7 Causes and Countermeasures for Low Power Generation Outputs in PVSyst

Table of Contents

• What to consider first when PVSyst shows low power generation

• Cause 1: Mismatch between meteo data and site assumptions

• Cause 2: Azimuth, tilt, and array layout not aligned with site conditions

• Cause 3: Shading conditions and 3D scene settings are too strict or unnatural

• Cause 4: Temperature loss assumptions do not match site conditions

• Cause 5: String configuration and PCS settings cause generation loss

• Cause 6: Loss factors set too conservatively or overlapping

• Cause 7: Shallow result interpretation leads to misattributing the low values

• How to translate PVSyst power improvements into practical decisions

What to consider first when PVSyst shows low power generation

When a PVSyst simulation produces a power generation result lower than expected, many practitioners first suspect a problem with the equipment configuration itself. Indeed, module capacity, PCS capacity, and array layout choices can greatly affect the outcome. However, in real projects there is rarely a single reason why the generation looks low. Meteorological data assumptions, azimuth and tilt premises, shading estimates, temperature losses, wiring losses, and utilization losses—many conditions overlap to determine the final numbers. In other words, when PVSyst shows low generation, what’s needed is not an immediate distrust of the equipment but a step-by-step isolation of which assumptions are pulling the result down.

A common practical mistake is deciding quickly whether a design is good or bad by looking only at the annual energy figure. That view alone makes it hard to distinguish between a reducible shortfall and a low result that correctly reflects site conditions. For example, a design that realistically includes shading may look conservative on the numbers yet be reasonable from a design perspective. Conversely, a design skewed toward ideal conditions may appear high but drop significantly when assumptions are adjusted closer to actual site conditions. The important thing is not to treat a low result as inherently problematic but to determine whether you can accept the level of that low result.

It’s also more practical to use PVSyst not as a tool that gives one definitive answer but as a tool to refine a design by comparing assumptions. Thus, when you feel the generation is low, treat that number as a signal indicating which assumptions have strong effects rather than as a failure. Looking beyond the annual value—at the loss tree, monthly generation, PR, and how each loss factor was set—makes the outline of causes much clearer. Reading which assumption differences create those numbers is more important than the absolute magnitude of the numbers themselves.

The seven causes and countermeasures explained below are a way to check settings with priorities when generation appears low, rather than randomly tweaking inputs. None of the causes necessarily exists alone; multiple causes may overlap. Still, if you organize them in order, you can see what to fix first and what to accept as site conditions. To use PVSyst effectively in practice, the most important thing is to identify why the output is low and be able to explain that reason.

Cause 1: Mismatch between meteo data and site assumptions

The first thing to check when generation looks low is whether meteorological data and the site assumptions are mismatched. In PVSyst, the baseline for annual generation is set by which site’s weather data you use. Therefore, even with the same equipment conditions, different choices of meteo data can lead to large differences in results. In practice, people sometimes reuse data from a nearby representative site or a previous project; if you proceed without confirming that it truly represents the current site, the generation can appear unnecessarily low.

Be especially careful about relying solely on annual irradiance figures for reassurance. Even if annual values are similar, differences in monthly distribution or temperature conditions can change how generation looks. For example, if the premise assumes weak winter irradiance, generation in that season will tend to be low; strict temperature assumptions will increase apparent temperature losses. When PVSyst shows low generation, before questioning modules or PCS, first confirm whether the meteorological assumptions are natural for the site.

Also, at the site comparison stage, minor differences in the chosen location can change the ranking of proposals. It isn’t fair to compare a proposal using detailed site conditions with another that uses a coarse representative location. When generation seems low, make sure to separate whether the low result comes from the design or from meteo assumptions by aligning the site location and weather data. Because PVSyst faithfully reflects input assumptions, checking the initial meteorological conditions is effective.

As a countermeasure, organize the site’s position information and confirm whether the meteo data truly represents that location. Check not only annual values but monthly irradiance trends and temperature conditions for any oddities; doing so will narrow down the cause. When generation looks low, doubting the meteo assumptions first—though it may seem indirect—is often the fastest way to resolve the issue.

Cause 2: Azimuth, tilt, and array layout not aligned with site conditions

Next, review whether azimuth, tilt, and array layout genuinely fit the site conditions. In PVSyst, setting ideal orientations and angles can make generation look promising at first glance. But in practice, if those orientations and angles do not match the site shape, slope directions, grading conditions, and access requirements, row spacing and shading issues may force compromises that lower generation. In other words, even if the orientation and tilt numbers are correct, if they are unnatural within the overall design, the generation can appear low.

For example, increasing the tilt to improve irradiance might increase shadowing between front and rear rows and force wider row spacing. Conversely, reducing tilt to squeeze row spacing may change seasonal output characteristics. Likewise, aligning to an ideal azimuth that differs from the site’s actual orientation can fragment the array and manifest as end-losses or difficulty providing access. When PVSyst shows low generation, check whether these design parameters are being forced unnaturally.

