How to Read PVSyst Results That Seem Too High: 4 Approaches｜Organizing Checkpoints

Table of Contents

• Initial mindset for what to check first when PVSyst results seem too high

• Approach 1: Check whether the meteorological data and irradiation assumptions are overly optimistic

• Approach 2: Check whether loss settings are set too low

• Approach 3: Check whether layout, orientation, tilt, and shading conditions are modeled more favorably than reality

• Approach 4: Check whether output limits, outages, degradation, and operational conditions are reflected

• When comparing high results, read PR and annual energy separately

• Improve the accuracy of interpretation by cross-checking with measured values and site information

• Summary: For overly high results, focus on the consistency of assumptions rather than the numbers themselves

Initial mindset for what to check first when PVSyst results seem too high

When reviewing PVSyst results you may sometimes think, “Is this annual energy estimate too high?”, “The PR looks too good compared with other analyses,” or “This is significantly above actual performance.” In solar power plant energy simulations, even for the same plant, results can vary greatly depending on input conditions, loss settings, meteorological data, how shading is handled, and how output limits are modeled. Therefore, rather than immediately concluding that high results are wrong, it is important to sequentially check which assumptions are pushing up the energy estimate.

In practice, be careful not to judge solely by annual energy. Annual generation is the cumulative result of many factors: system capacity, irradiation, temperature conditions, system losses, outage conditions, and more. If a result seems high, the appropriate response depends on whether the irradiation assumption is high, losses are set too low, shading is underrepresented, or output limits are omitted. Conversely, judging by PR alone is also risky. PR is a useful metric for comparing system efficiency, but its interpretation changes with meteorological conditions, definitions of irradiation, and evaluation scope, so it should be read together with annual energy.

What matters when reading PVSyst is not simply whether the result looks good or bad, but whether the assumptions and the result connect consistently. High generation could indicate good design, but it could also simply reflect overly optimistic loss settings. Especially when using results for business feasibility, bankable documents, EPC comparisons, O&M planning, energy guarantees, or performance assessment, adopting optimistic assumptions without scrutiny can create discrepancies with later actual results. Simulation outputs do not perfectly predict future generation, but organizing the assumptions makes the results more explainable.

This article organizes four approaches to check when PVSyst results seem too high. The key checkpoints are meteorological data, loss settings, layout and shading, and operational conditions. By reviewing these four areas in order, you can largely isolate the causes of overly high generation estimates. To help practitioners verify while reading a report, this article goes beyond describing settings and explains concretely how to interpret them and what differences affect results.

Approach 1: Check whether the meteorological data and irradiation assumptions are overly optimistic

The first thing to check when PVSyst results seem too high is the meteorological data and irradiation assumptions. Solar generation ultimately depends strongly on how much solar energy reaches the plant. Thus, even if the system-side settings are identical, higher annual irradiation inputs lead to higher generation. If the result is significantly above expectations, first verify whether the meteorological data used are appropriate for the site and whether the irradiation is higher than other references or neighboring data.

When reviewing meteorological data, it is important to look not only at annual global irradiation but also at monthly trends. Even if the annual value is similar, the monthly distribution may not match the local climate. For example, in snowy regions winter irradiation might be overestimated, in areas with a rainy season or frequent cloud cover summer irradiation might be overestimated, or in mountainous areas cloud influence might be underestimated. If annual generation is high, check which months are driving the excess. Whether only specific months are high or the whole year is high changes the cause estimation.

Next, check the representativeness of the meteorological site. If the plant coordinates and the meteorological representative point differ, effects of elevation, distance from the coast, mountain shadows, local clouds, snow, and fog may not be reflected. Especially in mountainous areas, basins, coastal zones, and snow-prone regions, irradiation and temperature can change over tens of kilometers. When PVSyst results seem high, do not assume “standard meteorological data are fine”; confirm whether the data reflect the actual site climate.

Temperature assumptions are also important. PV panel output decreases as temperature rises. If the meteorological temperature is set lower than reality, generation tends to be overestimated. When annual generation looks high, check not only irradiation but also monthly mean temperatures and conditions during hot periods. If summer generation looks too high, examine whether temperature, wind speed, and how temperature-related losses are modeled are more favorable than reality.

