6 Ways to Interpret PVSyst Energy Yield Differences | For Case Comparisons

In the design of solar power plants, feasibility assessments, materials for financial institutions, EPC proposals, and O&M improvement studies, it is common to compare multiple cases of energy production simulations performed with PVSyst. For example, this includes changing the modules, altering PCS capacity, changing the tilt angle, slightly rotating the azimuth, tightening the racking layout, or varying conditions such as snow and soiling.

In such cases, if you simply conclude that "the case with higher annual energy production is good" and "the case with lower production is bad," you may make incorrect design decisions. This is because differences in PVSyst's energy production arise from the accumulation of multiple factors, such as irradiance, temperature, shading, IAM, mismatch, wiring losses, PCS losses, clipping, and grid-side losses.

When comparing cases, it is important not to look only at the final annual energy generation, but to break down and read at which stage the differences arise. If you can explain the reasons for the generation differences, it becomes easier to justify design changes, validate loss settings, ensure the consistency of estimation conditions, and explain differences with other companies' analyses.

In this article, we outline six perspectives to check when interpreting differences in energy production in PVSyst case comparisons. Rather than simply comparing results, we explain how to read them so you can practically explain why the differences occurred.

Table of Contents

• Don't judge based on annual generation alone

• First check whether the comparison conditions are aligned

• Look at the difference between solar irradiation and effective solar irradiation

• Examine the difference in losses up to the array output

• Examine the difference in losses after the PCS

• Narrow down causes using monthly and time-of-day differences

• When comparing cases, document the reasons for the differences in writing

• Connect PVSyst comparisons to on-site judgment

• Summary

Do not judge based solely on annual power generation

In a PVSyst case comparison, the first things you notice are representative values such as annual energy production, specific yield, and PR. These are very important indicators, but when comparing cases it is essential not to draw conclusions based solely on those representative values.

For example, if Case A’s annual power generation is 1,000 MWh and Case B’s annual power generation is 1,020 MWh, at first glance Case B appears to be 2% better. However, the significance changes greatly depending on where that 2% difference comes from.

Is the increase in energy production due to larger module capacity, a change in solar irradiance conditions, reduced shading losses, a larger PCS capacity reducing clipping, or merely a change in the wiring loss settings? Even if they all amount to the same 2% difference, the design implications are completely different.

Particular attention should be paid to cases where the baseline assumptions for the items being compared are not completely identical. Even slight differences in weather data, albedo, soiling, module model, PCS model, DC capacity, AC capacity, azimuth, tilt, shading settings, wiring losses, transformer losses, and so on will result in differences in the final energy output.

Therefore, when comparing cases in PVSyst, after confirming the difference in final energy production, always return to the Loss Diagram and the monthly table to identify which items are causing the difference.

Annual power generation is a conclusion, not a cause. When comparing cases, you need to start from the difference in annual power generation and trace its breakdown.

First, check whether the comparison conditions are the same

Before interpreting differences in power generation, the first thing to confirm is the consistency of the comparison conditions. If you compare loss items without checking this, the root-cause analysis may be skewed.

The first thing to look at is the capacity under consideration. Check whether the DC capacity is the same, the AC capacity is the same, or the DC/AC ratio is the same. If you simply compare annual generation for cases with different DC capacities, the case with the larger capacity will appear to generate more. In this situation, you need to compare generation per kWp or PR rather than the annual generation itself.

Next, confirm whether the meteorological data are the same. In PVSyst, even at the same site the solar irradiance and temperature can vary depending on the meteorological data source used, the year, and the correction method. If the solar irradiance differs, then even if you examine downstream loss items in detail, the comparison will not be valid because the input energy itself is different.

Terrain and shadow settings are also important. Differences in near-field shading, far-field shading, terrain shading, obstructions, racking spacing, and the number of module rows can change the effective solar irradiance. In particular, in low-tilt systems, east–west layouts, snow-prone regions, and mountainous areas, differences in shading and angle of incidence can have a large impact on annual energy generation.

Also, check whether settings such as soiling, albedo, snow, IAM, temperature coefficient, wiring losses, transformer losses, and auxiliary equipment losses are the same. Although each of these differences may seem small individually, when several accumulate they can produce a non-negligible difference in annual energy production.

A common mistake in case comparisons is mistaking a difference in power generation for one caused by design variations when it is actually due to differences in input conditions. For example, if you want to assess the effect of changing the racking layout but the soiling settings or wiring loss settings have also been altered, you cannot evaluate the effect of the layout change alone.

