7 perspectives to improve the accuracy of comparative analyses in the PVSyst manual

Table of Contents

• Key Concepts to Understand Before Using the PVSyst Manual for Comparative Evaluation

• Determine the purpose of the comparison first and establish the evaluation criteria.

• Align differences between meteorological data and site conditions when interpreting them

• Do not overlook the assumptions regarding azimuth, tilt, and array layout

• Compare capacity design and DC/AC ratio on the same footing

• Break down the loss settings and identify the cause of the differences

• Link close-range shadows, distant terrain, and 3D conditions to the results.

• Assess by combining power generation, PR, and monthly results

• Operational rules to improve the accuracy of comparative evaluations

• Summary

Concepts to keep in mind before using the PVSyst manual for comparative evaluation

Many people who want to improve the accuracy of comparative evaluations in the PVSyst manual will want to compare multiple patterns of solar power plant design proposals, equipment configurations, layout conditions, meteorological data, and loss assumptions, and be able to explain which option is reasonable. Simply running simulations and selecting the option with the highest annual energy output is insufficient as a practical comparative assessment. It is necessary to interpret why the differences arose, which input conditions influenced the results, and whether a given option is truly easy to adopt given the project’s constraints.

PVSyst handles many factors that affect energy production, such as weather data, installation angle, modules, PCS, string configuration, shading, losses, and grid-side conditions. Therefore, the more cases you compare, the harder it becomes to tell which conditions are the same and which are different. Even if you follow the manual while working, creating multiple cases while leaving the meanings of input fields ambiguous makes it impossible to determine whether differences in results are due to design changes or simply to omitted settings.

What matters in comparative evaluation is not being able to tweak every single parameter in detail, but being able to manage the assumptions behind the comparison. For example, if you only want to compare orientation, capacity, losses, weather data, shading conditions, and equipment conditions should, in principle, be kept the same. If you want to compare PCS capacity, align the number of modules, array layout, weather conditions, and loss settings, and then check the effects of the DC/AC ratio and clipping.

The value of using the PVSyst manual is not just in learning how to operate each settings screen. It lies in understanding which inputs affect which results, and in arranging comparison results into a form that can be explained. In this article, with reference to the PVSyst manual, we outline seven perspectives that are particularly important in practice for improving the accuracy of comparative evaluations.

Decide the purpose of the comparison first and fix the evaluation criteria

The first point to confirm is to be clear about what you want to compare. PVSyst allows you to create multiple variants and cases, so once you become a little familiar with the software you may be tempted to change azimuth, tilt, PCS, modules, meteorological data, loss conditions, and so on one after another. However, if you change several conditions at the same time, you won’t be able to determine which factor caused the differences in the results.

The basic rule for comparative evaluation is to limit the main variables you change at one time as much as possible. For example, when examining differences in tilt angle, other conditions should in principle be kept fixed. The same applies when examining differences in orientation. If you are comparing module types, since the number of modules that can be installed and the capacity may change, you need to check not only the simple annual energy output but also the energy yield per kW and area efficiency.

The important thing here is that the metrics you should look at change depending on the purpose of the comparison. The optimal option can differ when you want to maximize energy generation versus when you want to prioritize land-use efficiency. The conclusion of the comparison will vary depending on whether you emphasize generation per unit of installed capacity, total generation across the entire site, the cost-effectiveness of PCS oversizing, or stability that accounts for shading and terrain constraints.

When reading the PVSyst manual, start by organizing which screens and which result indicators you should look at according to the comparison objective of the project you are handling; this will help you understand faster. Rather than simply following the settings screens in order, it is important to work backwards from the comparison objective to confirm the necessary input items. This will reduce unnecessary case creation and minimize rework during the review.

Also, when sharing comparison tables internally, pay attention to how you name the cases. If you simply use "Case1" and "Case2", you won’t be able to tell later what was changed in each case. For example, use names that indicate the main differences—azimuth, tilt, PCS capacity, meteorological data, loss conditions, etc.—to avoid confusion when reviewing reports. Aligning case management in PVSyst with case management in your internal documents is also essential for improving the accuracy of comparative analysis.