Conversion from horizontal plane irradiation to array-plane irradiation also affects results. PVSyst calculates the irradiation incident on the panel surface from the input meteorological data. At this stage, the ratio of direct to diffuse components, tilt angle, azimuth, and ground-reflected irradiance treatment are relevant. If results are high, verify whether the effective irradiation on the panel surface is reasonable. Even if horizontal-plane irradiation appears normal, the tilt-plane values after conversion may be high.

Ground albedo settings are an easily overlooked item. If ground reflectance is set high, reflected irradiance on the panels increases and generation rises. In snowy regions, winter reflectance is sometimes estimated high, but actual effects depend on snow cover duration, snow surface condition, snow removal on paths and under panels, and snow adhesion to panels. If albedo is increased without adequately accounting for losses due to snow accumulation or winter operational constraints, winter generation may be overestimated.

When reading meteorological data, it is insufficient to look only at “how many kWh of annual irradiation.” Confirm irradiation, temperature, monthly trends, site representativeness, conversion to tilted plane, and ground-reflected irradiance consistency. If the meteorological assumptions overly portray favorable site conditions, even very careful loss settings later will tend to yield optimistic overall results.

Approach 2: Check whether loss settings are set too low

The next thing to examine when PVSyst results seem too high is loss settings. In a solar plant simulation, various losses are deducted from incident solar energy to obtain the final electrical energy at the grid connection or delivery point. If loss settings are small, generation will naturally be high. When results look too optimistic, it is important to check “which losses are set too low,” “which losses are unaccounted for,” and “whether values are optimistic relative to reality.”

First, check wiring losses. Wiring losses depend on DC-side and AC-side routing, transformers, and distance to the grid connection point. In distributed layouts DC distances can be short, but when junction boxes, combiner boxes, converters, transformers, and substation equipment are far apart, losses occur. When results are high, verify that wiring losses are not uniformly set too low and that they are reasonable for actual cable lengths, conductor cross-section, voltage, and current conditions.

Even if wiring loss numbers look small, they affect annual generation. In large plants, distances from racks to electrical equipment differ by zone. Even if simplified with an average, confirm whether that average reflects the actual layout. If you set representative distances from drawings only, the effects of distant zones or long trunk lines may be omitted. When results are high, compare wiring loss assumptions with electrical design diagrams and single-line diagrams.

Next review module-related losses. Module nameplate power tolerance, low-light performance, temperature coefficients, mismatch, aging degradation, soiling, and incidence angle losses all affect final generation. The report shows these losses individually, but you must check the total loss magnitude. For example, if soiling loss is almost absent, mismatch is too small, incidence angle loss is understated, or degradation is not considered for the evaluation year, results will be biased high.

Soiling losses vary widely by region and environment. Near farmland, dusty sites, adjacent construction, snow or pollen-prone areas, and locations with bird fouling may require larger than standard soiling allowances. Conversely, in areas with natural rain-cleaning or with scheduled maintenance cleaning, large soiling losses may not be necessary. The key is whether soiling assumptions tie to local conditions and operational plans. If results are high, check whether soiling losses are set too low and whether monthly assumptions are realistic.

Temperature losses also need checking. Temperature losses depend on meteorological temperature data, wind speed, mounting configuration, rear-side ventilation, and racking design. Roof-mounted or poorly ventilated installations tend to have higher module temperatures and larger temperature losses. For ground mounts, racking height and rear-side airflow change the thermal conditions. If results are high, confirm that temperature losses are not underestimated and that thermal conditions suit the mounting type.

Converter/inverter losses also impact generation. If conversion efficiency is set too high, low-load efficiency decreases are not reflected, capacity ratio is not properly evaluated, or effects of output limits and power factor operation are omitted, results can appear high. In particular, confirm the DC-to-AC capacity ratio, the amount of clipping at peak times, and whether the converter’s input range and operating conditions match reality.

Auxiliary losses and standby power are also easy to overlook. Plants consume power for monitoring, communications, cooling, control, substation equipment, and nighttime standby. What PVSyst includes in these auxiliary consumptions varies by project, and different treatments among comparison studies cause result differences. For high-generation reports, verify whether auxiliary consumption is unaccounted for or underestimated.

When reading loss settings, avoid judging based on a single loss item. Slight underestimation in wiring loss, soiling, temperature, and auxiliary consumption can cumulatively create large differences in annual generation and PR. When PVSyst results seem high, review the loss list from top to bottom and check whether the losses are explainable for the plant in question.

