【How to Improve Design Accuracy with PVSyst｜8 Practical Tips for Practitioners】

Table of Contents

• What Increasing Design Accuracy in PVSyst Means

• Tip 1: Organize input conditions to match design drawings and on-site conditions

• Tip 2: Suppress annual generation variability by selecting the right meteorological data

• Tip 3: Input azimuth and tilt correctly based on local reference

• Tip 4: Verify array configuration and string design based on electrical conditions

• Tip 5: Assess shading effects from both nearby obstructions and terrain

• Tip 6: Set loss coefficients with justification rather than entering uniform values

• Tip 7: Establish a method for viewing result screens to detect anomalies quickly

• Tip 8: Cross-check with on-site survey data to further improve design accuracy

• Operational tips for mastering PVSyst in practice

• Summary: Simulation accuracy is determined by the quality of on-site information

What does it mean to increase design accuracy in PVSyst?

When people talk about improving design accuracy in PVSyst, some imagine simply making the energy generation numbers look larger or fine-tuning calculation results. However, what matters in practice is to make site conditions, design conditions, electrical conditions, and loss assumptions as close to reality as possible, producing simulation results that can be explained afterwards. In photovoltaic system design, even with the same site area, the same panel capacity, and the same azimuth, annual energy production and loss rates vary depending on the weather data, tilt angle, string configuration, shading settings, temperature conditions, wiring losses, soiling losses, and how terrain conditions are treated. In other words, the difference in using PVSyst lies less in the software operations themselves and more in organizing the conditions before input and in how the output results are interpreted.

What practitioners should be especially careful about is not accepting PVSyst results as "correct because they are calculation results." PVSyst runs simulations according to the input assumptions, so if those assumptions are ambiguous, the output results will be ambiguous as well. Conversely, if site survey data, equipment specifications, racking conditions, electrical design, construction constraints, surrounding shading objects, and maintenance/operation conditions are carefully reflected, the outputs become useful documents not only for energy yield prediction but also for design comparison, verification of equipment capacity, evaluation of shading losses, consideration of the oversizing ratio, and business feasibility assessment.

Many practitioners who search for "how to use PVSyst" are looking not just for basic operations but also for how to decide input values, where to check when results look wrong, and what to review to improve design accuracy. This article explains 8 practical tips that are especially effective in the field for improving design accuracy in PVSyst, organized along the design workflow. Rather than a mere operations manual, it organizes how to think about input conditions, commonly overlooked checkpoints, and how to connect with on-site data.

Tip 1: Organize input conditions to match the design drawings and site conditions

The first tip for improving accuracy in PVSyst is to organize the input conditions before opening the software. In a power generation simulation you enter many items such as meteorological data, installation site, azimuth, tilt angle, panel specifications, PCS specifications, string configuration, obstructions, and loss conditions. If you input these haphazardly, when you review the conditions later you will not be able to tell which figures came from drawings, which are assumptions, and which have been confirmed on site. To improve design accuracy, it is important to first organize the sources of the assumptions and manage assumed values and finalized values separately.

For example, the reliability of a site's latitude and longitude varies depending on whether they were obtained from project planning documents or from on-site surveys. The panel layout area is also treated differently depending on whether the drawings are at a conceptual layout stage or are detailed drawings created with construction in mind. Racking height and row spacing may also be changed—even if they are feasible on the drawings—because of the actual graded ground surface, drainage slopes, embankments, maintenance access routes, fences, and relationships with existing structures. If the assumptions entered into PVSyst remain based on outdated drawings, the simulation may produce neat results but the evaluation will be inconsistent with actual construction conditions.

In practice, it is effective to first create a list of design conditions. Confirm each item one by one: installation location, system capacity, number of modules, PCS capacity, DC/AC ratio, azimuth angle, tilt angle, mounting method, minimum ground clearance, row spacing, obstructions, treatment of terrain, loss coefficients, and types of meteorological data. Then, by categorizing the status of each item as finalized, under design, provisionally placed, or unconfirmed, you can more easily improve the accuracy of simulations in stages. Using assumed values in initial studies is acceptable, but as you approach detailed design and submission materials, you should reduce assumptions and replace them with substantiated values.

