【How to Set Soiling Losses in PVSyst｜Four Practical Approaches for Real-World Use】

Table of Contents

• What are soiling losses in PVSyst?

• Prerequisites to clarify before setting soiling losses

• Basic steps to set soiling losses in PVSyst

• Four practical approaches used in the field

• Points to note when deciding input values for soiling losses

• Seasonal variations and site conditions to check in monthly settings

• How to check soiling losses in simulation results

• On-site checks to avoid underestimating soiling losses

• Summary

What is soiling loss in PVSyst?

When simulating the energy production of a photovoltaic system in PVSyst, soiling loss is one of the most important input parameters. Soiling loss refers to the effect in which solar radiation reaching the PV module surface is partially blocked by sand and dust, pollen, yellow sand, bird droppings, fallen leaves, exhaust-derived particulates, salt, mud splashes, etc., adhering to the module surface, causing a reduction in energy production.

The power output of solar photovoltaic systems is affected by various conditions such as solar irradiance, module capacity, temperature, azimuth, tilt angle, shading, wiring, and conversion efficiency. Among these, soiling losses are a factor that varies greatly depending on the site environment and maintenance policy. At the design stage, even with the same modules, the same capacity, and the same tilt angle, the expected soiling patterns differ between mountainous areas, areas near farmland, industrial zones, coastal areas, arid regions, snowy regions, and urban areas.

In PVSyst you can enter soiling losses as a constant annual value or vary them month by month. In practice, it is important not only to state “what percentage to use for soiling loss” but also to be able to explain whether that figure is reasonable given site conditions, cleaning schedule, rainfall frequency, module tilt, surrounding environment, and the expectations of the recipient.

Soiling loss may seem like a small number, but its impact on project finances and generation forecasts cannot be ignored. For example, if soiling loss is underestimated by 1% relative to the assumed annual generation, the annual generation will be correspondingly overstated. The larger the plant, the more that 1% difference affects annual energy generation, power sales revenue, performance evaluation, and maintenance planning.

On the other hand, if soiling losses are overestimated, the design’s expected energy production may be underestimated, which can put you at a disadvantage in project evaluation and investment decisions. Therefore, when setting soiling losses in PVSyst, it is important to use assumptions that reflect the actual site conditions, neither to make the results look conservative nor optimistic.

Prerequisites to Clarify Before Setting Soiling Losses

Before entering soiling losses in PVSyst, the first thing to clarify is the installation environment of the photovoltaic system in question. Soiling losses depend more on the external environment of the installation site and on operational practices than on the modules' intrinsic performance. Therefore, verbalizing the site conditions before opening the input screen makes it easier to justify the numerical settings.

The first thing to check is whether there are sources of dust in the surrounding area. If sites such as land under development, unpaved roads, quarries, farmland, factories, logistics facilities, busy roads, or port areas are nearby, airborne particles are more likely to adhere to module surfaces. In particular, during dry periods and in windy regions, dirt tends to accumulate during periods with little rain.

Next, check the tilt angle of the modules. Steeply tilted modules tend to have surface dirt washed off more easily by rain. Conversely, on low-tilt modules, rain is less likely to run off the surface, and mud and dust tend to accumulate at the lower edge. On roofs and low-pitch mounting structures, this difference can affect power generation.

The frequency of rainfall is also important. Rain does not completely remove dirt, but it does have a certain cleaning effect. In regions with regular rainfall, the risk of dirt accumulating over long periods is relatively low. On the other hand, in regions that experience prolonged dry seasons, or during the period before the rainy season, when yellow sand is likely to arrive, or when pollen levels are high, it is worth considering monthly soiling loss.

Be sure to also clarify whether a cleaning plan exists. Even within the same region, soiling loss settings differ between plants that assume regular cleaning and those that are essentially operated relying on rain. Even when cleaning is performed, the items that should be reflected in the monthly loss settings change depending on how many times per year it is carried out, when it is carried out, and whether it is a full cleaning or a localized cleaning.

Additionally, take into account the purpose of the power plant and the intended recipient. Whether it is an initial simulation for internal review, for investment decision-making, a report to be submitted to financial institutions or the client, or material for equipment certification or design comparison, the level of accountability for the figures changes. In practice, being able to explain why you adopted those input values for PVSyst is more important than the PVSyst input values themselves.

Basic procedure for setting soiling losses in PVSyst

The procedure for setting soiling losses in PVSyst is: after creating the basic conditions of the photovoltaic system, open the soiling input screen from the detailed losses settings, and set the annual or monthly loss rates. The screen layout may vary slightly depending on the version, but the underlying concept is the same.

