7 Points to Note When Reading Energy Production in PVSyst

Table of Contents

• PVSyst's energy production is a "predicted value" and not a "guaranteed value"

• Point 1: Verify the source and representativeness of the irradiance data

• Point 2: Look at PR and Specific Production, not just energy production

• Point 3: Break down and check the assumptions for loss items

• Point 4: Do not judge the impact of shading solely by annual energy production

• Point 5: Check the relationship between temperature losses and installation conditions

• Point 6: Do not misinterpret PCS capacity and output limitations

• Point 7: When comparing with measured values, align the conditions

• Practical points that are easy to overlook when reading PVSyst's energy production

• Approach to linking on-site verification and simulation

• Summary

PVSyst's energy production is a "predicted value" and not a "guaranteed value"

PVSyst is a critically important simulation software for assessing the power generation of solar power plants. It is used in various situations such as plant design, financing documentation, energy-yield assessments, third-party reviews, and performance verification of existing plants. Especially for commercial-scale solar, it is often treated as the standard reference document for comprehensively checking annual energy production, PR, loss items, solar irradiance, PCS output, temperature losses, and other related factors.

However, PVSyst’s estimated energy production is a predicted value calculated based on the input conditions and does not guarantee the actual generation. If this is misunderstood, one may judge a plant’s quality too quickly by looking only at the annual energy production shown in the report.

The energy production displayed by PVSyst is determined by the accumulation of many underlying assumptions, such as meteorological data, module specifications, PCS specifications, azimuth, tilt angle, shading, wiring losses, mismatch, soiling, temperature conditions, and output limitations. Therefore, rather than looking only at the production numbers, it is important to read which assumptions those numbers are derived from.

For example, even when analyzing the same power plant, the energy production will change depending on the type of irradiance data. If the wiring loss or soiling loss settings differ, the results will change as well. Whether the shading model is detailed or simplified also alters how annual and monthly generation appear. In other words, when viewing PVSyst’s generation results, you should prioritize checking the conditions that led to those results rather than the results themselves.

This article organizes seven points to watch when interpreting PVSyst energy production, arranged in the order that is easiest to verify in practice. It is useful not only for first-time readers of PVSyst reports, but also for those comparing multiple proposals, comparing analyses from other firms, or validating against measured data.

Note 1: Verify the source and representativeness of solar irradiance data

When reading energy production in PVSyst, the first thing to check is the solar irradiance data. The energy production of a photovoltaic system is fundamentally strongly influenced by how much sunlight is incident. Therefore, simply changing which solar irradiance dataset you choose can significantly alter the annual energy production.

In PVSyst, various weather data sources may be used, such as Meteonorm, SolarGIS, satellite data, nearby station data, and measured meteorological data. If you look only at the energy production without checking which data were used, you cannot determine whether that production is conservative, typical, or somewhat optimistic.

One point to pay particular attention to is that, even at the same location, annual horizontal-plane solar radiation and tilted-surface solar radiation can differ depending on the data source. In some datasets the annual solar radiation is reported as high, while in others it is reported as low. This is because measurement methods, satellite-based estimates, statistical periods, interpolation methods, the distance to nearby observation points, and so on differ.

When reading PVSyst's energy production, first check values such as GlobHor, GlobInc, and DiffHor. GlobHor is usually read as the global irradiation on the horizontal plane, and GlobInc is often read as the irradiation incident on the module surface; these are important upstream indicators of energy production. If the annual energy production is high, it is necessary to distinguish whether that is due to high system efficiency or simply because the assumed solar irradiation is high.

Monthly solar radiation is also important. Even if the annual total appears reasonable, there can be cases where it is excessively high only in winter, too low only in summer, or where winter solar radiation and reflection conditions are unrealistic despite being a snow-prone area. Especially in Hokkaido, Tohoku, the Sea of Japan side, and mountainous regions, factors such as snow cover, fog, cloudiness, and surrounding topography can mean that standard meteorological data alone may not fully represent actual conditions.

When evaluating power generation, it is important to check not only the annual generation but also the source of the solar irradiance data, the statistical period, how close the measurement site is, and the monthly trends. Keeping in mind the obvious relationship that higher irradiance tends to result in higher generation and lower irradiance tends to result in lower generation will help you read PVSyst results objectively.

Point 2: Look at PR and Specific Production, not just energy production

When looking at a PVSyst report, many people first check the annual energy production. From the perspective of how many kWh will be generated annually, how much can be expected to be sold, and whether it will be sufficient for the business plan, energy production is of course important.

