5 Ways to Interpret PVSyst's Array Loss | Where Is It Reduced?

When checking power generation simulations for solar power plants, you will invariably encounter the concept of Array Loss on PVSyst result screens and in Loss Diagrams. If generation is lower than expected, differs from other companies' simulations, the PR does not improve as much as anticipated, or there is something odd about monthly generation, tracing the causes often yields many clues around Array Loss.

Array Loss, literally translated, means the loss on the array side. In other words, it refers to how much output is reduced from the ideal state during the process in which solar modules receive irradiance and produce DC power. It is a different area of concern from AC-side losses after the PCS or losses at transformers, transmission lines, or the grid interconnection point. In PVSyst, from the moment irradiance enters the module plane until DC power is obtained as array output, various factors accumulate: temperature, low irradiance, IAM, mismatch, soiling, wiring, and conditions before and after clipping.

In practice, it is insufficient to view Array Loss as a single number. For example, a large Array Loss does not automatically mean the module performance is poor. Whether the loss is due to large temperature-related losses, conservative assumptions about soiling or snow, conservative inclusion of DC wiring losses, significant proximity shading, or electrical mismatch will change both the mitigation measures and the explanations. Conversely, caution is also required when Array Loss is too small: site conditions may be being treated too optimistically, some loss factors may be omitted, or the meteorological data or array configuration may not match reality.

This article organizes five perspectives for interpreting PVSyst's Array Loss in practical work. Rather than a mere glossary, it explains—step by step and in a way that can be used for power generation reviews, internal explanations, materials for bank submissions, discussions with EPCs, and comparisons with measured data in O&M—where energy is being lost.

Table of Contents

• First, distinguish which range of losses Array Loss covers

• Check how much is reduced by temperature loss

• Look at reductions due to IAM, low irradiance, and module characteristics

• Assess to what extent site conditions are reflected by mismatch, shading, and soiling

• Check the consistency between DC wiring losses and array output

• Notes when using Array Loss for PR comparisons and measured-data comparisons

• Summary

First determine which range of loss Array Loss falls into

The first thing you should do when reading PVSyst's Array Loss is to distinguish which range of losses it covers—i.e., from where to where. Just because there is solar irradiance does not mean it directly becomes the amount of electricity sold. First there is horizontal-plane irradiance and tilted-plane irradiance; from these the effective irradiance incident on the module surface is determined, the module produces DC power, that DC power enters the PCS and is converted to AC, passes through transformers and AC wiring, and is finally sent to the grid.

Array Loss is the area in this workflow used mainly to read losses on the DC side, from the modules to before and after the PCS input. In other words, it is where you assess how much the actual array output has fallen relative to the DC energy the solar array should originally have been able to produce. Confusing this with AC-side losses or grid-side losses will misdirect the root-cause analysis.

For example, if the annual energy production in PVSyst's results is lower than another company's report, the first things you would want to check are the final Energy injected into grid and the Specific Yield. However, unless you isolate whether the difference is due to Array Loss, Inverter Loss, AC ohmic loss, or Transformer loss, you cannot provide a correct explanation. If Array Loss is large, focus on DC-side conditions such as module temperature, irradiance conditions, shading, soiling, DC wiring, and mismatch. On the other hand, if Array Loss is similar but only the final energy production differs, suspect PCS losses, output curtailment, AC-side wiring, transformer, auxiliary equipment, or grid constraints.

In PVSyst's Loss Diagram, energy is displayed as decreasing from top to bottom. What’s important here is to be aware of what baseline the displayed percentages are relative to. Even if a loss item is shown as -2%, how you interpret it changes depending on whether it is a ratio relative to the total irradiance or relative to the energy at the immediately preceding step. It’s dangerous to simply add multiple loss rates together and treat the result as the final cause of PR reduction.

A common misconception when interpreting Array Loss is to treat Array Loss as a single fixed coefficient. In reality, even if Array Loss appears as a single number on an annual average, it varies greatly by month, time of day, ambient temperature, solar irradiance, and shading patterns. In summer, temperature-related losses tend to increase, while in winter temperature losses tend to be smaller, but the effects of snow, low irradiance, and shading can become more noticeable. In the morning and evening IAM and shading effects are more likely to appear, while around noon temperature rise due to high irradiance and limitations on the PCS side tend to be more relevant.