First, open the target project and set the basic conditions such as system configuration, meteorological data, azimuth, tilt angle, and array configuration. Because soiling losses are treated as an additional loss condition on power generation, entering them while the basic conditions are still unorganized makes it difficult to assess the overall validity. It is more practical to first firm up the installed capacity, installation azimuth, tilt, the presence or absence of shading, and the electrical configuration to some extent before setting soiling losses.

Next, look for the soiling loss item on the loss settings screen. Soiling loss is represented by the idea that the irradiance reaching the module surface is reduced by soiling. The value you enter is typically expressed as a percentage. For example, if you assume a 2% soiling loss throughout the year, enter 2% as a constant annual loss.

For settings that are closer to actual practice, enter loss rates by month. With monthly input you can distinguish months that tend to be dry, months heavily affected by pollen or yellow sand, months with heavy rainfall, months after cleaning, and so on. For example, setting lower values for rainy periods, higher values for prolonged dry periods, and lower values immediately after cleaning will produce a simulation that more closely matches on-site conditions than using a single constant annual value.

After entering the data, always run the simulation and check how much the soiling loss is reflected in the loss diagram and the results report. In PVSyst, by viewing the breakdown of losses you can confirm how much impact the soiling loss has on the overall energy production. At this point, check whether it is unnaturally large compared to other loss items, or conversely too small for the site conditions.

Finally, we recommend keeping a note of the rationale behind the settings. The PVSyst configuration file alone can make it difficult to understand later why a particular soiling loss was adopted. Briefly recording site conditions, the assumed cleaning frequency, the surrounding environment, and the reasoning behind any month-by-month settings will be helpful for internal reviews, customer explanations, and comparisons when design changes are made.

Four Ways of Thinking for Practical Work

When setting soiling losses in PVSyst, it is easier in practice to organize the approach by distinguishing four main strategies. The first is to assume a standard constant annual value. This is often used for initial studies and rough simulations. When detailed site information is still limited, making overly detailed monthly settings tends to have weak justification. Therefore, for common installation environments it is realistic to first assume a constant annual soiling loss and use that to compare design options.

The advantage of using a fixed annual value is that it makes comparing conditions easier. When you want to compare tilt angle, azimuth, overloading ratio, string configuration, shading effects, and so on, fixing soiling loss as a common condition makes differences due to design parameters easier to see. Especially in the early stages, differences caused by layout and capacity design are often larger than small differences in soiling loss, so using a standard value to grasp the overall picture is effective.

The second is the idea of varying by month. This is effective for projects where site conditions and maintenance plans are somewhat known. Dirt does not accumulate uniformly throughout the year. During prolonged dry periods, windy periods, high-pollen seasons, times affected by agricultural work or land development, and periods of heavy rainfall, the condition of the module surfaces changes. Using month-by-month settings can reflect these seasonal variations.

In monthly settings, it's important not merely to enter different numbers for each month but to create a year-round narrative. For example, you might set higher impacts for pollen and yellow sand in early spring; lower values during the rainy season to account for the washing effect of rainfall; consider that dirt may be washed away more easily by typhoons and rain from summer through autumn; and increase values again in regions where dry conditions persist in winter. If cleaning is performed, you can also set losses to be lower immediately after the scheduled cleaning month and then gradually increase them afterward.

The third is an approach that works backwards from the maintenance policy. For solar power generation equipment, the extent of cleaning is directly linked to soiling losses. If regular cleaning is assumed, the risk of continuous dirt accumulation can be suppressed. Conversely, if cleaning is not performed and natural washing by rain is assumed, it is necessary to carefully anticipate dirt accumulation depending on the region and season.

From a maintenance policy perspective, balancing cleaning costs against the power recovery effect is also important. At sites where soiling losses are expected to be large, the amount of generation that can be recovered by cleaning may be significant. However, cleaning involves labor costs, travel, work time, safety measures, securing water, and other factors. PVSyst does not directly evaluate cleaning costs themselves, but by comparing simulations with and without cleaning, you can create material to inform maintenance policy decisions.

The fourth is a conservative way of thinking that accounts for risks. In submitted power generation forecasts and long-term financial projections, applying overly optimistic soiling losses can cause actual generation to fall short of forecasts. Especially at sites with nearby sources of soiling, low-slope roof installations, dry regions, locations where bird damage is a concern, or power plants near farmland or unpaved land, you should be more cautious than the standard values.

However, taking a conservative view and entering large values without justification are different. If you increase soiling losses, you need explainable reasons such as the surrounding environment, weather, tilt, cleaning frequency, and past similar projects. In practice, it can be effective to compare multiple scenarios—such as optimistic, standard, and conservative cases—and decide the final adopted value. At that time, creating multiple variants in PVSyst makes it easier to check how differences in soiling losses affect annual energy production.