Read meteorological data and differences in site conditions consistently

The second perspective is meteorological data and site conditions. In PVSyst comparisons, differences in meteorological data have a large impact on the results. If assumptions such as solar irradiance, ambient temperature, and wind speed change, the energy production will vary even with the same system configuration. Therefore, when you want to compare the merits of design proposals, if the meteorological data also differ you cannot disentangle the effects of design differences from those of meteorology.

What you should pay particular attention to are the location's latitude and longitude, elevation, time zone, the type of meteorological data used, the period, and the interpolation method. Even if you think you are using data from nearby areas, solar radiation and temperature trends can differ in mountainous, coastal, snowy, and urban areas. If there are multiple candidate sites for the project, you need to clearly distinguish whether you are comparing sites to examine differences in meteorological conditions or comparing design options at the same site.

When consulting the PVSyst manual, it is important to understand not only how to handle MET data and weather files but also which conditions are reflected in the simulation results. Placing a case where the weather data was changed alongside a case where the system configuration was changed on the same comparison axis can lead to incorrect conclusions. For example, if option A uses data with better solar irradiance conditions and option B uses different weather data, even if option A shows higher energy production, you cannot conclusively say that its system design is superior.

In comparative evaluations, it is easier to stay organized if you separate design comparisons that hold meteorological data constant from sensitivity analyses that vary the meteorological data. The former involves comparing azimuth, tilt, capacity, losses, and so on under the same meteorological assumptions. The latter examines how much the results change depending on the selection of meteorological data. Keeping these two separate is the first step to improving the accuracy of comparisons.

When looking at monthly results, the influence of meteorological data is also important. Even if annual energy production shows little difference, trends can differ between summer and winter. For projects with strong regional characteristics—such as snowfall, the rainy season, high temperatures in summer, or low solar irradiation in winter—it is necessary to check not only annual values but also monthly trends in generation and losses. When reading the results screen while referring to the PVSyst manual, cultivate the habit of checking monthly variations as well as annual values to increase the persuasiveness of comparison results.

Do not overlook assumptions about orientation, tilt, and array layout

The third perspective is orientation, tilt, and array layout. In solar power simulations, module orientation and angle directly determine the amount of solar radiation captured. When conducting comparative studies in PVSyst, even small changes in orientation or tilt can affect annual energy yield, monthly generation, shading patterns, array spacing, and on-site layout efficiency.

When comparing azimuth and tilt, simply searching for the angle that yields the highest energy output is insufficient. In real projects, there are various constraints such as site shape, site development conditions, racking specifications, wind loads, snow loads, maintenance access, relationships with neighboring properties, and the location of grid-connection equipment. Even a proposal that shows high energy output in simulations can be difficult to adopt if it has poor constructability or maintainability.

When using the PVSyst manual to check how to input azimuth and tilt, always confirm the scope to which those settings apply. Whether you are applying the same angle to the entire array or working with a configuration that has multiple sub-arrays or different faces will change how you interpret the results. On rooftop projects or sites with complex terrain, multiple azimuths and multiple tilts may coexist. In such cases, if you view the results with the same mindset as a simple single-face comparison, you may misidentify the cause of the differences.

In array layout, panel spacing and row spacing are also important. Narrowing the spacing can increase installed capacity, but it can increase shading from adjacent panels. Conversely, widening the spacing reduces shading effects but may reduce the capacity that can be installed on the same site. Therefore, what should be compared is not just annual energy generation, but the overall balance including installed capacity, generation per kW, site utilization efficiency, shading losses, and maintenance access.

When comparing orientation, tilt, and layout, you need to carefully record what is changing in each case. If you think you only changed the orientation but the layout or capacity has actually changed as well, you will not be able to correctly explain the differences in the results. In PVSyst comparisons, it is important to check not only the input values shown on the screen but also the assumptions output in the report, and to align them with internal documentation.