However, looking at generation output alone does not allow you to judge the validity of the design or analysis. Generation increases with larger installed capacity and in regions with higher solar irradiation. Therefore, to compare the performance of power plants, it is necessary to consider PR and Specific Production together.

PR stands for Performance Ratio and is an indicator of how efficiently a solar power generation system produces electrical energy relative to the incident solar irradiance. Generally, it is used to evaluate the overall performance of the system by normalizing differences in irradiance conditions and system capacity to some extent.

Specific Production is an indicator often viewed as the annual electricity generation per 1 kW of installed capacity. For example, if you divide the annual generation by the DC capacity, it is expressed in a form such as kWh/kWp. Looking at this value makes it easier to understand how much generation can be expected relative to the installed capacity.

For example, even if Plan A’s energy production is large and Plan B’s is small, Plan A might simply have a larger capacity. Conversely, Plan B may have better generation efficiency per unit of capacity. In such cases, looking only at energy production does not make for a correct comparison; you must also look at PR and Specific Production.

Also, be cautious if the PR is too high. Possible causes include loss settings being set excessively low, soiling loss not being included, wiring losses being too low, shading not being sufficiently reflected, or temperature conditions being more favorable than reality. Conversely, if the PR is too low, check shading, temperature, wiring, PCS clipping, output limits, snow, and excessive loss settings.

When reviewing generation figures in PVSyst, it's important to check energy production, PR, and Specific Production together. Energy production is the figure for project economics, PR is the figure for system performance, and Specific Production is the figure used to assess productivity per unit of capacity; distinguishing them in this way makes interpreting the results more consistent.

Point 3: Break down and verify the assumptions for loss items

PVSyst's energy production is calculated by subtracting various losses from the solar irradiation energy through to the final electrical energy at the grid connection point. Therefore, to determine whether the energy production is high or low, it is essential to break down and examine the loss items.

PVSyst's Loss Diagram displays irradiance-related losses, near shading, IAM losses, soiling losses, temperature losses, mismatch losses, wiring losses, inverter losses, transformer losses, auxiliary losses, and losses due to output limitation, among others. By checking how much each of these losses is accounted for, you can see the basis for the estimated energy production.

For example, in reports that show high power generation, the soiling loss may be set too low. If the site is in an area with a lot of dust, close to agricultural land, subject to snow or snowmelt soiling, or where bird-related soiling is expected, it is necessary to verify whether setting the soiling loss close to zero is reasonable.

Wiring losses are another item that is easy to overlook. Check which sections’ losses are included—DC wiring, AC wiring, MV wiring, etc. The reasonable loss rate differs between cases where PCS are distributed and cases with long-distance collection cabling. If wiring length, cable size, voltage, current, or circuit configuration differ, wiring losses will vary even for power plants of the same capacity.

Temperature losses are also important. Because higher module temperatures reduce output, factors such as the type of mounting structure, whether the system is roof-mounted or ground-mounted, the quality of rear-side ventilation, installation height, and wind speed conditions all have an impact. In PVSyst, temperature losses change depending on the temperature model settings, so you need to check that the assumptions match the installation conditions.

Also, inverter losses and clipping losses should be taken into account. If the PCS capacity is smaller than the DC capacity, output can be capped on sunny days and energy that could have been generated may be lost. This can commonly occur in oversized designs, but if you do not understand how large the losses are, you will misestimate the power generation.

Loss items cannot be judged simply by their size. It is important to verify whether a loss is reasonable in relation to site conditions, design conditions, equipment specifications, and operating conditions. When reading PVSyst's energy production, treat the Loss Diagram as a breakdown of the energy production and, as a basic approach, check in sequence where and by how much the production is being reduced.

Note 4: Do not judge the impact of shading solely by annual energy production

When evaluating power generation in PVSyst, the impact of shading is an aspect that deserves particular attention. Shading not only reduces annual energy production but can also have a significant effect during specific times of day or seasons. Therefore, even if the annual shading losses appear small, they may not be negligible in practice.

Causes of shading at solar power plants include surrounding trees, mountains, buildings, utility poles, fences, substation and transformer equipment, adjacent racking rows, and terrain undulations. In ground-mounted plants, inter-row shading occurs depending on rack spacing in the east–west or north–south direction, the tilt angle, and the azimuth. On roof-mounted installations, parapets, adjacent buildings, rooftop equipment, chimneys, and antennas are sources of shading.