Also, including condition-identifying information in PVSyst project names and variant names makes later comparisons easier. For example, if you manage separately the cases with different tilt angles, different PCS capacities, more detailed shading settings, or more conservative loss assumptions, it becomes easier to track which condition changes affected the energy yield. Improving design accuracy is not about creating a perfect model in one go. It is about organizing, comparing, and validating conditions while iteratively moving toward a model that more closely reflects reality.

Tip 2: How to choose weather data to minimize variation in annual power generation

PVSyst's design accuracy is strongly influenced by meteorological data. In solar power generation simulations, meteorological conditions such as solar irradiance, ambient temperature, and wind speed are directly linked to energy output. No matter how finely you refine the panel layout and electrical design, if the selection of meteorological data is inappropriate, estimates of annual energy production can differ significantly. In practice, it is important not only to use data from the nearest location but also to verify how well that data reflects the climatic characteristics of the installation site.

Especially in mountainous areas, coastal areas, snowy regions, basins, industrial zones, and areas with large elevation differences, nearby observation points may differ from the actual site conditions. Even if the distance is short, a large difference in elevation can change temperature and cloud formation, and along the coast there are situations where you need to consider the effects of fog, salt damage, humidity, and wind. Inland, high summer temperatures can cause significant thermal losses, and in snowy regions you also need to consider winter solar radiation, generation stoppages due to snow cover, soiling, and post‑snowmelt conditions.

When selecting meteorological data in PVSyst, it is important to look not only at annual solar irradiation but also at monthly trends. Even if annual values are similar, datasets that are stronger in summer versus those stronger in winter will change generation patterns and the assessment of self-consumption. In particular, for self-consumption projects, not only annual generation but also monthly and hourly generation are important. Depending on the project objectives, the points to check differ—for example, whether the generation peak coincides with periods of high load, and whether the generation required in winter can be secured.

In practice, it is also effective to compare multiple meteorological datasets. By comparing standard data, somewhat conservative data, and data that reflect local observational tendencies, you can grasp the range of power generation. Adopting only a single result does not reveal the sensitivity to small changes in conditions. Comparing multiple scenarios makes it easier to explain expected values and risks at the proposal stage, and to make conservative decisions at the detailed design stage.

Also, when you change weather data, it is important to fix the other conditions for comparison. If you change the weather data, tilt angle, loss coefficients, and shading settings at the same time, you will not be able to identify the cause of differences in power generation. To improve design accuracy, you should change one condition at a time and check how much it affected the results. PVSyst is well suited for comparing multiple variants, so saving cases with different weather data can be useful for later presentation materials.

Tip 3: Enter azimuth and tilt angles correctly using the on-site reference

Azimuth and tilt angles are aspects that users most commonly stumble over when first learning how to use PVSyst. In photovoltaic systems, the amount of solar radiation received depends on which direction the panels face and the angle at which they are installed. Particularly for typical fixed installations in Japan, there is a tendency to achieve higher annual energy production the closer the panels are to facing south; however, in practice constraints such as site shape, roof shape, racking layout, roads, neighboring property boundaries, maintenance access routes, and grading slopes mean you cannot always achieve ideal azimuth and tilt. That is why you need to enter values into PVSyst that reflect the actual installation conditions rather than ideal values.

When entering azimuths, be careful about how north on the drawing and north on site are treated. If you input data without confirming whether the orientation shown on the drawing is referenced to true north, magnetic north, or merely a convenient drafting direction, the azimuth can be offset. Small deviations may have only a limited effect on energy generation, but for layouts close to an east–west orientation or projects with strong shading effects, an azimuth error will affect the results. In particular, for multi-plane roofs, corrugated metal roofs, or ground-mounted projects on complex terrain, each surface can have a different azimuth, so summarizing with a single representative value reduces accuracy.

The same applies to the tilt angle. In some cases you can simply enter the racking tilt angle from the design drawings, but for rooftop installations you need to check the relationship between the roof pitch and the racking angle. For ground-mounted projects, in addition to the racking’s tilt angle, the slope of the prepared surface and differences in ground elevation can affect how shadows appear. Whether you treat the site as flat in PVSyst or reflect the terrain and per-row height differences will change the results for near shading. Especially on sloped terrain, even if the row spacing is the same, the way shadows fall changes depending on the relationship between the upper and lower rows, so a simple planar model may diverge from reality.

