03:32.545 --> 00:03:37.944
So the report then details the energy produced by the system for each month of the year beginning in January, and this is known as seasonality.
00:03:38.085 --> 00:03:50.775
So it's looking at how much energy the system is capable of producing in each month of the year, depending on the climate conditions and other system conditions present at the site during the different times of the year.
00:03:50.775 --> 00:04:06.007
For financiers building monthly cash flow models, it's incredibly important to know the monthly estimated generation, as it is the annual estimated generation, so you know specifically what you're looking to make, what you're looking to produce in each month of the year.
00:04:07.028 --> 00:04:07.229
Yeah.
00:04:08.216 --> 00:04:25.536
The final page of the report contains what we call an energy waterfall diagram, which explains each component of energy production or loss from each system parameter on both the AC and the DC sides of the system production seasides of the system production.
00:04:25.536 --> 00:04:32.882
So the final result of the report is usually labeled energy injected into the grid, and this represents the energy that is finally transmitted out of the system into the electric grid where it can be consumed.
00:04:33.684 --> 00:04:35.276
And the project can get paid for that energy.
00:04:35.276 --> 00:04:44.367
So, to summarize, the report itself starts with a list of inputs, let's say the equipment, the layout, the location, the meteorological data set, etc.
00:04:44.367 --> 00:04:52.589
And at the end you have a monthly and annual estimations of the energy that would be delivered to the revenue meter.
00:04:53.774 --> 00:04:55.060
Yeah, that's a very good summary.
00:04:55.855 --> 00:05:01.127
Okay, now I would like to talk about the energy production estimates that the report contains.
00:05:01.127 --> 00:05:10.776
As we've explained, the PVsyst report presents the estimated energy production of the system in its first year of operation, let's say, with emphasis on first year.
00:05:10.776 --> 00:05:16.297
Can you give some detail on how the model calculates these estimated energy production values?
00:05:17.319 --> 00:05:18.663
Definitely so.
00:05:18.663 --> 00:05:28.903
The estimated energy production of the PVsyst system in the first year of operation, which is the final point on the energy waterfall of a PV system report, is known as the P50 value.
00:05:29.584 --> 00:05:34.341
Right, yeah, the P50 is really important in project contracts and financing.
00:05:35.684 --> 00:05:43.267
So, simply put, the P50 value is the output that the model calculates as essentially the long-term average system energy production.
00:05:43.267 --> 00:05:54.127
As essentially the long-term average system energy production, it's used most for modeling and financial prediction purposes because it represents the best statistical average for expected PV system production, given the system parameters.
00:05:54.127 --> 00:06:09.189
So since we know that weather conditions are never going to be the same from year to year, it's impossible to produce an estimation for energy production per PV system for a specific year and expect that it's going to match that production exactly.
00:06:09.189 --> 00:06:19.257
So instead we use a probabilistic approach to determine the relative likelihood of certain weather conditions and the associated impacts on energy production at that system.
00:06:19.257 --> 00:06:37.365
So during 50% of years the system is expected to produce less than the P50 estimate, and during 50% of years the system is expected to produce more than the P50 estimate, so that averages out to the P50 production estimate over the lifespan of the system.
00:06:38.485 --> 00:06:43.166
Right, okay, so let me interpret that and how the project is actually going to function.
00:06:43.166 --> 00:06:59.692
We know the weather is variable year to year, month to month, and the P50 is really about the weather more than anything else and it seems like in one year we have basically a 50-50 chance of doing better or worse than the P50 generation, really based on the weather.
00:06:59.692 --> 00:07:00.793
Is that correct?
00:07:05.375 --> 00:07:05.656
Yes, exactly so.
00:07:05.656 --> 00:07:09.343
The P50 is technically called an exceedance probability that is calculated from the weather and the system parameters.
00:07:10.605 --> 00:07:13.699
Okay, now I know from looking at a lot of PVsyst reports.
00:07:13.699 --> 00:07:17.937
There are other levels of probabilities that are considered in these models as well.
00:07:17.937 --> 00:07:20.545
Right, Could you discuss those briefly?