Therefore, in PVSyst case comparisons, clarify the purpose of the comparison before looking at differences in energy production. Decide whether you are comparing module changes, PCS capacity, tilt angle, or layout, and as a basic principle, keep conditions unrelated to that purpose as consistent as possible.

Examine the difference between solar radiation and effective solar radiation

After confirming the comparison conditions, the next thing to examine is the difference in solar irradiance. In PVSyst you can check the flows of horizontal-plane irradiance, tilted-plane irradiance, and effective irradiance. Differences upstream in these flows greatly affect the overall energy generation.

First, check whether solar irradiance values such as GlobHor and GlobInc are the same. GlobHor can be understood as horizontal-plane solar irradiance, while GlobInc is the solar irradiance incident on an inclined plane. Differences can arise at this stage when meteorological data, tilt angle, or azimuth change.

For example, when comparing a south-facing 15-degree case with a south-facing 25-degree case, the annual solar irradiation on the tilted surface changes. While solar gains in winter increase, gains in summer may also vary. When comparing east–west and south-facing layouts, not only does the annual total change, but the distribution of generation in the morning and evening also changes.

Next, check the effective irradiance that reflects shading and IAM. Even if the irradiance on the tilted surface is the same, the irradiance actually available to the modules will change if shading losses or IAM losses differ. In cases where racking spacing is narrowed, installed capacity may increase while shading losses also increase. In such cases, annual generation may increase, but energy yield per kWp and PR may decrease.

IAM is the loss caused by light incident at oblique angles. Its impact varies with tilt angle, azimuth, module surface characteristics, and morning/evening irradiance conditions. If IAM losses differ significantly between cases, differences in incident-angle conditions — rather than merely capacity or PCS differences — may be influencing the variation in energy production.

Also, in snowy regions the treatment of albedo and snow losses is important. While snow cover reduces power generation, under some conditions reflection from surrounding snow surfaces can increase solar irradiance capture. How this is configured in PVSyst can significantly alter winter energy production.

Checking solar irradiance and effective irradiance is the task of confirming the entry point of power generation differences. If the discrepancy is large at this stage, looking only at downstream PCS losses or wiring losses will not lead to the root cause.

Examine the difference in loss up to the array output

After checking the effective irradiance, next we look at the losses up to the array output. Here we check how much of the solar energy received by the modules can be extracted as DC power.

On the array side, important factors are temperature losses, module quality losses, LID, mismatch losses, DC wiring losses, and electrical losses due to shading. These directly affect the final power generation and are items that tend to show differences in case comparisons.

In particular, temperature losses vary depending on module type, mounting method, ventilation conditions, and ambient temperature. Even with the same solar irradiance, output decreases under conditions that raise module temperature. For roof-mounted, ground-mounted, bifacial modules, and low-tilt configurations, the way temperature conditions are considered changes, so care is needed when making comparisons.

When the module model is changed, temperature coefficients, nominal output, low-irradiance characteristics, bifaciality, and other parameters may change. In such cases, differences in energy yield are not simply due to capacity differences, but also arise from differences in module characteristics.

Mismatch losses are also easy to overlook. Losses vary depending on string configuration, how shading occurs, module variability, and the approach to MPPT partitioning. In particular, in layouts with complex terrain or multiple orientations, differences in string configuration can affect energy yield.

DC wiring losses are very important when comparing cases. In PVSyst, wiring losses can either be set as a fixed percentage or calculated from cable length and cross-sectional area. If the DC wiring loss settings differ between cases, part of the difference in energy production will be due to differences in settings rather than differences in design.

For example, if one case assumes DC wiring losses of 1.5% and another assumes 1.0%, that alone will cause a difference in annual energy production. If the PCS layout was changed so that the actual cable length became shorter, the difference is reasonable, but if it is simply that the settings are not consistent, the comparison is inappropriate.

When examining losses up to the array output, it is important to distinguish which losses will naturally change as a result of design modifications and which losses are merely differences in input conditions.

Examining the Loss Difference After PCS

After confirming the array output, we next look at the losses after the PCS. Here we check how much the DC power is reduced as it is converted to AC power and transmitted to the grid interconnection point.

Particularly important for PCS are inverter efficiency, MPPT range, voltage conditions, clipping due to oversizing, power factor settings, AC wiring losses, transformer losses, and auxiliary losses.