Comparing capacity design and the DC/AC ratio on a level playing field

The fourth perspective is capacity design and the DC/AC ratio. When conducting comparative studies in PVSyst, module capacity, PCS capacity, string configuration, MPPT allocation, and the degree of oversizing all have a major impact on energy generation and losses. In particular, when comparing a proposal to increase the number of modules on the same site, a proposal to change PCS capacity, and a proposal to adjust the DC/AC ratio, you cannot judge them solely by differences in annual energy production.

Increasing the DC-side capacity increases annual energy production under certain conditions. However, if the DC capacity becomes too large relative to the PCS capacity, output limiting and clipping are more likely to occur. In the PVSyst results, it is necessary to check where these losses are appearing. Even if energy production increases, if losses also increase substantially, whether it is truly advantageous in terms of economics and equipment utilization must be evaluated separately.

When comparing options, it is important to consider total generation, generation per unit of installed capacity, PR, and the breakdown of losses together. A large-capacity option may appear advantageous in terms of total generation, but it can be disadvantaged in efficiency per kW. Conversely, a reduced-capacity option may look smaller in total generation yet have fewer losses and allow for more stable operation. Reviewing the PVSyst manual and understanding the relationship between capacity-related inputs and resulting indicators will improve the quality of comparisons.

String design is another easy-to-overlook point. If the number of modules, the series count, the parallel count, or the assignment per MPPT are not appropriate, they will affect the voltage range, current conditions, mismatch, PCS input conditions, and other factors. Even when you duplicate a case and change only part of it, you need to verify that the string configuration is in the intended state. In particular, when changing the module type or PCS type, electrical conditions are not necessarily optimized automatically, so it is important to recheck while referring to the manual.

When comparing capacity designs, the project's objectives are also important. The desirable DC/AC ratio will vary depending on whether you want to maximize revenue from electricity sales, minimize initial investment, aim to maximize generation within the grid connection capacity, or assume peak shaving. PVSyst simulation results are a basis for decision-making, not the final conclusion itself. To improve the accuracy of comparative assessments, you need to organize technical results and business objectives separately and then link them together when making a judgment.

Decompose the loss settings to identify the cause of the differences

The fifth perspective is loss settings. One of the most important points to watch when comparing PVSyst cases is how loss conditions are handled. Temperature loss, wiring loss, mismatch, soiling, degradation, IAM, low-irradiance characteristics, PCS losses, transformer losses, grid-side losses and other factors all affect energy yield. If these settings differ between cases, the comparison results can vary significantly.

A common problem with loss settings is using the standard/default value in some cases and a project-specific value in others. Also, copying settings from past projects and reusing them for new ones can leave loss conditions that do not reflect the actual situation. Even when working with the PVSyst manual open, entering values without understanding the meaning of each loss item makes it difficult to explain differences in the results.

When comparing options, it is important not to view losses as a single total value, but to break them down and examine each item. If there is a case of low annual power generation, isolate whether it is caused by poor insolation conditions, large temperature losses, significant shading losses, substantial PCS-side limitations, or large wiring losses. By reviewing the loss diagram and the breakdown in the report, you can identify which factors are influencing the results.

Especially when there are multiple alternatives to compare, checking differences in loss settings is essential. If you want to compare only the module layout between Plan A and Plan B, but the soiling loss or wiring loss values differ, the comparison of layouts will be inaccurate. Conversely, if you want to compare differences in wiring routes, changing the wiring loss itself becomes an item under consideration. In this way, it is important to separate beforehand which losses can be changed and which should be kept fixed.

Loss settings are an area that is often questioned in internal or client presentations. Being able to explain why a particular value was adopted — whether it is a standard value, based on project conditions, chosen conservatively, or derived from actual measurements or design documents — increases the credibility of the comparison results. When referring to the PVSyst manual, you should not only check the location of the input fields but also make a point of understanding the rationale behind how each loss item is set.