Therefore, when reviewing Array Loss, the practical workflow is to first grasp the overall picture using annual values, then examine monthly values and the breakdown in the Loss Diagram, and finally verify that the configured settings match the site conditions. Looking only at individual loss figures from the start makes it difficult to see where losses are occurring. By first distinguishing whether the issue is on the DC side, the AC side, related to weather conditions, or due to equipment settings, you improve the accuracy of the review.

See how much it has decreased due to temperature loss

One of the items that tends to account for the largest share of Array Loss is temperature loss. Solar cell modules generate more power the more irradiance they receive, but at the same time the module temperature rises. In typical crystalline silicon modules, as cell temperature increases the voltage drops and, as a result, the output falls. In other words, although power generation increases during periods of strong irradiance, some output is lost due to the temperature rise.

PVSyst uses the thermal loss coefficient when calculating module temperature. Depending on the racking type, ventilation conditions, whether the system is roof-mounted or ground-mounted, and whether the rear side is well ventilated, module temperature can vary even under the same ambient temperature and the same irradiance. The assessment of thermal losses differs between a ground-mounted installation with good back ventilation and a system mounted in close contact with the roof. In ground-mounted megasolar installations, ventilation is generally relatively good, so temperature rise may be lower than for rooftop systems; however, the actual situation varies with snow-prone areas, mountainous regions, coastal locations, and differences in wind conditions.

When examining temperature losses, it is important to look not only at the annual loss rate but also at monthly trends. It is natural for temperature losses to be larger in summer and smaller in winter. If temperature losses are large even in winter, or if summer temperature losses are extremely small, you need to check the inputs for meteorological data, the temperature coefficient, thermal loss settings, and the mounting/installation method. In particular, when comparing with other companies' simulations, even if the same module is used, differences in meteorological data such as ambient temperature and wind speed, and differences in PVSyst's thermal model settings, can lead to discrepancies in temperature losses.

Temperature losses also have a large impact on PR. PR is an indicator of how efficiently solar irradiation is converted into electrical energy, but in regions with high ambient temperatures temperature losses tend to be larger, so PR can appear lower even with the same system design. Conversely, in cold regions temperature losses are smaller and PR can appear higher. However, in cold regions other factors such as snow cover, low solar altitude, long shadows, albedo, and winter solar conditions come into play, so you should not judge advantage or disadvantage based solely on temperature losses.

In reviewing Array Loss, you should not regard large temperature losses as a problem in themselves, but rather assess whether their magnitude is reasonable for the region, module, racking, and meteorological data. For example, it is natural for temperature losses in summer to be relatively large in regions with high ambient temperatures. Conversely, if temperature losses are excessively high in a cool region, you should check whether the thermal loss coefficient is set too stringently, whether the temperature in the meteorological data does not match local conditions, or whether the module temperature coefficient is the value for a different model.

When explaining in practice, it's clearer to say, "Of Array Loss, the temperature loss is the amount by which the module's DC output decreases because the module becomes hot." Additionally, adding, "It tends to be larger in summer and smaller in winter, and varies depending on the local ambient temperature and the racking's ventilation conditions," makes it easier to understand as a physically unavoidable loss rather than merely a negative factor.

Observing reductions due to IAM, low irradiance, and module characteristics

When reading Array Loss, the next items to check are the losses related to IAM, low-irradiance characteristics, and module quality. These are not as intuitive as temperature losses, but they are important when explaining differences in PVSyst's energy yield.

IAM stands for Incidence Angle Modifier and denotes the reduction in the solar irradiance that can actually be effectively utilized due to the angle at which sunlight is incident on the module surface. When sunlight strikes the module surface head-on, losses are small; when it strikes at an oblique angle, reflection at the glass surface increases and the light reaching the cell is reduced. This effect is more likely to occur in the morning and evening, in winter, and during periods of low solar elevation.

IAM losses can look small when you only examine annual values. However, they cannot be ignored when comparing generation in the morning and evening, winter generation, or differences in tilt angle. In particular, when comparing cases with the same system capacity but different tilt angles or azimuths, the way IAM behaves changes. In PVSyst the results vary depending on the module glass properties and the IAM model, so when comparing with other companies' reports it is worth checking whether the same assumptions were used.

Low-irradiance loss is also an important component of Array Loss. Solar modules are evaluated based on their rated output under standard test conditions, but in real-world sites they do not always receive strong irradiance of 1000 W/m^2 (317.1 BTU/hr/ft^2). There are many periods when they operate under low irradiance, such as overcast skies, mornings and evenings, winter, rainy weather, and hazy sunlight. Modules still generate power at low irradiance, but their efficiency is not necessarily the same as under standard conditions. This difference manifests as low-irradiance loss.