To improve design accuracy, it is important to determine the azimuth and tilt angles not as "ideal values to maximize energy yield" but as "values that are actually constructible." In the initial study, compare multiple angle proposals, and in the detailed design phase, finalize the values to match the construction drawings and on-site survey results. Furthermore, verifying the difference in energy yield when the azimuth or tilt angles are slightly changed makes decision-making easier when making design changes. For example, if you lower the tilt angle to prioritize constructability, wind loads, or maintainability, comparing the extent of the energy yield reduction in PVSyst will provide a basis for the design decision.

Tip 4: Verify array configuration and string design based on electrical conditions

To improve design accuracy in PVSyst, it is essential not only to set the layout conditions but also to correctly enter the array configuration and string design. If the number of modules, number in series, number in parallel, PCS capacity, allocation of MPPT units, DC/AC ratio, operating voltage range, and so on are not appropriate, the simulation results will not match the actual plant behavior. In practice especially, it is necessary to verify that the number of modules shown on the layout and the number of electrical connections in the electrical design exactly match.

In string design, you first check the module's open-circuit voltage, operating voltage, temperature coefficient, the PCS input voltage range, maximum input current, and the number of MPPTs. Because open-circuit voltage rises at low temperatures and operating voltage falls at high temperatures, it is necessary to verify that they remain within the PCS's allowable range throughout the year. If PVSyst issues a warning, do not simply adjust numbers to clear the warning; instead, check the cause against the actual equipment specifications. It is important to determine whether the number of modules in series is too high or too low, whether the combination with the PCS capacity is inappropriate, or whether the temperature condition settings are too conservative.

Setting the DC/AC ratio also affects design accuracy. When oversizing is implemented, the PCS may impose output limits during peak periods. PVSyst can check clipping losses caused by oversizing, but their assessment depends on module capacity, PCS capacity, irradiance conditions, temperature conditions, azimuth, and tilt angle. Simply increasing DC capacity does not necessarily increase annual energy production; the optimal ratio varies with the installation environment and electricity usage conditions. For self-consumption systems, consider the relationship with daytime demand patterns, while for feed-in systems you must also account for grid interconnection conditions and the possibility of output control or curtailment.

When there are multiple array surfaces, it is also important to organize the azimuth, tilt, shading conditions, and string assignments for each surface. Grouping modules with different azimuths into the same input circuit can increase losses due to differences in generation characteristics. If you simplify them in PVSyst as a single representative surface, you need to understand how that simplification will affect the results. Simplification may be sufficient for small projects, but for large projects or complex roof projects, modeling each surface separately yields higher design accuracy.

To improve the accuracy of electrical design, it is important to cross-check PVSyst input values with the single-line diagram, equipment specifications, layout drawings, and construction drawings. If the simulation engineer and the electrical design engineer are different people, changes to the number of modules or the number of PCS units may not be reflected in PVSyst. When a design change occurs, establishing a process to update layout, electrical, and simulation simultaneously will prevent inconsistencies in downstream processes.

Tip 5: Check the impact of shadows from both near-field shading and terrain

A setting that often causes practical differences in PVSyst is the shading configuration. In solar power generation, surrounding buildings, trees, utility poles, signs, chimneys, tower-like structures, fences, embankments, and adjacent rows of panels create shadows. The effects of shading are not limited to a simple reduction in solar irradiance but also lead to non-uniform generation at the string level and electrical losses. Therefore, to improve design accuracy in PVSyst, it is important not to roughly ignore shading but to set it at an appropriate level of granularity according to the project's risk.

In shadow assessment, you first consider distant shading and near shading separately. Distant shading refers to elements such as mountains or obstacles on the horizon that affect conditions when the solar altitude is low. In contrast, near shading consists of items close to the installation, such as buildings or adjacent rows, that cast direct shadows on the generating surface depending on the season and time of day. In particular, near shading can have a major impact on energy production when part of a panel is shaded. By modeling near shading in PVSyst, you can evaluate shading losses in a way that more closely reflects reality.

