Complete Guide to Getting Started with PVSyst | 9 Steps from Initial Setup to Analysis

Table of Contents

• Get an overview before starting PVSyst

• Step 1 Decide the analysis objectives and decision criteria first

• Step 2 Create a new project to set up the project container

• Step 3 Set the site and meteorological data

• Step 4 Decide the azimuth and tilt

• Step 5 Enter the system configuration

• Step 6 Adjust the detailed losses for the actual project

• Step 7 Set the shading conditions

• Step 8 Run the simulation and refine the outputs

• Step 9 Interpret and compare the results to inform decisions

• Common pitfalls in PVSyst and how to avoid them

• Summary

Understand the overall picture before starting PVSyst

When you start using PVSyst, items such as orientation, system, losses, shading, output, and comparison appear one after another. However, the overall picture is surprisingly simple. First, create a project that holds the location and meteorological data as the project's basic information; within that project create analysis cases; for each case define the orientation and system configuration; add losses and shading as needed; and finally view reports, loss diagrams, monthly values, daily values, and hourly values to compare. In other words, the core of PVSyst is the "project container" and "multiple comparable cases".

If you understand this overall picture first, you will reduce the common confusion that occurs during the initial setup. For example, you do not have to fill in everything at the outset, such as the 3D depiction of near-field shading or all detailed losses. The official tutorial also shows an approach where you first run one calculation with minimal conditions, and then create separate cases adding distant shading, near-field shading, and individual losses for comparison. In practice, this workflow is also the least likely to fail and makes it easier to follow the meaning of the analysis.

Step 1: Decide the analysis objectives and decision criteria first

The first thing you should do when using PVSyst is not to open the interface, but to decide what you are trying to assess with the analysis. Whether you want a rough estimate of energy production, to compare azimuths and tilts, to examine the impact of shading, or to prepare materials for a report, the level of detail required in the initial settings will vary. Because PVSyst can produce many outputs—loss diagrams, performance ratio, monthly values, daily values, hourly values, custom tables, graphs, and so on—if you start using it with an unclear objective, you may generate numbers but be unable to reach a decision.

In practice, work proceeds more smoothly if you can state the initial analysis objective in a single sentence whenever possible. For example: "Create a baseline case for annual power generation under these site conditions," "Check how much difference arises from variations in tilt and azimuth," or "Update to a realistic case that includes the effects of obstructions." Having that one sentence makes it natural to create the first case simply as the baseline and then add conditions one by one in subsequent cases. If the axis of comparison is clear, you can avoid situations where, when you review things later, you can no longer tell which case was intended to test what.

Step 2 Create a new project and set up the project container

Next is creating a new project. In PVSyst, a project serves as the basis for an assignment and contains multiple analysis cases. The project mainly includes common parameters such as geographic location, meteorological data, and, if necessary, reflectivity, while each case differs in orientation, system configuration, losses, and shading. In other words, the initial project you create is not a one-off estimation file but the parent container for developing that assignment. If you are careless here, it will be difficult to make comparisons later.

When creating a new project, you should consider how to name it from a practical perspective. The official tutorial also treats project names on the assumption that they will appear in final reports and project lists. Therefore, naming projects so that the location, project category, and whether it is a baseline case or a comparative case are clear at a glance will prevent confusion as the number of cases grows. At the initial stage, it is effective to create one baseline case first, save it, and then derive cases such as a shading-added version, a loss-adjusted version, and an orientation-comparison version. PVSyst is designed with this comparative workflow in mind.

Step 3 Set the location and meteorological data

The factors that most strongly affect PVSyst's analysis quality are the site and the meteorological data. According to the official documentation, for detailed simulations the required hourly meteorological data are global horizontal irradiance and ambient air temperature, while diffuse horizontal irradiance and wind speed are treated as optional items. These data are managed as meteorological files and can be prepared from the built‑in meteorological database, external meteorological sources, or a company's own measured data. In other words, PVSyst is a tool where you first decide where to build and under what weather conditions to perform the calculations before proceeding to the main computations.

What's important here is not to be reassured by matching only the location name. In practice, you may be tempted to substitute data from a nearby city, but assumptions change depending on elevation, whether the site is coastal or inland, and whether there is snow or reflective conditions. For initial settings, it is realistic to adopt the most appropriate weather data closest to the project and reflect any special conditions in subsequent cases. The same applies to reflectivity: under general conditions, start from standard values and only revise them when there is snow-surface reflection or unusual ground conditions. Rather than incorporating all exceptional conditions from the outset, fixing the baseline case first makes it easier to explain the meaning of any differences later.