A common result in case comparisons is the difference in energy production caused by variations in the DC/AC ratio. When the PCS capacity is smaller than the DC capacity, the PCS reaches its output limit during periods of strong solar irradiance and clipping occurs. As a result, a portion of the power that could have been produced on the array side is discarded.

On the other hand, designing for a higher DC/AC ratio also has the advantage of securing annual power generation while keeping PCS capacity down. Therefore, rather than simply judging that clipping losses are bad, it is necessary to make a decision by considering equipment costs, grid capacity, electricity selling price, output control, and constructability.

In cases where PCS capacity has been changed, it is necessary to distinguish whether the increase in annual energy production is due to a reduction in clipping losses or to differences in PCS efficiency. Large PCS and distributed PCS can have different efficiency curves and may exhibit different behavior under low-load conditions.

Power factor settings also require attention. How power factor and output limits are treated in PVSyst changes the way losses appear. In particular, whether the PCS's rated output is regarded as active power or viewed in relation to apparent power can alter the interpretation of output limits and losses.

AC wiring losses and transformer losses are also items that tend to exhibit differences in settings when comparing cases. If the distance from the PCS to the power receiving and transforming equipment changes, the number of PCS units changes, or the transformer configuration changes, the actual losses will change as well. However, if only the loss settings differ while the design has not changed, the comparison conditions may not be consistent.

Don't forget auxiliary losses. Monitoring devices, communication equipment, PCS standby power, transformer no-load losses, air conditioning and snow-melting equipment — depending on how auxiliary power is accounted for, PR and the amount of electricity sold will change. In small-scale projects, auxiliary losses may appear relatively large.

Losses downstream of the PCS are related not only to design but also to operational conditions and grid conditions. When comparing cases, it is important to distinguish whether differences downstream of the PCS reflect the power plant’s actual performance or differences in electrical design conditions.

Narrow down causes by monthly and time-of-day differences

After checking the difference in annual generation and the Loss Diagram, next we look at the monthly differences. Viewing them by month makes it easier to determine whether the generation difference is due to seasonal factors or to a consistent loss throughout the year.

For example, if the difference is large only in winter, the effects of snow cover, low solar altitude, shading, albedo, and tilt angle can be considered. If the difference is large only in summer, temperature-related losses, PCS clipping, and restrictions during peak solar irradiance may be related.

When differences are large in spring or autumn, variations in solar altitude, angle of incidence, and azimuth may be responsible. In comparisons between east–west and south-facing layouts, not only the annual power generation but also the monthly and time-of-day generation patterns differ.

When examining month-to-month differences, check not only the absolute difference in generated energy but also the relative difference. Months with higher generation tend to show larger absolute differences, while months with lower generation tend to show smaller absolute differences. Therefore, in addition to looking at the monthly generation difference in kWh, check the percentage difference as well to more easily identify the cause.

If you want to look in more detail, check the differences by time of day. If differences occur in the morning and evening, azimuth, shading, IAM, and terrain shading may be involved. If differences occur around noon, PCS clipping or temperature losses may be involved.

If PVSyst's outputs can be reviewed as time-resolved data, comparing representative days or monthly averages makes it easier to explain which time periods show differences. In particular, when comparing PCS capacities or east–west layouts, examining time-of-day generation curves reveals differences that annual values alone do not show.

Monthly and time-of-day differences are a means to verify the causes identified in the Loss Diagram. For example, if the Loss Diagram shows an increase in clipping losses, check whether the differences are larger during daytime in summer. If shading losses are increasing, check whether the differences appear in winter or in the mornings and evenings.

In this way, by linking the annual figures, loss items, monthly differences, and time-of-day differences, the explanation for differences in power generation becomes more convincing.

When comparing cases, document the reasons for the differences.

PVSyst comparison results can be difficult to convey with numerical tables alone. In practice, it is important to organize the results so you can explain in writing the reasons for differences in energy production.

For example, simply stating that "Case B's annual power generation is 2.1% higher than Case A's" is not sufficient. By itself, this does not clarify why it is higher—whether it is the effect of a design change or due to differences in conditions.

In more practical terms, an explanation such as the following is necessary: "In Case B, the DC capacity is the same, but proximity shading losses decreased due to the increased spacing between racks, resulting in higher effective irradiance. Also, because the PCS capacity is the same, there is no large difference in clipping losses, and the primary cause of the difference in generation output is the reduction in shading losses."