In practice, impressions from a site visit can also differ from what was entered into PVSyst. A site that felt roughly south-facing may actually be tilted more east-west; a slope’s direction may have been interpreted differently when laying out arrays. These small mismatches, combined with shading and row spacing issues, can reduce annual generation. Installation parameters are not just input values but an interpretation of site conditions.

As a countermeasure, review azimuth, tilt, and array layout together rather than optimizing each separately. Verify whether they truly fit naturally within the site—whether row spacing and access can be secured without compromise, and whether they conflict with slopes or existing conditions. When improving generation in PVSyst, it’s more important to relate orientation and layout back to site reality than to chase the numbers themselves.

Cause 3: Shading conditions and 3D scene settings are too strict or unnatural

Shading conditions and 3D scene settings are also frequent causes of low generation. PVSyst allows detailed shading representation via Near Shading and 3D scenes, but if those assumptions are overly strict or misaligned with actual site conditions, the perceived generation changes significantly. In practice, people sometimes over-model obstacles when trying to be cautious, or they omit primary obstacles—both distort results.

For example, if slopes, retaining walls, buildings, trees, or adjacent arrays are not modeled correctly as primary shading sources, shading losses will not match reality. Also, if an array layout with unrealistic row spacing or tilt is directly imported into the 3D scene, shading conditions may become excessively severe, reducing PR and generation. PVSyst’s shading analysis is powerful, but it requires that the input scene naturally represents the site conditions.

Furthermore, underestimating the impact of partial shading by looking only at shaded area can lead to misdiagnosis. Even if the shaded area is small, if shading is concentrated on specific rows or strings, the resulting loss can be large. Conversely, shading that looks widespread visually may have limited annual impact if it occurs only at limited times or seasons. When generation is low, review not only the total shading loss but also where the shading comes from, when it occurs, and which parts of the array it affects.

As a countermeasure, reconstruct the 3D scene accurately without excess or omission and identify the primary shading sources. Prioritize review of nearby arrays, slopes, buildings, and trees, and create comparison scenarios to see which shadows are dominant—this makes root-cause separation easier. When PVSyst shows low generation, shading should be treated not just as a loss item but as a point to reassess the validity of the layout itself.

Cause 4: Temperature loss assumptions do not match site conditions

Temperature loss assumptions are an often-overlooked cause of low generation. Because temperature is less conspicuous than shading or orientation, it is frequently handled as a lumped loss at the end. However, PVSyst’s generation outcome is not determined solely by ambient temperature; how modules heat up in their installation environment also matters. In other words, temperature loss is not just a matter of temperature data but also a function of array layout and ventilation conditions.

For example, a design with high array density, many surrounding structures, and poor ventilation will tend to show harsher temperature losses under the same weather conditions. Conversely, an array with generous row spacing placed in an open area will show relatively better temperature performance. Because PVSyst reflects these differences in annual generation and PR, when generation appears low you should not dismiss temperature loss as a minor item but read it together with the layout conditions.

Module characteristics also interact with temperature assumptions. A module with a higher nameplate output may not translate into the expected advantage once temperature effects are included. Conversely, even small differences elsewhere can make temperature loss the deciding factor between designs. When comparing proposals in PVSyst, including temperature losses can reveal which proposal is actually superior for the site.

As a countermeasure, do not determine temperature losses solely from meteorological data; review array density, surrounding conditions, and ventilation potential as well. If you have comparison scenarios, check how row spacing and tilt differences affect temperature losses to identify causes. When PVSyst shows low generation, question temperature environment assumptions in addition to shading and equipment factors.

Cause 5: String configuration and PCS settings cause generation loss

When generation is low, also check whether string configuration and PCS settings are causing avoidable loss. In practice, people may assume module count and PCS capacity are sufficient and move on, but PVSyst results are sensitive to the naturalness of those groupings. With the same site and equipment capacities, unnatural string arrangements or PCS connections can create mismatches or apparent output limitations, making generation and PR look low.

For example, if rows with different conditions are forcibly combined into the same string, shading, orientation, or temperature differences can concentrate on one group and show up as mismatch loss. Similarly, if PCS settings are not appropriate for the project—if the DC/AC ratio is unnatural—output clipping or underutilization of the PCS may occur. In PVSyst you should check not only whether the calculation holds mathematically but whether the configuration makes sense as a design.

PCS settings also affect the impression of PR as well as annual generation. Loading the DC side heavily can make annual energy appear higher while PR drops due to clipping; conversely, leaving too much margin on the PCS side can look safe but may not fully utilize module potential. In practice, the goal is not to choose one over the other but to find a natural balance for the site and operational assumptions.