When evaluating low-irradiance losses, you need to consider them together with the distribution of the meteorological data. Even if the annual solar irradiation is the same, the module’s operating conditions differ between regions where strong sunlight is concentrated on clear days and regions where cloudy conditions are frequent and low-irradiance periods are long. In regions with many cloudy days, the impact of low-irradiance characteristics may become relatively large. If low-irradiance losses are a concern in PVSyst results, check the characteristic values in the module database, the measured data of the selected module, and the temporal distribution of the meteorological data.

Losses related to module characteristics include the tolerance of nominal power, degradation, quality variability, and how LID and initial degradation are treated. In PVSyst settings and reports, you need to check which items these are included under. Your interpretation will differ depending on whether you are looking at first-year energy production, long-term average energy production, or whether the degradation rate is considered separately.

A common practical situation is when someone looks at PVSyst's Array Loss and asks, "Why is the module-side reduction so large?" In such cases, explaining that it includes not only temperature losses but also IAM and low-irradiance losses helps them understand. Modules do not operate continuously at their catalog output; their output changes slightly according to field operating conditions such as angled incidence, low irradiance, high temperature, and module-to-module variation. Array Loss should be read as the result reflecting those field conditions.

IAM and low-irradiance losses may appear small individually, but when multiple factors accumulate they affect annual energy production. Especially when discussing differences of a few percent in PR comparisons, differences in the settings for these small losses cannot be ignored. When reviewing PVSyst results, after temperature losses, checking that IAM, low-irradiance, and module characteristics are within reasonable ranges makes the structure of Array Loss easier to understand.

See to what extent mismatches, shadows, and dirt reflect on-site conditions

When reading Array Loss in practical work, it is important to check how well site conditions are reflected, namely mismatch, shading, and soiling. These factors can vary greatly depending on the design and site environment and are prone to producing differences compared with other companies' simulations.

Mismatch loss is the loss that occurs due to differences in characteristics between modules connected to the same string or the same MPPT. Even solar modules of the same model do not have completely identical output and current–voltage characteristics. When characteristics deviate within a string, the whole is pulled down to the weakest condition, causing the total output to be slightly lower than the ideal sum. Furthermore, if shading or soiling is concentrated on some modules, the electrical mismatch becomes larger.

In PVSyst, there are items configured as module quality losses and mismatch losses. What should be noted here is that if mismatch losses are underestimated, the calculated energy production tends to be higher than the actual. Conversely, if they are overestimated, the design energy production becomes overly conservative. For large-scale power plants, the appropriateness of the settings is judged by reviewing string configuration, MPPT partitioning, azimuth differences, tilt differences, shading patterns, and module lot variations.

Losses caused by shading are one of the more difficult items to explain among Array Loss. Shading includes far-field shadows from terrain and surrounding objects, near-field shading between racking rows, local shadows from utility poles and trees, and partial obstructions similar to snow or soiling. In PVSyst, the effect of shading can be calculated using the 3D scene and shading settings, but the results change depending on the accuracy of the input model.

When evaluating shading losses, check not only the annual values but also which seasons and times of day the shading occurs. Large shadows on winter mornings and evenings are natural, but if substantial shading also appears during daytime or in spring and autumn when energy production is high, the layout’s shading effects may be impacting output. In particular, for sites on slopes, in mountainous areas, on reclaimed or developed land, or where trees remain, shading modeling is directly tied to the validity of the energy production estimates.

Soiling loss is also an important part of Array Loss. In PVSyst it is often set as Soiling Loss and takes into account sand and dust, pollen, yellow sand, bird droppings, soil dust around agricultural land, coastal salt, and dirt from snowfall and after snowmelt.

The way soiling occurs varies greatly by region and by operation. Some areas are naturally cleaned by rain, while in arid regions or at sites with a lot of soil dust immediately after land development, soiling can accumulate more easily.

Soiling loss is particularly important when comparing with measured data. Even if a simulation assumes a constant annual soiling rate, it can vary greatly on site depending on the season. For example, there may be heavy pollen and yellow sand in early spring, lingering snow in winter, abundant dust during land development work, or high traffic on nearby roads; when there are such site-specific conditions, a single uniform annual setting may not be sufficient to explain the differences.