In ground-mounted projects, the shadows between rows of panels are also important. If the distance between rows is narrow, at the low solar altitudes in winter the shadow from the front row falls on the rear row. Tightening row spacing to increase installed capacity improves land-use efficiency but can increase shading losses. By comparing row spacing, racking height, and tilt angle in PVSyst, you can confirm the balance between increased energy production from higher capacity and shading losses. In practice, it is important not to look only at total capacity but to make a comprehensive judgment based on energy generation per unit capacity, shading losses, maintenance access aisles, and constructability.

Also, the effects of topography cannot be overlooked. If there are elevation differences on the site, the heights of the panel surfaces will vary by row, and the way shadows appear will differ from flat ground. Pre-development topographic data, planned post-development elevations, slopes, retaining walls, and drainage structures can affect layout and shading. Especially for projects in mountainous or sloped areas, it is difficult to assess shading risk from plan views alone, so on-site inspections that include height information are effective.

When setting shadows, it's important to note that making the model more detailed does not necessarily increase accuracy. If the entered positions or heights of obstructions are inaccurate, even a detailed model will produce incorrect results. What matters is prioritizing and accurately entering obstructions that are likely to affect power generation. For example, buildings close to the installation, obstacles on the south side whose shadows extend in winter, trees that cast shadows for long periods, and rack height and row spacing that affect inter-row shading are high-priority items. Conversely, over-modeling distant low obstacles that have little effect offers limited benefit relative to the time spent.

Tip 6: Set loss coefficients based on evidence rather than entering them uniformly

An important factor that determines PVSyst’s design accuracy is the various loss coefficients. In photovoltaic installations, energy yield is reduced by many elements such as temperature losses, wiring losses, mismatch losses, soiling losses, reflection losses, equipment conversion losses, degradation, and downtime rate. If these are not entered appropriately, the energy yield will be over- or underestimated. In practice, it is important not to input loss coefficients uniformly based only on past project conventions, but to set them with supporting justification tailored to the project’s conditions.

Temperature losses are strongly affected by the installation configuration. When panels are mounted flush to the roof, rear-side ventilation is poor and module temperatures tend to rise. Conversely, for ground-mounted systems where rear-side ventilation is ensured, temperature increases may be relatively suppressed. When setting temperature conditions in PVSyst, do not simply use the default initial values; check the installation type, racking height, roof material, ventilation conditions, and the surrounding environment. In high-temperature regions and rooftop projects, pay particular attention because temperature losses can have a large impact on annual energy yield.

Wiring losses depend on cable length, cross-sectional area, current, voltage, and wiring route. In preliminary design, approximate values may be used, but in detailed design it is necessary to make them closer to the actual wiring plan. In projects with long distances to the PCS, distributed rooftop projects, or large-scale ground-mounted projects, differences in wiring losses may become significant and cannot be ignored. Verify that the wiring losses in PVSyst match the actual cable design, and reconcile them with the electrical design calculations as necessary.

Soiling loss is another parameter that should be adjusted for each project. Near agricultural land, along unpaved roads, around factories, in coastal areas, and in regions with low rainfall, soiling can have a greater impact. Conversely, where regular cleaning is planned or natural washing by rain can be expected, the handling of losses will differ. However, if soiling loss is set excessively low, the discrepancy between predicted and actual operational generation can easily increase. At the proposal stage, it is important to consider this together with the maintenance plan.

Mismatch losses are influenced by module-to-module variability, differences in aging degradation, how shadows fall, uneven soiling, temperature differences, and other factors. In projects where partial shading is likely, you should check not only the simple mismatch losses but also the electrical effects caused by shading. In PVSyst, results vary depending on the shading settings and how electrical effects are handled, so for projects with significant shading it is important to review shading losses and mismatch losses separately.

The basic rule when setting loss coefficients is to document the basis for the numbers. Record why a value was chosen, which sources or design conditions it is based on, and whether it is an assumed or a definitive value, so that later review and explanation are easier. The reliability of simulation results is assessed not only by the final power generation figures but also by the validity of the loss settings.