Also, when you are not yet familiar with using PVSyst, it is recommended to save it as a separate case as soon as you swap the meteorological data. The reason is simple: if the causes of differences in energy production are mixed—whether it is the site, the orientation, or the loss conditions—it becomes difficult to determine. Comparing projects is not merely looking at which numbers are larger or smaller, but creating a state in which you can explain what change produced the difference. In that sense, meteorological data is the first step and an important parameter that continues to affect the final interpretation.

Step 4 Determine orientation and tilt

Once the site and meteorological data are determined, the next step is orientation and tilt. Even in PVSyst’s basic workflow, it is assumed that the surface orientation is defined for each analysis case. A fixed-tilt surface is the most basic configuration, and its characteristic is that you can easily create a reference case simply by specifying the tilt angle and azimuth. For the initial setup, it is best to first set representative conditions for the project in this simplest form. What is important here is not to make the first case a complete reproduction of the field, but to make it a meaningful baseline for comparison.

PVSyst includes a simple optimization tool for fixed-tilt configurations that lets you get a rough sense of how tilt angle and azimuth affect energy production. Rather than serving as a precise calculation for deciding the final design, it is meant to help you understand how far the conditions you plan to use are from the optimum. Especially for beginners, using this feature to first check the sensitivity to azimuth and tilt makes it easier to grasp why creating comparative cases is necessary. The first step to mastering PVSyst is to look not at a single number but at the response to changes in conditions.

In practical work, due to roof-surface constraints and multi-surface configurations, a single orientation may not be sufficient. Recent documentation has strengthened the approach of handling multiple independent orientations within a single project and managing their correspondence with electrical systems and 3D scenes. However, if you try to fully reproduce a complex multi-orientation project in the initial setup, you tend to spend a lot of time on consistency checks alone. First run a calculation on a representative surface, and then create orientation-specific cases and multi-surface cases; this sequence stabilizes both understanding and work efficiency.

Step 5 Enter the system configuration

Once the orientation is decided, enter the system configuration. In PVSyst, a grid-connected system is defined as a whole, including the components on the irradiated surface, the series and parallel connections, the power conversion equipment, and the grid connection. Moreover, you can assign multiple sub-arrays as needed and change the conditions for each sub-array. In other words, this is not simply a screen for entering capacity; it is the place to create the generation-side configuration logic itself.

At initial setup, the starting inputs are basic information such as the target capacity or available area, the equipment on the irradiated surface, and the power conversion equipment. According to the official documentation, when these requirements are entered, PVSyst automatically proposes compatible configurations and is designed to show acceptable ranges for the number of series and parallel strings. What is important here is to quickly create a baseline configuration that is physically and electrically feasible, rather than refining every detail on the first attempt. Once the minimum conditions are met, PVSyst helps carry the design to a state suitable for calculation.

A common pitfall for practitioners is insisting on reproducing the exact physical configuration, which prevents the initial case from ever being completed. However, the primary goal at the start is to build the foundation for the analysis. First create a baseline case with a representative configuration, and then adjust capacity allocation, connection conditions, multiple sub-arrays, output constraints, and so on in subsequent cases. The official tutorial also recommends running the first simulation with only the minimal parameters and then adding individual conditions afterward. It becomes easier to understand how to use PVSyst if you view it not as finishing all inputs at once but as progressively maturing comparable cases in sequence.

Step 6: Adjust detailed loss for real-world projects

After entering the system configuration, the next step is setting the losses. What you should remember here is that PVSyst’s loss parameters come with reasonably sensible default values from the start, and the official guidance recommends carefully reviewing them after the first simulation to match the specific project. In other words, you don’t need to finalize all losses at the initial setup stage. Rather, it’s more sensible to run the simulation once, review the loss diagram, and then decide which parts and to what extent to adjust them to the actual project.