00:07:22.071 --> 00:07:22.572
There are.
00:07:22.572 --> 00:07:32.730
There are definitely some other levels of p-values, which you sometimes see in financial models for PV systems such as the P75 and the P99.
00:07:32.730 --> 00:07:36.963
But the concept is the same and again, it's based on weather variability.
00:07:36.963 --> 00:07:50.569
So the number after the p indicates the percentage of years in which the production of the PV system will exceed that value based on the amount of sunlight or irradiance incident on the modules.
00:07:51.312 --> 00:08:01.480
So the higher the number is after the letter P, the smaller the estimated energy value is, meaning the probability of exceeding that estimated energy value is higher.
00:08:01.480 --> 00:08:07.437
The output parameters of PVSYS models can be adjusted to export these p-values.
00:08:07.437 --> 00:08:19.315
So aside from the P50, the P99 is the most commonly used p-value that we look at because the P-99 is the energy production quantity that the system would be statistically expected to exceed.
00:08:19.315 --> 00:08:24.579
In 99% of years, the system would be expected to produce less than the P-99.
00:08:43.150 --> 00:08:53.905
So if we were a bit more technical, when you say the P50 number produced is using a probabilistic model, what are the probabilities that you are considering in that model?
00:08:54.549 --> 00:09:03.424
The factors included in probabilistic models used to calculate p-values are primarily related to the climate data used to determine the energy system production.
00:09:03.424 --> 00:09:06.806
So environmental conditions and climate data accuracy are prone to much more variability than other aspects of PV system production.
00:09:06.806 --> 00:09:10.923
So environmental conditions and climate data accuracy are prone to much more variability than other aspects of PD system design.
00:09:11.322 --> 00:09:21.951
You're not changing the modules or the inverters you're using year after year, or significantly changing other areas of loss or gain for the system.
00:09:21.951 --> 00:09:28.644
But the weather is going to be the most commonly adjusted input into system performance.
00:09:28.644 --> 00:09:43.110
So these uncertainties can be caused by variances in data collection equipment and analysis, or different changes in climate, data modeling systems and data sets used, and then other changes in the local environment.
00:09:43.110 --> 00:09:50.558
If you're seeing different weather patterns emerge or things like that, that can all change how the system behaves.
00:09:51.280 --> 00:09:52.443
Yeah, I see.
00:09:52.443 --> 00:09:54.549
So I guess there's two things there.
00:09:54.549 --> 00:09:57.317
One is the predicted performance and then I'm guessing.
00:09:57.317 --> 00:10:07.524
Another important question is how can this information be used to determine if an operating PV system is meeting its production expectations or is performing as designed?
00:10:09.211 --> 00:10:20.283
So, as we've been discussing the PV, the P50 and other P values are probabilistic estimations of power production which means that some individual years may have lowered production than the P50.
00:10:20.283 --> 00:10:29.572
This is when the P99 should be consulted, since the P99 effectively represents the floor of expected energy production for the system.
00:10:29.572 --> 00:10:36.195
If the energy production of the system in any one year is lower than the P value under consideration.
00:10:36.195 --> 00:10:38.860
The first aspect you should check is the environmental data.
00:10:39.322 --> 00:10:39.582
Right.
00:10:40.009 --> 00:10:47.791
Was this a cloudier year than normal for that location, or are there new shading issues on site that were not contemplated by the PVSys model?
00:10:47.791 --> 00:10:56.919
So if you had a new building that popped up right next to your PV system that is casting shade on the system that you weren't expecting, that's definitely something to consider.
00:10:57.600 --> 00:10:57.861
Right.
00:10:58.509 --> 00:11:05.599
So, after you've looked at all of those impacts to the system, other aspects of system under performance should be considered, such as equipment failure.
00:11:06.630 --> 00:11:08.916
Yeah, I mean we often see this in performance reports.
00:11:08.916 --> 00:11:13.197
You have generation versus budgeted and the budget is usually based on a P50.
00:11:13.197 --> 00:11:24.697
And we'll often see the actual generation versus weather adjusted, which takes account of the actual measured installation versus the P50 installation.