Tip 7: Decide how to view the results screen to quickly spot abnormal values

When reviewing PVSyst results, it's important not to stop at checking only the annual energy production. To improve design accuracy, verify how much each type of loss is occurring, whether there are any unnatural biases in the monthly energy production, whether the system performance ratio is reasonable, and whether clipping or shading losses are larger than expected. Agreeing on a standard way to read the results screen within your company or team will help you spot anomalous values more quickly.

The first things to check are the annual energy production and the system performance ratio. If the annual production is higher than expected, loss assumptions may be too lenient, shading may be unconfigured, the meteorological data may be optimistic, or the capacity input may be incorrect. Conversely, if the production is too low, shading losses may be large, the azimuth or tilt may be incorrect, there may be issues with PCS capacity or string configuration, the meteorological data may be inappropriate, or loss factors may be double-counted. When the figures differ from expectations, it is important not to immediately try to correct only the final result, but to check the breakdown of losses step by step.

Monthly generation is also important. In solar power systems, output varies by season, but if there are unnatural monthly patterns for the region or installation conditions, you should review the meteorological data, azimuth, tilt angle, and shading settings. For example, if winter generation is extremely low, inter-row shading or surrounding obstruction settings may be affecting it. If summer generation underperforms expectations, temperature losses or PCS output limits may be involved. Examining monthly trends can reveal problems that annual values alone do not show.

Checking the loss diagram is also indispensable. In PVSyst, you can review the flow of losses from solar irradiance to the final output. If a particular loss is too large there, verify whether its settings are appropriate. By examining shading losses, temperature losses, wiring losses, mismatch losses, PCS losses, clipping losses, and so on in sequence, it becomes easier to isolate the causes of reduced energy production. Especially after design changes, comparing which losses have increased or decreased allows you to quantify the impact of those changes.

Also, PVSyst results should not be presented in isolation; it is important to compile them into a form that can be explained as design documentation. In proposals and internal reviews, showing not only the annual energy production but also the key input conditions, assumed losses, treatment of shading, assumptions about meteorological data, and design considerations will increase the credibility of the results. In practice, being able to explain those results to stakeholders is more important than merely producing the calculation outputs.

Tip 8: Improve design accuracy further by cross-checking with on-site survey data

Verifying against on-site survey data is effective for further improving design accuracy in PVSyst. In solar power plant design, site boundaries on drawings, graded elevations, existing structures, roads, fences, slopes, trees, and building locations can be offset from the actual site. Especially when designing based on old drawings or schematic maps, on-site checks can reveal unexpected elevation differences or obstacles. If the model in PVSyst does not match the site, the accuracy of shading assessment and layout planning will also decline.

Items to verify during an on-site survey include the location of the installation area, elevation/height, slope/gradient, surrounding obstructions, existing equipment, boundary conditions, maintenance/access paths, drainage routes, and so on. Even if a solar panel layout fits on the plan, it may need to be adjusted in practice due to elevation differences, slopes, drainage ditches, or inspection/maintenance walkways. Also, knowing the heights, distances, and relative positions of nearby buildings and trees allows shading models to be made more realistic; in particular, at the low solar altitudes in winter, obstacles located some distance away can still have an impact, so on-site verification is especially valuable.

Using high-precision positional information, you can organize the orientation and positions of obstructions entered into PVSyst more accurately. For example, if you measure on site the coordinates of the panel layout area, the row direction, the reference lines of the racking, and the positions of surrounding structures, you can check discrepancies between the design drawings and the site. By utilizing point cloud data and coordinate-tagged photos, it becomes easier to share site conditions among stakeholders and to reduce misunderstandings between design engineers, construction personnel, and power generation simulation personnel.

Also, field survey data are useful not only during the design stage but also for post-construction verification. By checking the actual panel azimuth, tilt, racking positions, and relationships with obstructions after construction, you can compare PVSyst’s design model with the as-built condition. When comparing design simulations with operational data, these surveys make it easier to confirm whether the installation matches the design, facilitating analysis of the causes of any differences in energy production. Even if output is lower than expected, they provide clues to distinguish whether the cause is different meteorological conditions, shading, soiling, or errors in installation positioning.

The accuracy of PVSyst is not something that can be achieved by software settings alone. The reliability of a design depends greatly on how precisely site location and elevation information are reflected. In particular, for projects with complex terrain, many surrounding obstructions, multiple roof surfaces, or intricate existing installations, coordination with on-site surveying is the most direct way to improve design accuracy.