PVSyst's detailed losses include incidence-angle correction, soiling, low-irradiance effects, temperature, initial degradation, quality variability, mismatch, wiring, auxiliaries, and downtime. Because these are reflected in hourly, daily, and monthly results and in loss diagrams, how each item is set affects not only annual values but also the seasonal profile. In particular, soiling is highly dependent on the surrounding environment and rainfall conditions; as the official documentation shows, its meaning changes between residential areas, agricultural surroundings, and industrial dust environments. Therefore, it is important to understand that loss settings should not be fixed by general rules but should be carefully chosen to reflect the project’s site-specific conditions.

As a practical workflow, for the baseline case run the calculation once with values close to the initial settings, and then in subsequent cases update in order from the most reliable information — soiling, temperature conditions, wiring, mismatch, downtime rate, etc. — which makes it easier to manage. Following this order makes it easier to trace in the loss diagram how much each loss item contributed. To make the final analysis convincing, it is important not only to present the annual energy production figure but also to be able to explain the breakdown of losses that lead to that figure. PVSyst is precisely the tool for that purpose, so detailed losses should be treated not as a tedious extra task but as a central step that underpins the reliability of the analysis.

Step 7 Set shielding conditions

Shading settings are one of the parts of using PVSyst that tend to cause misunderstandings. The first thing to grasp is to treat far-field shading and near-field shading separately. Far-field shading is used when distant terrain or obstacles affect the entire receiving surface uniformly; the official guideline suggests targets roughly ten times the installation scale or more away. On the other hand, objects such as nearby buildings or structures that cast partial shadows on part of the receiving surface are treated as near-field shading. If you confuse these, the calculation logic will not align and the meaning of the results will become unclear.

Far-field shading is a relatively simple mechanism defined as a horizon profile made up of a sequence of elevation-angle and azimuth-angle points. You can create the horizon on-screen by adding and editing points, and the official documentation also outlines an approach for importing values based on on-site measurements. Therefore, if you want to examine the effects of mountain ranges or distant terrain, it is easier to first enter the overall conditions here. For the initial setup, unless far-field shading is clearly the dominant factor in a project, it is practical to grasp the trend with the simple input first and refine it in subsequent cases if necessary.

Near-field shading becomes considerably more difficult. Even the official tutorial states that near-field shading is one of the most challenging parts of PVSyst and requires building a 3D scene. Here, because not only the presence or absence of shadows but also which parts are shaded and to what extent affect power generation, it is better to proceed by completing a baseline case and then adding other cases separately. Also, when using multiple orientations or 3D scenes, you must check consistency between the system and the shading definitions. If the correspondence of area or orientation is off, it may look plausible, but the calculation assumptions will be inconsistent. The more complex the project, the safer it is to create cases step by step in the order of baseline case, far-field shading added case, and then near-field shading added case.

Step 8 Execute the simulation and format the output

Once you reach this point, run the simulation. In PVSyst’s detailed simulations, you can manage the calculation period and output contents according to the project conditions. The official documentation explains that the data required as detailed outputs should be defined before the simulation. In other words, the correct order is to decide what you want to see and set the output options in advance, rather than computing first and then deciding what to save. In particular, if you want to analyze time-series results in detail later, you need to prepare the output files from the beginning.

Regarding reports, PVSyst is also quite practical. The results are summarized in a printable report, allowing you to review the main conditions and key results together. In addition, if you want to perform detailed external analysis of hourly, daily, or monthly data, you can export it in CSV format. Officially, this output file is generated at the next simulation run, so you need to decide which variables to include before execution. Deciding at the initial setup whether you only need annual values or also want to check hourly behavior will reduce the amount of rework.

Also, for grid-connected projects, it is possible to include a single-line wiring diagram in the report. This is useful for on-site briefings and internal reviews. In the early stages of estimates, internal discussions, and client explanations, it is often quicker if you can see at a glance how the calculations are configured rather than just the power generation figures. PVSyst is not merely a tool for producing numbers; it also aims to turn analysis results into shareable documentation. Even during the initial setup, if you proceed while considering how much to output to make later processes easier, operational efficiency will improve significantly.

Step 9 Interpret and compare the results and use them to make a decision

What matters most in PVSyst analysis is not running the calculations but correctly interpreting the results. Among these, the loss diagram is an important view that the official documentation positions as "quickly grasping design quality"; it is always included in the annual report and can also be checked on a monthly basis. This allows you to follow, on a single chart, not only the final annual energy yield but also how solar irradiation reaches the receiving surface, passes through each loss, and results in the final output. A shortcut to becoming proficient with PVSyst is to get into the habit of reading the loss diagram to understand, not just the magnitude of the numbers, but which losses are dominant.