Link near-field shadows, distant terrain, and 3D conditions to the results

The sixth perspective is shading and 3D conditions. In comparative evaluations of solar power generation, nearby shading, distant topography, surrounding structures, and mutual shading between arrays all affect energy output. PVSyst can handle 3D scenes and shading settings, so if you configure conditions carefully you can identify effects that are difficult to see in simple planar analyses. On the other hand, comparing results with insufficient input of shading conditions may lead to overlooking real-world risks.

In near-field shading, array spacing, racking height, surrounding buildings, fences, utility poles, trees, slopes, and equipment can all have an impact. The effect of shadows can be especially large in mornings and evenings and during winter when the solar altitude is low. When viewing PVSyst results, it is important to pay attention not only to annual shading losses but also to monthly and time-of-day trends. Shadows that appear minor in annual figures can still affect generation and profitability in particular months.

Distant terrain is another element that is easy to overlook. For projects near mountains, hilly areas, or valley landforms, horizon-direction shading can affect solar gain in the morning and evening. If you compare cases that do and do not take distant terrain into account on an equal footing, you may misinterpret the differences in results. It is necessary to clarify whether the comparison is intended to compare design proposals or to assess sensitivity to terrain conditions.

When using a 3D scene, attention must also be paid to the accuracy of the model. If the dimensions, position, height, or orientation of objects that cast shadows deviate from reality, the resulting shading losses will also be off. In particular, when setting obstacles based on on-site surveys or drawing/plan information, alignment of coordinates and heights is important. To improve the accuracy of comparative assessments, you need to check not only the operation of PVSyst but also the quality of the source on-site data and drawing/plan information.

When comparing shading conditions, it is important not simply to compare shaded and unshaded cases, but to understand which shadows are having what effect and to what extent. Depending on whether inter-array shading, surrounding buildings, or terrain is the main cause, the mitigation measures will differ. To translate findings into concrete design decisions — for example, widening array spacing, changing the layout, rotating module orientation, adjusting the installation area, or excluding heavily shaded sections — you need to read the breakdown of shadows in connection with the results.

Determine by combining power generation, PR, and monthly results

The seventh viewpoint is how to interpret the result indicators. When comparing in PVSyst, people tend to judge based only on annual energy yield, but that alone is not sufficient. It is only by looking at annual energy yield, Specific Yield, PR, monthly generation, the loss diagram, output limitations, shading losses, temperature losses, and so on in combination that you can understand the meaning of the comparison results.

Annual power generation is an easy-to-understand indicator for grasping the overall scale of a project. However, when comparing multiple options with different installed capacities, be careful with simple comparisons, because options with larger capacity tend to have greater annual generation. In such cases, also check generation per unit of installed capacity and PR. Even if Option A is advantageous in total generation, Option B may be superior in efficiency per unit of capacity.

PR is an important metric for assessing how efficiently an installation converts incident solar irradiance into electricity. However, maximizing PR alone is not necessarily desirable. Reducing capacity or using only areas with less shading can make PR appear higher, but these choices can be disadvantageous for the project’s total energy generation and profitability. PR is a measure of efficiency and should not be used alone to determine the overall optimality of a project.

Monthly results are highly effective for improving the accuracy of comparative evaluations. Even two options that show little difference in annual values can diverge when viewed by month—one may perform better in summer, another in winter, or one may be more sensitive to morning and evening effects. In particular, for projects aimed at end users, those paired with battery storage, peak-shifting measures, or cases that consider seasonal profitability, monthly or time-of-day trends can be more important than annual totals.

The loss diagram is essential for explaining the results. Because it lets you visually trace why energy production has decreased and how much is lost at each stage, it makes it easier to identify directions for design improvements. When using the PVSyst manual, it is important not only to read the figures on the results screen but also to understand the flow of losses. When you find a difference in energy production, check whether the cause lies in the input conditions, equipment parameters, shading conditions, or loss settings.

When explaining comparison results to internal and external parties, it is important to specify not only the conclusion but also the metrics used for the judgment. Rather than just saying "Plan A has a larger annual energy production," explaining it in a way such as "Plan A is advantageous in total energy production but has large shading losses, while Plan B has more stable generation per unit capacity and PR" conveys the depth of the analysis. Because PVSyst's results contain many numerical values, rather than listing everything, it is necessary to select the important metrics according to the purpose of the comparison.