Even if annual shading losses are a few percent or less, shadows can be concentrated in the mornings and evenings or during winter. In typical projects where the feed-in tariff does not vary by time of day, the focus is on the impact on annual power generation, but when considering battery storage co-location, self-consumption, peak shaving, or time-of-day electricity value, the times when shading occurs are also important.

Shading can also be accompanied by electrical losses. When part of a module is shaded, output can drop by more than the simple area ratio due to bypass diode behavior, current limiting within the string, impacts on MPPT units, and other effects. In PVSyst, the accuracy of the results depends on the 3D model of near shadows and how electrical shading effects are treated.

What’s important when interpreting shadows is to confirm that the 3D model reflects actual conditions. If the terrain, surrounding obstacles, mounting-structure height, module surface, or row spacing are oversimplified, shadow-loss estimates can deviate from reality. In particular, on complex terrain or sloped sites, a simple model designed for flat land may not sufficiently represent the effects of shadows.

When checking shading losses in a PVSyst report, do not judge based only on the annual value; check the monthly and hourly breakdowns, the 3D scene, the shading objects, and the settings for electrical impact. Instead of assuming there is no problem because the shading looks small, it is important to verify in which seasons, at what times, and over what areas the shading occurs.

Point 5: Confirm the relationship between temperature loss and installation conditions

Solar modules produce less power as temperature increases. Therefore, when reading energy production in PVSyst, you need to check how much temperature loss is included. The impact of temperature loss is particularly significant when examining summer generation, annual PR, and peak output.

Temperature loss is not determined solely by ambient air temperature. The module’s mounting method, rear ventilation, racking height, distance to the roof, wind speed, and the module’s thermal characteristics all affect it. A module that is ground-mounted with good ventilation will have a different temperature than one installed close to a roof, even at the same ambient temperature.

In PVSyst, module temperature is calculated from the parameters of the temperature model. If these do not reflect actual conditions, temperature losses can be underestimated or overestimated. For example, if an installation actually has poor ventilation but is set as a well-ventilated condition, temperature losses will be smaller and the generated energy may be reported as higher.

Also, the module temperature coefficient is important. Modules with larger temperature coefficients experience greater output reduction as temperature rises. It is also necessary to verify that the values in the module datasheet are correctly reflected in PVSyst’s database and that the module model you are using is correct.

When examining temperature loss, it's easier to understand if you check monthly trends as well as the annual value. It's natural for temperature loss to be greater in summer and smaller in winter. However, if the temperature loss is unusually small or large for the region and installation conditions, you should check the temperature model, the meteorological data, and the installation condition settings.

Temperature losses are also related to improvements in power plant design. Design measures such as ensuring racking height, maintaining clearance from the roof surface, arranging components so as not to obstruct airflow, and avoiding excessively dense layouts can sometimes suppress temperature rise. When reading PVSyst's energy yield, it is important not to view temperature losses simply as calculation results, but to interpret them in connection with the design conditions.

Note 6: Do not misread PCS capacity and output limits

When reading PVSyst's energy yield, the treatment of PCS capacity and output limits is extremely important. In solar power generation, the DC-side module capacity and the AC-side PCS capacity do not necessarily match. Indeed, for commercial solar projects, an oversized design that makes the DC capacity larger than the PCS capacity is often used.

When the system is oversized, the PCS utilization rate during low irradiance and on cloudy days increases, making it easier to raise annual energy generation. On the other hand, on sunny days DC power exceeding the PCS's rated output is fed in, and the AC output becomes capped. The losses caused by this capping appear as clipping losses or losses due to inverter limitations.

When viewing energy production in PVSyst, check the PCS capacity, DC/AC ratio, maximum input voltage, MPPT range, string configuration, and output limiting settings. If the energy production is lower than expected, the PCS capacity may be too small, causing large clipping losses. Conversely, if the energy production appears higher than expected, output limits or power factor conditions may not be adequately reflected.

One thing to pay particular attention to is output limits on grid interconnection. Not only the PCS’s rated output, but also conditions such as the point of supply, the interconnection point, contracted capacity, power-factor operation, and remote output control can change the actual amount of active energy that can be exported. In PVSyst, if you do not check at which location the output is being limited, you may misread the generation output.