Points to consider when determining input values for soiling losses

When deciding on input values for soiling losses, the thing you most want to avoid is reusing fixed values without justification across all projects. Of course, having internal standards and experience from past projects is useful. However, mechanically using the same numbers can cause you to overlook differences between sites. When using PVSyst, the important point is not the numeric entry itself but the judgment to adjust input values to the site conditions.

For example, in locations where the slope is sufficiently steep, there are no major dust sources nearby, and periodic rainfall can be expected, soiling losses may be set relatively low. Conversely, in low-slope areas with an unpaved road nearby and dry seasons prone to blowing dust, you should assume higher soiling losses even within the same region. This is especially true for ground-mounted installations, where exposed ground after site development, nearby agricultural activities, and dust from vehicle traffic are more likely to have an impact.

Localized soiling such as bird droppings and fallen leaves also requires attention. PVSyst’s soiling loss is basically treated as an average loss over the entire module surface. However, on actual sites, bird droppings may adhere only to specific modules, or fallen leaves and tree sap may affect modules located near trees. Such localized soiling can be difficult to represent with a simple uniform loss. Together with the effects of shading and mismatch, it should be considered a separate risk.

It is also important not to confuse soiling losses with shading losses. Soiling losses are losses in which the transmission of solar irradiance is hindered by dirt deposited on the module surface. On the other hand, shading losses are losses in which direct solar irradiance is blocked by buildings, trees, mounting structures, utility poles, or surrounding structures. When bird droppings or fallen leaves cause strong local obstruction, the phenomenon is classified as soiling, but electrically they can have an effect similar to partial shading. In PVSyst it is important to understand the meaning of each loss item and to avoid double-counting or underestimating losses.

When determining input values, it is necessary to clarify whether they should be treated as an annual average loss or as the maximum monthly risk. Entering a constant 3% for the year versus entering 6% in one month and 1% in another not only affects the annual total differently, but also alters the seasonal distribution of power generation. For projects where generation in summer or early spring is important, monthly settings can affect the financial assessment.

Furthermore, the appropriate level of detail also depends on the purpose of the simulation. For initial, rough evaluations, it is more efficient to compare design options using a simple, constant annual value rather than detailed month-by-month settings. Conversely, for final designs and submission reports, monthly settings that reflect site conditions and are consistent with maintenance policies may be required. Practitioners using PVSyst should be mindful of adjusting the granularity of inputs according to the project stage.

Seasonal variations and site conditions to consider in monthly settings

When setting monthly soiling losses, it is important to consider seasonal variations in susceptibility to soiling. In many regions of Japan, early spring is prone to pollen and yellow sand, and during dry periods dust is more likely to adhere. During the rainy season and typhoon periods, rainfall can provide a cleaning effect, but mud splashes and re-deposition from the surrounding environment may also occur. In winter, effects vary by region, including dust from dry conditions, cleaning by snowfall, and soiling after snowmelt.

When setting monthly parameters, consider not only the climate but also activities around the site. Near farmland, tillage, harvesting, soil movement, irrigation, and vehicle traffic on farm roads vary with the seasons. Near construction sites or development areas, dust levels may increase during certain periods. In ports and coastal areas, salt deposition can be a concern depending on wind direction and season. Near factories and roads, dust and exhaust-related soiling may occur continuously.

The timing of cleaning also has a major impact on monthly settings. For example, at sites where dirt tends to accumulate in early spring, if cleaning is planned before the period when power generation increases, the months after cleaning can be set with lower soiling losses. Conversely, if cleaning is not performed and the dry season continues, you might configure losses to increase slightly from month to month.

However, making monthly settings too detailed can actually make explanations more difficult. When there is not enough measured data, you should avoid placing excessive weight on month-by-month figures. In practice, it is easier to explain by setting broad trends—for example, periods that tend to get dirty, periods that recover easily with rain, and periods following cleaning.

Also, when using monthly settings, you need to check the monthly power generation in the simulation results. If you only look at annual power generation, the effect of monthly soiling losses can be hard to see. In particular, setting high soiling losses during seasons with high generation will have a large impact on the annual power generation. Conversely, even if soiling losses are high in seasons with low generation, their impact on the annual total may be relatively small.

In this way, monthly settings can provide expressions closer to practical conditions, but they require a basis for the input values. As a way of using PVSyst, it is better to think of monthly settings not as something that will necessarily increase accuracy, but as an effective method when you have information that can reflect site conditions.

How to check soiling losses in simulation results

When you input soiling losses into PVSyst, verify their impact in the simulation results. Points to check are the annual energy production, the loss diagram, monthly energy production, and the balance with other loss items. Confirm not only that the values you entered are reflected in the results, but also that they do not appear unnatural within the overall picture.