Operational Tips for Mastering PVSyst in Practical Work

To master PVSyst in practice, it is important not only to memorize individual settings but also to understand how to operate it within the overall design workflow. The required level of accuracy differs for initial studies, basic design, detailed design, pre-construction checks, and post-construction validation. In initial studies, the goal is often to compare estimates, so it is more important to quickly compare multiple options than to refine every detail. On the other hand, for detailed design and submission documents, justification for input parameters, loss settings, handling of shading, and consistency with equipment specifications are required.

First, clarify the purpose of the PVSyst model for each design stage. The preliminary study model is used to compare the site's generation potential and proposed capacity options. The basic design model is used to compare azimuth, tilt, PCS capacity, array configuration, and shading effects, and to determine the design direction. The detailed design model reflects finalized equipment specifications, layout drawings, electrical design, and loss conditions, and is used as an explainable energy yield forecast. Reusing the same model without differentiating purposes carries the risk that outdated assumptions remain and are used in detailed documentation.

Next, it is important to manage the change history. In solar PV design, the number of panels, PCS capacity, layout, inter-row spacing, racking height, loss coefficients, and so on often change midway. If you overwrite the PVSyst model every time a change is made, you will lose the ability to compare before and after. By separating variants, giving them names that indicate the changes, and noting the key conditions, it becomes easier later to trace why the energy generation turned out the way it did.

Furthermore, establishing checklist items for internal reviews will stabilize quality. If you define a process to verify the installation site, meteorological data, azimuth, tilt angle, number of modules, PCS capacity, DC/AC ratio, string configuration, shading settings, temperature conditions, wiring losses, soiling losses, mismatch losses, annual generation, monthly generation, and the main breakdown of losses, you can reduce variation between personnel. Especially when multiple people handle a project, it is important not to leave the use of PVSyst to individuals.

In practice, it is necessary to make a habit of viewing PVSyst results together with design drawings and on-site information. If you look only at PVSyst, something may appear to work on the screen but in reality be a layout that cannot be constructed on site or a design that is difficult to maintain. Conversely, if you understand the on-site constraints, it becomes easier to explain why differences in simulation results occurred. When using PVSyst in practice, it is important to judge not only power generation but also constructability, maintainability, safety, and commercial viability.

Summary: Simulation accuracy is determined by the quality of on-site information

To improve design accuracy in PVSyst, rather than just memorizing special procedures, you need to be conscious of aligning the input conditions with the site and the design. By checking each item—installation site, meteorological data, azimuth, tilt angle, array configuration, string design, shading settings, loss coefficients, and how to read the results—the reliability of the energy yield simulation can be significantly affected. In practical work especially, it is important not to judge solely by the annual energy yield figure, but to be able to explain which assumptions were used in the calculation, which losses are large, and whether it matches the on-site conditions.

The eight tips introduced here may all seem basic. However, in real projects it is common to proceed without thoroughly checking meteorological data, to determine the azimuth solely from the value on the drawings, to oversimplify shading settings, to input loss coefficients out of habit, or to forget to update the PVSyst model after design changes. When these small deviations accumulate, discrepancies arise between the predicted energy generation and the actual system conditions.

PVSyst is not intended to replace a designer’s judgment; it is a tool for comparing design conditions and visualizing their impact on power generation. To perform high-accuracy simulations, it is necessary to correctly understand the site’s terrain, obstructions, orientation, height, and equipment layout, and to reflect those in the input conditions. In other words, the accuracy of PVSyst is determined not only by the software itself but by the quality of the site information.

To further improve the design accuracy of solar PV systems, it is effective to utilize precise location information obtained on site. LRTK, as a GNSS high-precision positioning device that can be attached to an iPhone, can be used for acquiring on-site coordinates, verifying equipment locations, geotagging photos, point cloud measurements, and cross-checking with design data. If you want to increase simulation accuracy in PVSyst as well, accurately capturing on-site conditions, obstructions, and racking positions will raise the reliability of input conditions. As a means of connecting desk-based design with on-site realities, combining LRTK makes it easier to carry out the design, construction, inspection, and maintenance of solar PV systems in a more practical way.