For example, even if the total capacity of the PCS is large, if the limit at the point of interconnection is set low, output will be constrained there. Also, when power factor conditions restrict the range of apparent power, it is necessary to correctly understand how active power is handled. It is important to verify what the power factor setting in PVSyst means and where losses are occurring in relation to the PCS rating.

PCS capacity and output limits affect not only annual energy production but also peak generation characteristics, monthly generation, and PR. When reading generation figures in PVSyst reports, you should not simply look at the inverter loss %; you need to check DC capacity, AC capacity, any limiting conditions, and whether clipping occurs.

Note 7: When comparing with measured values, make sure the conditions are consistent

When reviewing energy yield in PVSyst, you may compare simulation results with measured data. For tasks such as performance evaluation of existing plants, PR tests, investigations into the causes of reduced generation, O&M improvements, and verification of the impact of output curtailment, it is important to reconcile PVSyst’s predicted values with the actual generation.

However, when comparing against measured values, you cannot make a correct judgment unless the conditions are aligned. PVSyst's standard analysis is often an annual forecast based on long-term average meteorological conditions and does not match the weather of any particular year. If the solar irradiation in a given year is lower than the long-term average, it is natural that the measured energy production will be lower than PVSyst. Conversely, if the irradiation is higher in a given year, the measured values may exceed PVSyst.

Therefore, when comparing with measured values, first check the solar irradiation during the measurement period. If possible, use on-site pyranometer data, nearby meteorological data, and satellite data to compare PVSyst’s assumed solar irradiation with the actual solar irradiation. Rather than comparing only the energy production, it is important to look at the PR corrected for solar irradiation and the energy production efficiency per unit of solar irradiation.

Also, the location at which the energy being compared is measured is important. Confirm whether the PVSyst results correspond to PCS output, after the transformer, at the point of receipt, or at the export meter. If the measured value is the export meter reading and PVSyst corresponds to PCS output, there may be inconsistencies in the treatment of transformer losses, on-site consumption, and wiring losses.

Missing data also require careful attention. If there are missing pyranometer measurements, PCS stoppages, communication failures, output curtailment, maintenance outages, or accident-related shutdowns, how those periods are handled can change the comparison results. If you simply compare annual totals, stoppages that are not related to equipment performance may appear as reductions in generation.

When comparing measured values with PVSyst, the basic principle is to match the same period, the same solar irradiance conditions, the same measurement points, and the same system operating conditions. Comparisons with mismatched conditions become mere differences in numbers rather than true cause analysis. If you plan to use PVSyst’s generation estimates in practice, it is essential to clarify and organize the comparison conditions with the measured values.

Practical points commonly overlooked when interpreting PVSyst energy yield

When reading PVSyst's energy yield, you need to pay attention not only to the major losses and PR, but also to the detailed practical assumptions. Especially when comparing multiple reports, the energy yields may look similar while the underlying assumptions can differ significantly.

First, what I want to confirm is the definitions of DC capacity and AC capacity. I will check whether the module capacity is based on STC, whether the PCS capacity is based on active (real) power, and whether there are capacity limits at the point of interconnection. If the capacity standards differ, comparisons of Specific Production and PR may be skewed.

Next, confirm that the module and PCS model numbers are correct. Even if model numbers are similar, output, temperature coefficient, efficiency, voltage range, and MPPT specifications can differ. Because the equipment information registered in PVSyst’s database may be outdated or may use an approximate model, it is necessary to cross-check with the datasheet.

Handling soiling losses and snow losses is also important. In Japan, the impact of soiling and snow varies greatly by region. In coastal areas, agricultural areas, mountainous areas, snowy regions, and areas prone to yellow sand, standard loss settings alone may not adequately reflect the actual situation. In particular, in snowy regions it is necessary to check, on a monthly basis, how much power generation can be expected during winter.

Also, the handling of auxiliary equipment losses and transformer losses is something that is easily overlooked. Verify how monitoring devices, air conditioning, communication equipment, trackers, PCS standby power, and transformer no-load and load losses are accounted for. Even if they appear to be small losses, over the course of a year they can add up to a difference that cannot be ignored.

When comparing multiple proposals, it is particularly important to align the comparison conditions. If, for example, Option A includes soiling losses while Option B does not; Option A uses detailed shading while Option B uses simplified shading; or Option A includes transformer losses while Option B is calculated up to the PCS output, then the comparison of energy yields is unfair. When comparing PVSyst energy yields, you should first create a table of input conditions and clearly indicate which conditions are the same and which are different.