The important point here is not to see mismatch, shading, and soiling as "bad things" but to look at how much the site conditions are being incorporated into the simulation. For example, if your company's PVSyst yield is lower than another company's report, you may be modeling shading and soiling more realistically or conservatively. In that case, the lower yield can be explained not simply as reduced generation but as a result of having factored in risk. Conversely, if these losses are extremely small, you should verify whether site conditions are being adequately reflected.

When reading Array Loss, check the mismatch, shading, and soiling items and cross-check them against the design drawings, site photographs, topography, racking layout, and surrounding environment. Rather than judging solely by the numbers in PVSyst, going back to the actual site and inspecting the real conditions will clarify the meaning of the losses.

Assessing consistency between DC wiring losses and array output

Among Array Losses, the one most directly tied to design conditions is DC wiring loss. The direct current power generated by the solar modules flows through wiring from the string cables, junction boxes, and trunk lines to the PCS input. Because the cables have resistance, when current flows a voltage drop and Joule losses occur. This is the DC ohmic loss, in other words the DC wiring loss.

DC wiring losses vary depending on cable length, cable cross-sectional area, current, voltage, string configuration, combiner box placement, and PCS placement. In designs that place distributed PCS near the racking, the DC distance to the PCS becomes shorter, which can make it easier to suppress DC wiring losses. On the other hand, when the distance from the combiner box to the PCS is long, cable size is small, current is large, or the number of circuits is high, losses increase.

When reviewing DC wiring losses in PVSyst, check whether the input is a fixed percentage or calculated from cable parameters. In the preliminary stage you may use representative values such as 1% or 1.5%, but in detailed design it is easier to explain if you calculate them based on cable length and cross-sectional area. In particular, for reports submitted to banks or for third-party reviews, you may be asked to justify the DC wiring losses.

What matters when evaluating DC wiring losses is not just the magnitude of the loss rate. It is important whether it aligns with the design philosophy. For example, if PCS are distributed across each block but the DC wiring loss is set high, it may be a conservative setting. Conversely, if the site is wide and the distance to the PCS is long yet the DC wiring loss is set too low, it may be an underestimation.

Also, be aware that DC wiring losses tend to have a greater effect during periods of high power generation. Because wiring losses are related to the square of the current, losses increase during periods of strong solar irradiance when current is high. Even if they look small when viewed as an annual average, they impact output during high-irradiance periods. For projects where PCS clipping or output limiting occurs, it is also a check point to see how much loss is occurring on the DC side before entering the PCS.

When reading PVSyst's Array Loss, it's easier to understand if you also look at the relationship between the array output and the PCS input. The DC power produced by the modules reaches the PCS after passing through temperature, IAM, low irradiance, mismatch, shading, soiling, and DC wiring. Then conversion losses, MPPT operation, and clipping occur in the PCS, and the power proceeds to the AC side. If Array Loss shows a large reduction, that means generation has already dropped before entering the PCS. If the Array Loss looks reasonable but the final energy output is low, you should check the PCS and everything downstream.

DC wiring losses are prone to change due to on-site design modifications. Changes to the PCS location, combiner box layout, cable routing, cable size, or the number of strings will also change the losses. If the PVSyst report remains based on outdated design conditions, it will become inconsistent with the as-built design. During power generation reviews, you must always verify that PVSyst’s DC wiring losses match the latest electrical design.

Precautions when using Array Loss for PR comparisons and comparisons with actual measurements

Array Loss is very useful for comparisons between PVSyst cases and for comparing PVSyst with measured data. However, if used incorrectly, it can lead to incorrect conclusions.

First, when examining Array Loss in PR comparisons, it is important not to compare only the final PR. Even if the PR is low, it does not necessarily mean Array Loss is the cause. PR is affected not only by losses on the array side but also by PCS losses, AC-side losses, transformer losses, auxiliary equipment losses, curtailment, downtime, grid constraints, and other factors. Therefore, when explaining PR differences, you need to look at the breakdown of Array Loss separately from the breakdown of AC-side losses.

When comparing PVSyst reports from other companies, check that the same terms refer to the same scope. In one report, Soiling Loss may appear to be included on the irradiation side, while in another report it may appear as an array-side loss. Differences in notation or aggregation can make the same loss appear in a different place. Therefore, you need to look not only at the item names but also at the flow of the Loss Diagram to confirm at which stage the reduction occurs.

In measured comparisons, you cannot directly compare Array Loss with measured values. What is typically obtained from measurements are PCS output, energy at the point of interconnection, energy sold, solar irradiance, ambient temperature, and so on. If the array output is not measured directly, it becomes difficult to decompose each component of Array Loss from measurements. If you have a pyranometer and thermometer, PCS DC data, and string monitoring data, you can perform more detailed verification; but if those data are not available, the result will be an estimate.