00:11:24.697 --> 00:11:31.937
We can also look at long-term production system degradation, which is honestly very important for the owner in their revenue projections.
00:11:31.937 --> 00:11:36.394
What are your thoughts there, at least as regards the PV sys performance model?
00:11:37.918 --> 00:11:57.082
So the industry standard rule of thumb is to assume a half a percent loss in DC power generation capacity per year of operation and this comes from the standard assumption for PV module annual degradation which assumes a half a percent loss in efficiency and power production per year at the module level.
00:11:57.082 --> 00:12:09.892
So this can be used as a rule of thumb estimation to determine estimated energy production for subsequent years by derating the P50 value for the first year's energy production by half a percent per year.
00:12:09.892 --> 00:12:18.422
The exact amount of annual derating of energy production estimation may vary depending on the type of equipment you're using or the environment you're using them in.
00:12:18.422 --> 00:12:23.522
Most of them that I've seen are in the half a percent per year ballpark range.
00:12:24.710 --> 00:12:28.440
Okay, yeah, well, I see this a lot.
00:12:28.440 --> 00:12:34.639
What I see is often between 0.4% and 0.7%, depending on the module warranted maximum degradation.
00:12:34.639 --> 00:12:43.375
It's also important to note that these module degradation loss are on the DC side, where the P50 estimates the energy production on the AC side of the system.
00:12:43.375 --> 00:12:54.072
And honestly, I guess at this point we'd be getting into a long discussion on system availability and other system degradation parameters, but that's a whole other subject we're not going to discuss today.
00:12:55.135 --> 00:13:01.895
Yeah, exactly, we know that DC and AC losses don't exactly correlate one-to-one due to various design reasons.
00:13:01.895 --> 00:13:09.441
But you can assume that they're going to be pretty close because the electrical energy going into the AC part of the system comes from the DC side.
00:13:09.461 --> 00:13:09.923
Yeah.
00:13:10.149 --> 00:13:16.744
So this half a percent per year value is useful as a general assumption for annual system degradation.
00:13:16.744 --> 00:13:27.284
So if the system production decreases more intensely or more quickly than this, it's a good snapshot of potentially systemic issues that require remediation.
00:13:27.970 --> 00:13:29.173
Right, that makes sense.
00:13:29.173 --> 00:13:42.981
Okay, now if we were to summarize this technical discussion, what do you think are the most important points for someone who's relying on PV values to either put together a financial model or a performance contract with their service provider?
00:13:44.009 --> 00:14:03.975
So my top three points to summarize our conversation today would be number one, to know that PVSys reports contain details on system inputs and expected production, and that understanding how to interpret PVSys reports is essential to understanding PV system utilization and operation.
00:14:03.975 --> 00:14:19.778
Number two, that P values represent probabilistic energy production estimates and can be used as baselines for energy and revenue guarantees, and that P values are very useful for quick checks of system performance against expected prediction values.
00:14:19.778 --> 00:14:34.352
And number three, pvsyst reports and P values are only as accurate as their input information, so make sure that the reports and the values that you're using to make energy production estimations are as accurate and up-to-date as possible.
00:14:34.874 --> 00:14:35.855
Yeah, that makes sense.
00:14:35.855 --> 00:14:38.383
Well, thank you for that discussion, Frances.
00:14:38.383 --> 00:14:45.491
This was very helpful to understand how PVSYST reports can be interpreted and used in establishing contracts.
00:14:45.491 --> 00:14:59.179
If you have any questions for me or 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, at e3cocom.
00:14:59.179 --> 00:15:02.475
We look forward to hearing from you and have a great day.
CARI EPISOD LEPAS
Cari episod lepas bagi The Top 3 by E3 .
Penafian: Podcast dan karya seni yang dibenamkan pada halaman ini adalah daripada E3 Consulting, yang merupakan hak milik pemiliknya dan tidak bergabung dengan atau disokong oleh Listen Notes, Inc.
SUNTING
Terima kasih kerana membantu memastikan pangkalan data podcast dalam keadaan terkini.