Also, when comparing cases, it is important not to assume the difference is due to a single factor. Because differences in power output are often caused by multiple factors acting simultaneously, it is clearer to explain primary and secondary causes separately.

For example, a summary like: "The primary cause is the increase in irradiance on the inclined surface, with a secondary contribution from the reduction in shading losses during winter. On the other hand, temperature-related losses increase slightly in summer, and over the entire year these effects offset each other, leaving the difference in power generation at about 1%."

When written this way, it becomes not merely a statement that power output has increased, but a basis for design decisions.

When explaining to financial institutions, project owners, EPC contractors, O&M teams, or internal approval committees, clearly identifying the reasons for differences in generated output increases confidence in the analysis results. Conversely, if the reasons for the discrepancies cannot be explained, the PVSyst settings and comparison conditions may be called into question.

In case comparisons, it is important to always conclude by writing, "what changed, which losses changed as a result, and by how much the final power generation was affected."

Connecting PVSyst Comparisons to On-Site Decision Making

PVSyst case comparisons are not meant to end as mere desktop comparisons of energy production. Their purpose is to inform actual design, construction, operation and maintenance, and profitability assessments.

For example, widening the rack spacing reduces shading losses, but it may decrease the capacity that can be installed on the same site. Increasing PCS capacity reduces clipping, but it may increase equipment costs. Increasing the tilt angle may increase winter generation, but it may affect wind load, constructability, and racking cost.

In other words, a case that shows higher energy production in PVSyst is not necessarily the optimal option from a business perspective. When comparing cases, differences in energy production need to be evaluated together with differences in design, cost, constructability, maintainability, and risk.

Also, to reflect the PVSyst comparison results at the site, verifying the local conditions is essential. Even if drawings appear to show little shading, surrounding trees, slopes, fences, utility poles, snow accumulation, soiling, drainage conditions, and other factors can affect power generation.

In such on-site verifications, positioning that combines a smartphone with high-precision GNSS and AR-based drawing overlays can be useful. Using a system that leverages an iPhone and GNSS like LRTK to overlay drawings and design information while confirming position on site makes it easier to reconcile desk-based PVSyst conditions with actual site conditions.

For example, if you can confirm on site the racking positions, PCS positions, cable routes, surrounding obstructions, extent of site development, and inspection access paths, it becomes easier to determine whether the conditions assumed in PVSyst correspond to the actual construction conditions. It is also useful for clarifying whether differences in energy yield are due to desk-based assumptions or to on-site conditions.

PVSyst is a powerful tool for evaluating power generation, but input conditions must be based on actual site conditions. To improve the accuracy of case comparisons, it is important to have a perspective that links simulation results with on-site verification.

Summary

When interpreting PVSyst generation differences, it is important not to judge based solely on annual energy production. The final generation difference is the cumulative result of many factors, such as irradiance, effective irradiance, array losses, PCS losses, wiring losses, transformer losses, and auxiliary equipment losses.

When comparing cases, first confirm that the comparison conditions are consistent. If DC capacity, AC capacity, weather data, tilt angle, azimuth angle, shading settings, soiling, albedo, wiring losses, PCS conditions, etc. remain inconsistent, you will not be able to tell whether the differences are due to design or to configuration.

Next, look at the solar irradiance and the effective solar irradiance to confirm the entry point of the power generation difference. The effects of tilt angle, azimuth, shading, IAM, snow, albedo, and other factors tend to show up here.

After that, verify the losses up to the array output, examining temperature loss, mismatch, DC wiring loss, and differences in module characteristics. Additionally, as losses downstream of the PCS, check clipping, PCS efficiency, AC wiring loss, transformer loss, and auxiliary equipment loss.

If you can't determine the cause from annual values alone, examining monthly and time-of-day variations makes it easier to isolate winter shading, summer clipping, morning/evening IAM, temperature losses, and so on.

Ultimately, it is important to explain the difference in power generation not only with numbers but also in words. If you can organize which conditions changed, which losses changed, and how they affected the final power generation, you will have a comparison that is easy to use for design decisions and for explaining to customers.

PVSyst case comparisons are not simply judgments of which simulation result is better. They are analyses to interpret the reasons for differences in energy output and link them to design, construction, cost, operation, and site conditions for decision-making. By being able to explain why output increased, why it decreased, and why differences are small, PVSyst results become easier to use for practical decision-making.