As a countermeasure, confirm that stringing does not mix differing conditions and that PCS settings connect naturally to the DC side. Create comparison scenarios and slightly vary string arrangements and DC/AC ratios to see where generation is being lost. When PVSyst shows low generation, don’t just look at equipment capacities—reexamine the coherence of the whole system.

Cause 6: Loss factors set too conservatively or overlapping

A common cause of low generation is setting various loss factors too conservatively or with overlap. In practice, driven by the desire to be cautious, teams tend to tighten losses such as temperature, soiling, wiring, utilization, and shading a little more than necessary. Each may seem reasonable on its own, but stacking them can produce an outcome that is overly pessimistic compared to reality. In PVSyst, this accumulation directly reduces final generation.

For example, if you model shading very conservatively and also apply conservative soiling, strict temperature losses, and thick utilization losses, each item may look plausible but the total becomes overly conservative. Conversely, reusing fixed values for wiring or shutdown conditions can produce a loss structure that is incongruent with the site. Without clarifying what is excessive and what is reasonable, simply increasing numbers makes it hard to see the relative merits of comparison scenarios.

Also, overlapping losses may be hard to notice even when inspecting the loss tree. You might be accounting for shading via Near Shading and then also applying a strong overlapping assumption in another loss item. Or you might assume conservatism in soiling and then double up similar adverse conditions under utilization losses. When PVSyst shows low generation, you need to check not only which losses are large but whether loss items are redundantly covering the same effect.

As a countermeasure, compare the loss tree with input assumptions and confirm that each loss has a distinct role. Being able to explain what each loss represents helps detect overly conservative settings and overlaps. To improve generation in PVSyst, reorganizing how losses are allocated and ensuring overall consistency can be more effective than simply trying to reduce individual loss items.

Cause 7: Shallow result interpretation leads to misattributing the low values

The final cause to mention is not the settings themselves but how results are interpreted. In practice, when annual generation or PR looks low, people often rush to change settings. But if the reason for the low appearance is misattributed, tweaking inputs will steer improvements in the wrong direction. PVSyst returns many results, and without a plan for what to read and how, you can become more confused.

For example, if you assume the annual shortfall is due to temperature, the actual primary cause might be shading. Conversely, if you think a low PR means shading is bad, wiring or utilization losses might be the real driver. Monthly generation plots might reveal that a specific season shows divergence, or the loss tree might reveal the dominant loss. In practice, which screens you view and in what order greatly affects the speed of cause identification.

Also, when reading comparison scenarios, if it’s unclear what was changed and what was kept constant, you will misread the reason for the low result. If you want to see azimuth differences but loss factors also differ, or if you want to compare PCS options but the array layout also changed, the meaning of the numeric differences becomes unclear. Using PVSyst in practice requires being able to organize the relationship between input assumptions and results. It’s necessary to track which assumptions produced which numbers rather than just looking at figures.

As a countermeasure, set an order for reviewing results. If you first look at annual generation and PR, then the loss tree, then monthly trends, and finally return to individual settings, cause identification becomes much easier. When PVSyst shows low generation, the biggest problem is not the settings themselves but taking measures while misunderstanding the reason for the low values. Improving how you read results is itself an important practical improvement.

How to translate PVSyst power improvements into practical decisions

A common thread in the seven causes and countermeasures above is not to attribute low generation to a single cause. Meteorological data, installation conditions, shading, temperature, stringing, PCS, loss factors, utilization losses, and result interpretation all interact to determine the final outcome. Therefore, when PVSyst shows low generation, what’s needed is not immediate changes to settings but a systematic ordering of which assumptions are dominating. Doing so reveals the priority for design changes and clarifies the meaning of comparison scenarios.

For practitioners, the real value is not producing the highest generation number but being able to explain why that generation has that value. A low result that correctly reflects site conditions is a perfectly valid design. Conversely, a high-looking figure based on overly ideal assumptions is fragile. When using PVSyst in practice, don’t fear low numbers—develop the ability to read their causes.

Also, to improve the accuracy of generation estimates, it is essential not to rely solely on desktop simulation. If site boundaries, slopes, access ways, surrounding buildings, existing equipment, and maintenance routes are ambiguous, assumptions tend to skew ideal. To connect PVSyst numbers to practical decisions, repeatedly iterate between site understanding and simulation to confirm which assumptions are realistic. Many causes of low generation stem not from outright input errors but from a lack of proximity to actual site conditions.

In that sense, when you need more reliable on-site position confirmation or coordinate acquisition, using an iPhone-mounted GNSS high-precision positioning device such as LRTK can be effective. If site position information and conditions captured on site are organized clearly, assumptions about layout, obstacles, and access in PVSyst become more definite when investigating low-generation causes. If you can raise desktop comparison accuracy with PVSyst and support on-site measurement accuracy with LRTK, improving generation becomes less about tweaking numbers and more about grounded practical decisions. Carefully identifying why generation is low not only improves simulation accuracy but also enhances the design capability that connects desk work and field work.