Approach 3: Check whether layout, orientation, tilt, and shading conditions are modeled more favorably than reality

When PVSyst results seem too high, always check layout, orientation, tilt, and shading treatment. In a solar plant, panel orientation and tilt, row spacing, terrain, surrounding obstacles, inter-row shading, trees, utility poles, buildings, and slope faces affect generation. If these are modeled more favorably than reality, even reasonable irradiation and loss assumptions can produce high results.

First check azimuth and tilt angles. Panels close to ideal south-facing with optimal tilt produce higher generation. However, actual construction drawings may constrain rack orientation and tilt due to site grading, roads, plot boundaries, slopes, drainage, or rack layout requirements, so not all racks will have identical azimuth and tilt. In plants split into multiple plots, azimuth and tilt may vary by plot. If results are high, confirm whether modeled azimuth and tilt match the actual layout and racking drawings.

Be especially careful about optimistic averaging when using representative values. Analyzing the entire site with a single azimuth and tilt averages out variations from plots with differing orientations or tilts. If the representative values are close to the best conditions, the overall generation may be overestimated. For large plants or complex terrain, check whether plot-by-plot conditions are reflected or whether capacity-weighted aggregation is performed appropriately.

Next check how near shading is handled. Narrow row spacing causes front-row shading on rear rows in mornings, evenings, and winter. Shading losses affect both annual and monthly generation. If PVSyst results are high, confirm whether inter-row shading is adequately reflected and whether terrain slope and rack height are correctly input. A model assuming simple flat terrain can miss real elevation differences and slope-induced shading.

Far shading and surrounding obstacles are also important. Mountains, trees, buildings, transmission equipment, slopes, and retaining walls can create shading, especially during mornings and evenings. If these are not included in the model, generation will be overestimated. In mountain valleys or wooded sites, far shading cannot be ignored. When reviewing high-generation reports, check that shading is not set to “none” or oversimplified.

Terrain modeling must not be overlooked. On flat sites simplification has limited impact, but on sloped or terraced sites elevation differences, slope direction, cut-and-fill, slope faces, and steps change shading and incidence angles. A model with neat flat layouts can hide portions of the site that are lower, surrounded by slopes, or have varying rack heights. If results are high, verify how comprehensively the model represents real terrain.

Also beware differences between theoretically available area and actual rack placement. Early-stage studies may model an ideal layout across the entire plot, but detailed design must account for setbacks, maintenance paths, drainage, O&M space, regulations, ground conditions, and electrical equipment areas that alter rack placement. If an old layout or initial plan is used without updating, the modeled site may show more capacity or better conditions than actually feasible.

Azimuth, tilt, and shading are items that are hard to judge from report numbers alone. Rather than only viewing numerical values, check layout drawings, racking diagrams, topographic maps, site photos, and survey data to find the cause of overestimation. When PVSyst results seem high, determine whether the simulated plant has become an “idealized plant” rather than the “actual plant.”

Approach 4: Check whether output limits, outages, degradation, and operational conditions are reflected

A frequently overlooked area when PVSyst results seem high is output limits, outages, degradation, and operational conditions. Plants do not operate at design performance continuously throughout the year. Grid-side curtailment, converter output caps, power factor operation, equipment outages, inspections, failures, communication issues, snow, cleaning frequency, and aging degradation all reduce generation in practice. If these are not included in the analysis, PVSyst results look optimistic.

First check the relationship between DC capacity and AC capacity. If the AC converter capacity is smaller than the DC array capacity, the plant will hit output limits during favorable irradiance periods and experience peak clipping. This is a common design approach, but if not properly reflected in the analysis, generation will be overestimated. When results are high, check how output limits, capacity ratios, and peak clipping losses are reported.

Power factor operation is also important. Some plants are required by grid connection conditions to operate at a certain power factor. Depending on converter capacity definitions and control methods, this can affect the maximum active power extractable. If power factor settings in the analysis differ from reality, apparent peak limits and losses will change. When results seem too high, read not only whether a power factor number is entered but how that setting impacts available active power.