Operational rules to improve the accuracy of comparative evaluations

When conducting comparative evaluations while consulting the PVSyst manual, you need rules not only for understanding each setting but also for how to proceed with the work. Much of what reduces the accuracy of comparisons stems not from shortcomings in the software’s functionality but from unclear case management, recording of input conditions, change histories, and interpretation of results.

First, before creating comparison cases, clearly define the fixed conditions and the changed conditions. Fixed conditions are those you keep the same because they are not the subject of comparison. Changed conditions are the ones you intentionally alter in this analysis. If you carry out the work without this distinction, you may not be able to tell later what you were comparing when you look at the results. Even when duplicating cases in PVSyst, it is important to check that no unnecessary configuration differences remain after duplication.

Next, align the case name with the document name. If the names used within PVSyst, the names of output reports, and the names of internal comparison documents are all different, it will cause confusion during review. Make them in a format that clearly shows the project name, date, purpose of the study, and the main changed conditions so they are easier to refer back to later. Especially when multiple people are working, simply standardizing the naming rules can greatly improve the quality of comparative evaluations.

Recording the input conditions is also important. Meteorological data, azimuth, tilt, capacity, PCS, strings, loss settings, shading conditions, terrain conditions, and other key assumptions should be recorded not only in the report but also in a separate summary document for comparison. PVSyst reports generate a large amount of information, but in internal reviews and client explanations you will often need to extract and present only the information required for comparisons. Therefore, organizing the assumptions at the time of work helps streamline later processes.

Deciding review criteria in advance is also effective. For example, by having checkpoints such as whether the meteorological data are appropriate, whether the capacity conditions are consistent, whether there are any unnatural differences in the loss settings, whether the shading conditions reflect the on-site conditions, and whether the result indicators are suitable for comparison purposes, it becomes easier to detect omissions in the settings. The PVSyst manual can be used not only to verify operations but also to organize these review criteria.

Also, in comparative evaluations, before deciding on a single "best option," it is important to be able to explain "why the other options are not being adopted." For example, Plan A has a high energy output but large shading losses, Plan B has good efficiency but insufficient total capacity, and Plan C has high constructability but weak winter energy output; by organizing the strengths and weaknesses of each option, the conclusion becomes more persuasive. The practical objective is to use PVSyst results not merely as a numerical comparison but to inform design decisions.

Summary

To improve the accuracy of comparisons in the PVSyst manual, merely memorizing the operating procedures is not enough. You must clarify the purpose of the comparison and organize meteorological data, site conditions, orientation, tilt, capacity, DC/AC ratio, loss settings, shading conditions, and result indicators on the same footing. Because simulation results are determined by the accumulation of input conditions, comparing numbers alone while the assumptions remain ambiguous will not lead to a correct judgment.

What's particularly important is to narrow down the conditions you change at one time. If you want to compare orientation, fix everything except the orientation; if you want to compare PCS capacity, keep all conditions other than the capacity design the same; if you want to compare shading conditions, align all assumptions except the shading. By sticking to this basic rule, it becomes easier to explain the differences in the results. Conversely, if you change too many conditions at once, it becomes unclear what is causing the difference in power generation.

Also, it is important not to rely solely on annual energy production, but to consider a combination of PR, energy generation per unit of capacity, monthly results, loss diagrams, shading losses, and output limitations. A proposal with higher total energy production is not necessarily the optimal one; decisions need to be made taking into account efficiency, stability, constructability, maintainability, and alignment with project objectives. The values from PVSyst are material to support design decisions and do not automatically determine the final conclusion.

To improve the accuracy of comparative evaluations, it is important to organize case names, input conditions, change history, output reports, and internal documents so that the rationale for decisions can be traced later. By using the PVSyst manual to understand which settings affect which results, and by reading the results with evaluation criteria suited to the purpose of the comparison, you can produce simulations that are useful in practice.