PVSyst reports, if you're not familiar with them, tend to draw attention to the final energy production. However, in practice, what matters more than the final energy production is the assumptions under which that figure was produced. Looking at differences in assumptions before looking at differences in numbers is the basic principle for correctly reading PVSyst.

Approach to Connecting On-site Verification and Simulation

PVSyst is a powerful simulation tool, but you cannot fully grasp the site from desk-based input conditions alone. To improve the accuracy of energy production estimates, it is important to reflect on-site inspection information in PVSyst’s assumptions.

For example, the positions of surrounding trees and buildings, terrain undulations, racking orientation, actual tilt, aisle widths, the layout of PCS and cubicles, cable routes, and so on may not be accurately determined from drawings alone. By reviewing shading, wiring losses, temperature conditions, maintainability, and other factors based on information confirmed on site, PVSyst’s energy yield assessment will more closely reflect reality.

Especially when investigating a decline in power generation at an existing plant, it is essential not to rely on simulation alone but to verify on-site conditions. By checking module soiling, shading from weeds, tilt of the mounting structure, damage, wiring abnormalities, PCS stop history, output curtailment history, and the installation condition of the pyranometer, you can concretely identify the causes of discrepancies with PVSyst.

For on-site checks like this, high-precision positioning using smartphones and AR displays is also effective. For example, using a system that combines an iPhone and GNSS, such as LRTK, to obtain highly accurate on-site positions makes it easier to confirm the positional relationship between equipment locations on drawings and their actual positions in the field. If racking, PCS, boundaries, cable routes, inspection points, and so on within the power plant can be linked to and recorded with location information, it becomes easier to check differences between the layouts and shading conditions assumed in PVSyst and the actual on-site conditions.

Also, if you can use AR to overlay drawings and design information onto the site for verification, you are more likely to notice that layouts that appear acceptable on the drawings may actually be affected by obstacles or terrain. To improve the accuracy of power generation simulations, it is important not to rely solely on the figures in PVSyst, but to make assessments by combining on-site location information, photographs, point clouds, drawings, and measured data.

To correctly interpret PVSyst's energy output is not just a matter of reading the numbers displayed on the software screen. It involves checking whether the site conditions and the input parameters match, applying corrections as necessary, and evaluating them against measured values and on-site information. By taking a perspective that connects plant design, construction, O&M, and performance evaluation, a PVSyst report becomes a more practical basis for decision-making.

Summary

When reading energy production in PVSyst, it is important not to judge based solely on the final annual yield. Energy production is the result of calculations based on many assumptions—irradiance, system capacity, loss conditions, shading, temperature, PCS capacity, output limits, evaluation points, and so on. Therefore, judging right or wrong only by the size of the numbers risks overlooking the essence of the analysis.

The first thing to verify is the source and representativeness of the solar irradiance data. Because higher solar irradiance tends to result in higher power generation and lower solar irradiance tends to result in lower power generation, you should check for differences in meteorological conditions before comparing differences in power output.

Next, we will look not only at energy production but also at PR and Specific Production. Energy production is a figure that indicates the scale of a project's commercial viability, but by examining PR and energy production per unit of capacity, it becomes easier to compare system performance and the appropriateness of the design.

Checking loss items is also indispensable. By breaking down and examining how much each loss—soiling, wiring, temperature, mismatch, shading, inverters, transformers, auxiliary equipment, output limits, etc.—contributes, you can discern the reasons why generation is high or low.

The effects of shading must be checked not only on an annual basis but also seasonally, by time of day, using 3D models, and for electrical impacts. For temperature losses, it is important to interpret them in relation not only to outdoor temperature but also to installation and ventilation conditions.

PCS capacity and the handling of output limits also greatly affect how generation output is interpreted. If the DC/AC ratio, clipping, point-of-interconnection limits, power factor conditions, and so on are not checked, differences in generated output cannot be correctly interpreted.

When comparing with measured values, it is necessary to align conditions such as the period, solar irradiance, evaluation points, causes of stoppage, output control, and missing data. A comparison in which conditions are not aligned does not evaluate power plant performance; it simply becomes a difference in numbers.

The ability to read PVSyst's generation figures is not about memorizing the numbers in the report but about checking the context behind them. By sequentially reviewing the annual energy production, PR, Specific Production, Loss Diagram, monthly values, and input conditions, and comparing them with on-site conditions, you can use PVSyst's results as a practical basis for decision-making.