Grid-side curtailment must be checked. In some regions grid conditions or system constraints cause generation curtailment. Standard PVSyst simulations do not automatically reflect future curtailment events. Therefore, for business evaluation or performance comparison it is necessary to clarify whether curtailment was considered separately, and whether the reported result is pre-curtailment theoretical generation or expected delivered energy after curtailment. If results are high, the curtailment component may be missing.

Outage losses are another practical check. Plants experience downtime for routine inspections, equipment replacement, communication faults, protection trips, grid outages, converter failures, and substation maintenance. In simulations these are sometimes handled as availability or outage loss. If results are high, check whether outage losses are included and whether the values are realistic. New installations often assume ideal availability, so be cautious for long-term business plans.

Aging degradation is important. PV output gradually declines over time. Whether you are evaluating first-year generation, multi-year averages, or generation after several years of operation changes which numbers to compare. If a high result is actually the pre-degradation first-year estimate, and you compare it to measured performance years later, discrepancies will naturally appear. Verify whether you are comparing like with like.

In snowy regions, check how snow accumulation and snow removal are modeled. High ground albedo may increase winter generation in models, but you must also consider snow adhesion on panels, time until snow slides off, rack height, snow removal practices, and shading from piled snow. If you include only the albedo benefit without snow-covered losses or winter stoppages, the result will be too optimistic. If winter monthly generation seems higher than reality, investigate this point closely.

Operational assumptions about cleaning and maintenance also matter. Whether the model assumes scheduled cleaning or relies on natural rain-wash, and the frequency of mowing and vegetation control, all affect long-term generation. If results are high, check whether the report implicitly assumes ideal maintenance. In practice, budget, task frequency, access, and seasonal constraints often prevent continuous ideal conditions.

Output limits, outages, degradation, and operational conditions may not be fully apparent from a PVSyst report alone. Therefore, check grid connection requirements, operation plans, maintenance plans, contractual evaluation scope, and past operational data along with the report numbers. When results appear high, differentiate between “theoretical plant generation potential” and “operationally deliverable energy” and read the results accordingly.

When comparing high results, read PR and annual energy separately

When PVSyst results seem too high, many practitioners check annual generation and PR. These two metrics are related but serve different purposes. Annual energy directly affects project revenue and sales volume, while PR is an indicator of system performance and loss magnitude. To interpret results correctly, consider these separately.

If annual generation is high, the causes can broadly be: higher input irradiation, larger system capacity, smaller losses, less shading, or smaller outage assumptions. PR, on the other hand, indicates how effectively the system converts received irradiation into electrical energy. High irradiation does not necessarily produce a high PR. You can have high annual generation due to abundant irradiation but ordinary PR, or normal irradiation with an optimistic loss assumption producing a high PR.

For comparisons, it’s helpful to look at annual generation per unit capacity first. If plant capacities differ, simple annual energy comparisons are meaningless. Looking at annual generation per 1 kW of capacity makes it easier to compare meteorological and loss conditions. Then look at PR to assess whether system losses are reasonable.

However, PR is not perfect. PR varies depending on which irradiation reference is used, which energy point is used as the numerator, and which losses are included. For example, PR differs if evaluated at inverter AC output, at the point of common coupling, including auxiliary consumption, or after curtailment. When PVSyst results look high, check the PR value itself and the scope over which PR is evaluated.

When comparing with other analyses, organize whether the same meteorological data, capacity, evaluation point, and loss scope were used. Comparing “this PR is higher” or “this annual generation is lower” without aligning conditions will not lead to a correct conclusion. When preparing comparison tables, list annual irradiation, tilt-plane irradiation, system capacity, wiring losses, temperature losses, shading losses, soiling losses, outage losses, curtailment losses, and the final evaluation point so it is clear where differences originate.

When PVSyst results seem high, do not arbitrarily increase losses to reduce annual energy. Separately read PR and annual generation to determine whether the excess is due to irradiation, system losses, or omitted operational conditions; this clarifies which assumptions to adjust.

Improve the accuracy of interpretation by cross-checking with measured values and site information

Determining whether PVSyst results are too high cannot be done solely from the report. The most effective method is cross-checking with measured values and site information. For existing plants, review actual irradiation, generation, outage history, curtailment records, temperature, snow, and cleaning history to identify specific discrepancies with the simulation. For new projects without measured data, use nearby meteorological trends, terrain, design drawings, site photos, and survey results to improve the validity of assumptions.