When comparing PVSyst with measured data, it is important to first compare the same period. PVSyst often uses standard-year or typical-year meteorological data, which do not match the weather of the measured year. If the measured year experienced persistent cloudiness, heavy snowfall, frequent output curtailments, or shutdown periods, a simple comparison with PVSyst’s annual values becomes of little meaning. Measures such as correcting with the measured solar radiation, comparing on a monthly basis, or excluding shutdowns and output curtailments are necessary.

Among Array Loss items, temperature loss is one that is easy to compare if measured ambient temperature or module temperature is available. On days with high solar irradiance and high ambient temperatures, if output is lower than expected, temperature loss may be larger than assumed. Conversely, output differences during cloudy conditions or at dawn and dusk can be related to low-irradiance characteristics, the IAM, or shading effects. If there is a localized drop in the output of a string, suspect mismatch, soiling, faults, shading, or poor connections.

Also, when using Array Loss in explanatory materials, simply listing technical terms can be hard to understand. For internal teams and customers, it becomes clearer to explain it as "the loss from the solar irradiance entering the module to the DC power delivered to the PCS." On top of that, if you explain in order that "the main components are temperature, angle of incidence, low irradiance, shading, soiling, mismatch, and DC wiring," it becomes easier to share where power generation is being reduced.

PVSyst's Array Loss is not simply a number to judge the quality of a design. Rather, it serves as a diagnostic metric to verify how well the simulation reflects site conditions and which assumptions affect the energy yield. When comparing PR or measured results, examining the breakdown of Array Loss greatly improves the ability to explain differences in energy production.

Summary

PVSyst's Array Loss is an important indicator for identifying where power is being lost between a PV array receiving solar irradiance and producing DC power, and the point at which that power is delivered to the PCS. When generation is low, PR is low, or there are discrepancies with other reports, carefully reviewing the Array Loss makes it easier to determine whether the cause lies with the modules, the site conditions, or the wiring design.

The first step in reading is to understand that Array Loss is a loss on the DC side and not to confuse it with AC-side losses, transformer losses, or grid-side constraints. Next, check the temperature loss. Temperature loss varies with region, weather data, the racking’s ventilation conditions, and the module temperature coefficient, and it has a large impact on the annual PR.

By examining IAM, low-light performance, and module characteristics, you can understand the difference between catalog output and field output. Modules do not always operate under standard test conditions, and their output changes with conditions such as angled incidence, cloudy skies, morning and evening, and high temperatures; these small losses can become a non-negligible difference over the course of a year.

When assessing site conditions, mismatch, shading, and soiling are important. These factors are closely related to the design drawings and the on-site environment, and they are aspects that tend to differ from other companies' simulations. How realistically shading and soiling are included changes how the expected power generation is perceived. Even if the predicted power generation is low, if that outcome appropriately incorporates site risks, it can be considered a simulation with greater explanatory power.

DC wiring losses are a verification item directly tied to design changes. If PCS placement, combiner box placement, cable length, cable size, or string configuration change, DC losses will also change. Confirming that PVSyst settings match the latest electrical design increases the reliability of the report.

Finally, Array Loss is particularly useful when used for PR comparisons and measured-data comparisons. However, rather than drawing conclusions based only on the final PR or annual energy production, it is important to track the flow of losses step by step. By clarifying what within Array Loss is having an effect, how losses are split between the AC side and the PCS side, and to what extent these can be verified with measured data, explanations for differences in energy production become more concrete.

The Array Loss in PVSyst should not be viewed as a single number but read as the cumulative result of multiple factors: temperature, IAM, low irradiance, shading, soiling, mismatch, and DC wiring. By tracing, in sequence, where losses occur, PVSyst’s results become not just an energy production table but a practical document for verifying design conditions and site conditions.

In designing and reviewing solar power plants, it is important not only to rely on PVSyst analysis results but also to accurately understand on-site positional information, drawings, and construction status. By combining a system that uses iPhone and GNSS to acquire high-precision site positions—such as LRTK—and can cross-reference those positions with drawings and point clouds, it becomes easier to check racking layout, PCS locations, cable routes, and nearby objects that are likely to cause shading. Reading numerical losses in PVSyst and confirming actual conditions on site with high-precision positional information can strengthen the connection between power generation simulation and construction and maintenance.