On the results screen, you can review data from multiple perspectives—not only a loss diagram but also monthly tables, custom tables, graphs, performance ratios, normalized metrics, and hourly and daily behavior. The official documentation likewise describes the results as being organized so that many variables can be visualized by month, day, and hour rather than only in the main report. Therefore, even when two cases have similar annual energy production, you only learn by inspecting the detailed views whether the summer temperature effect is large, morning/evening shading is significant, or assumptions about soiling or wiring losses are driving the difference. Making decisions based solely on comparisons of total energy output tends to overlook opportunities for design improvement.

When making comparisons, it is important to have a single difference that carries a clear meaning. PVSyst has a feature for comparing reports from different projects or analysis cases, and you can also operate it to quickly load and compare multiple cases within the same project. Precisely for that reason, when comparing two cases you should not change everything—location, orientation, losses, shading, and configuration—but should narrow the changes according to what you want to verify. For example, if you want to examine orientation differences, keep the loss conditions fixed; if you want to examine shading differences, keep the configuration fixed. By following this comparison practice, PVSyst results become not just tables of numbers but materials for decision-making.

Common Pitfalls in PVSyst and How to Prevent Them

What beginners using PVSyst often stumble over is trying to build a complete, all‑in‑one model on their first attempt. However, the official tutorial recommends the opposite: first create a baseline case with the minimum necessary parameters, and then create derived cases that add far shading, near shading, detailed losses, and so on. Simply following this order makes isolating input inconsistencies much easier. The initial case is not a finished product but a starting point for comparison. Recognizing this is the single biggest tip for stabilizing your use of PVSyst.

The second stumbling block is confusing far-field shading with near-field shading. If you do not keep the division of roles—horizon profile for distant terrain and 3D near-field shading for nearby obstacles—the way shadows are handled will change. In particular, if you simplify nearby buildings or structures as far-field shading, you may not be able to adequately represent the electrical effects of partial shading. Conversely, trying to model all distant terrain in detailed 3D is too heavy for an initial setup. Shading should be included for accuracy, not to stop the first pass.

The third is fixing loss settings at a generic level. PVSyst provides initial values, but the official documentation also recommends reviewing each loss and adjusting them for the project after the first simulation. Even something like soiling changes meaning depending on environmental conditions, and wiring, downtime rates, and temperature conditions also vary by site. Therefore, don’t stop at the initial case—look at the loss diagram and decide which parts should be brought closer to real conditions. The quality of the analysis is determined not by how much detail you entered, but by whether you can explain the correspondence between your assumptions and the results.

The fourth issue is that the design of multi-case comparisons becomes sloppy. PVSyst includes mechanisms to compare multiple conditions, including batch calculations using multiple weather files and sites, but if you proceed to multivariate comparisons without having the basics down, the meaning of the differences becomes unreadable. First, start by stacking a baseline case and comparison cases that each contain only one change. After that, moving on to multi-year weather, multiple sites, and parametric comparisons makes it easier to reconcile depth of analysis with explainability.

Summary

If I were to summarize how to get started with PVSyst in one sentence, it comes down to creating the project framework, running a baseline case once, and then bringing the conditions one by one closer to reality to compare. Decide the site and meteorological data, set the azimuth and tilt, input the system configuration, add detailed losses and shading step by step, and finally judge based on loss diagrams, performance indicators, and time-series behavior; simply following this order turns PVSyst from a difficult piece of software into a very powerful analytical foundation that supports practical decision-making.

And if you truly want to improve the accuracy of the analysis, it is also essential to make the desktop input conditions themselves accurate. If topography, elevation, installation location, orientation, and obstacle positions remain unclear, no matter how carefully you configure PVSyst the underlying assumptions will be unstable. In situations where you want to streamline the acquisition and verification of such site conditions, using an iPhone-mounted GNSS high-precision positioning device, LRTK, to capture point clouds with absolute coordinates and to perform on-site position checks is an effective way to raise the input accuracy for the analysis itself. Proceed with the desktop analysis in PVSyst and improve the precision of on-site information with LRTK. This combination reduces the discrepancy between design and the field and makes it easier to reach more convincing conclusions.

lefixea.com