When comparing with measured data, align the comparison period. Simply comparing PVSyst annual values with a partial measured period is not valid. When measured data exist, compare on a monthly or daily basis and match the same irradiation conditions. If the measured period had lower-than-normal irradiation, lower generation compared to simulation is expected. Conversely, if irradiation was higher than normal yet generation is low, this suggests losses or outages.

Pyranometer data are particularly useful. Combining measured irradiation with measured generation lets you verify actual PR and system efficiency. If PVSyst results are high, pyranometer data help distinguish whether the meteorological baseline is high or the system efficiency is low. Note, however, the pyranometer’s mounting angle, cleanliness, calibration, missing data, and shading effects must be considered; if the instrument itself does not reflect site conditions well, comparisons will be inaccurate.

When examining generation data, confirm the measurement point. Values differ depending on whether they are inverter output, substation meter, sales meter, or SCADA aggregation. To compare with PVSyst, match the simulation’s output point with the measured point. For example, comparing PVSyst inverter-equivalent output with measured substation energy will show differences due to transformer losses and on-site consumption.

Review outage histories. Low measured generation relative to simulation does not necessarily indicate poor performance. Scheduled inspections, equipment faults, communications failures, grid outages, curtailment, snow, or construction-related stops reduce generation. When PVSyst results seem high, the measured side may include many outage periods, making the simulation appear optimistic. Separate theoretical performance differences from operational outage impacts.

Site photos and survey data are also effective. Photos reveal nearby trees, mountain shadows, slope faces, utility poles, row spacing, vegetation, soiling, snow, and drainage that are not visible from reports. Shading and soiling especially are difficult to judge from drawings alone. High-precision position information and survey data make it easier to confirm rack positions, azimuth, tilt, and elevation differences and ensure PVSyst inputs match the site.

In practice, do not stop at checking the PVSyst report; link it to site information. When results appear high, focus on identifying “which site conditions are missing from the model.” Combining meteorological, electrical, civil, and operational information increases the explanatory power of generation estimates.

Summary: For overly high results, focus on the consistency of assumptions rather than the numbers themselves

When PVSyst results seem too high, do not judge solely by annual generation or PR numbers. Instead, sequentially check the connection between assumptions and results. The four most important areas are meteorological data and irradiation, loss settings, layout and shading, and output limits and operational conditions. Reviewing these four areas will largely clarify the main causes of high generation estimates.

For meteorological data, inspect monthly trends, site representativeness, temperature, tilt-plane irradiation, and ground reflectance, not just annual irradiation. For loss settings, confirm that wiring losses, temperature losses, soiling losses, mismatch, auxiliary consumption, and conversion losses are not set unrealistically low. For layout, verify azimuth, tilt, inter-row shading, far shading, terrain, and actual rack placement. For operational conditions, ensure output limits, outages, degradation, power factor operation, snow effects, and maintenance assumptions are reflected.

Avoid increasing losses arbitrarily without identifying causes or simply accepting optimistic simulation results uncritically. Simulation outputs are calculations based on input assumptions, so their validity depends on the validity of inputs. If the assumptions are explainable, you can justify results even if they differ from other analyses. But adopting high generation with unclear assumptions makes later discrepancies with measured values difficult to explain.

Also, when reading PVSyst, separate PR from annual energy: determine whether high annual generation is due to irradiation, loss settings, or system capacity. If PR is high, check the scope of losses and evaluation points. When preparing comparison materials, list each loss item and show how each assumption affects results to facilitate internal and client explanations.

For existing plants, cross-checking with measured values is essential. Combining irradiation, generation, outage history, curtailment, site photos, and survey data helps concretely identify why PVSyst results are high. For new projects, carefully reflecting drawings and site conditions helps avoid overly optimistic analyses.

To accurately capture site conditions, do not rely only on desk-based reports. Confirm position, elevation, azimuth, rack placement, surrounding shading, and as-built conditions on site. LRTK, as a GNSS high-precision positioning device that can be attached to an iPhone, can be used on site for position checks, point acquisition, drawing verification, and post-construction as-built confirmation. If you want PVSyst results to better match reality, precisely measuring site conditions that form the basis of simulation assumptions is crucial. Measuring site conditions and linking them to drawings and simulation inputs leads to more defensible solar plant evaluations, not just relying on the generation analysis numbers.