In a loss diagram, the losses from solar irradiance to the final output are displayed stepwise. Soiling loss affects power generation as an upstream loss that occurs before solar irradiance is effectively utilized at the module surface. If this loss is large, it means that the effective irradiance reaching the module is reduced in the first place, prior to later losses such as temperature loss, wiring loss, and conversion loss.

When comparing annual energy production, it becomes easier to understand if you create multiple cases with different soiling losses. For example, by comparing three conditions—low, standard, and high soiling loss—you can grasp how much the annual energy production changes. This difference can serve as a basis for decision-making when considering cleaning schedules and maintenance policies.

When examining monthly generation, check how much generation drops in months with high soiling losses. If high soiling losses occur in months with high generation, they can have a large impact on annual generation. Conversely, applying high soiling losses in winter months with low generation may not make much difference to the annual total. For this reason, it is important to know whether the months prone to soiling coincide with the months of high generation.

We also look at the relationship with other loss items. For example, in a project with large shading losses, if soiling losses are also set high, the annual power generation may drop significantly. If those settings reflect the actual site conditions, that’s fine, but it’s necessary to check whether the effects of shading and soiling are being double-counted or assessed too strictly. We also check whether they are being confused with other factors, such as mismatch losses or low-irradiance losses.

In practical reviews, it is important to be able to explain the validity of input values while looking at PVSyst’s output. Rather than only stating “what percentage was set for soiling loss,” summarize “why that value was chosen,” “which elements of the site conditions it reflects,” “whether it is consistent with the cleaning plan,” and “the reason for using monthly settings,” so that it will be easier to respond if conditions change later.

On-site verification to avoid underestimating soiling losses

PVSyst's soiling loss is determined by the input values, but on-site verification is essential to improve the accuracy of those inputs. Desk-based simulations alone may not fully capture surrounding dust, rainwater flow, bird damage, fallen leaves, soil dust, vehicle traffic patterns, ground conditions, and so on. In practice, risks that are immediately apparent on site can be overlooked when relying only on drawings and satellite images.

During the on-site inspection, first observe the area around the planned module installation location. Check whether there is an unpaved road nearby, whether vehicles pass frequently, whether soil is exposed around the site, whether wind easily raises dust or sand, and whether nearby trees or power lines make the site prone to attracting birds. For rooftop installations, also check surrounding vents, equipment, trees, adjacent buildings, and the way rainwater flows.

Next, examine the terrain and how the wind flows. If there is a dust source on the windward side, dirt is more likely to adhere to module surfaces. On sloped sites, rain can cause mud splashes in particular locations and may lead to dirt concentrating on lower-row modules. The effects of rain and wind also vary depending on whether the ground is gravel, soil, grass, or pavement.

In the case of an existing power plant, observing the actual module surfaces is the most straightforward approach. Check whether the soiling is spread uniformly, has accumulated in bands along the lower edge, or is concentrated in specific rows or areas. If the change in power output before and after cleaning can be determined, that provides important information for revising the assumed soiling loss values.

Even for new projects where modules are not yet installed, observing how nearby roofs, vehicles, equipment surfaces, and existing structures are soiled can sometimes allow you to infer tendencies for dust, salt, and bird damage. Because the surrounding environment can change between immediately after site development and after the start of operations, it is also important not to judge solely based on conditions during construction.

High-precision location information and photographic records are extremely useful for streamlining these on-site checks. If you keep geo-tagged records of where potential soiling factors were observed, which direction dust sources lie, and which areas require extra attention for cleaning or inspection, they can be used not only as justification for PVSyst settings but also for post-construction operation and maintenance. Power generation simulation is a desk-based task, but the credibility of its input values is enhanced by on-site observations.

Summary

Setting soiling losses in PVSyst may seem like simply entering a number on the screen. However, in practice it is important to organize the site conditions, cleaning plan, seasonal variations, surrounding environment, and the purpose of submission that lie behind that number. Soiling loss is a factor that affects long-term energy yield forecasts for photovoltaic installations, and whether it is underestimated or overestimated will influence design decisions and financial assessments.

First, check the risks at the site such as dust, pollen, yellow sand, bird damage, fallen leaves, salt, and mud splash, and decide whether to consider them as a constant annual value or to set them on a monthly basis. In initial assessments, compare simply using a constant annual value; for detailed design and submission reports, using monthly settings based on site conditions makes it easier to carry out evaluations that are closer to practical work.

As approaches used in practice, there are methods such as using a standard constant annual value, varying it by month, deriving it from the maintenance policy, or setting it conservatively to allow for risks. Whichever method is used, what is important is being able to explain "why that soiling loss was adopted." In PVSyst outputs, review the loss diagram, annual energy production, monthly energy production, and the balance with other loss items to reassess whether the input values are reasonable for the actual site.

To improve the accuracy of soiling loss estimates, on-site verification is essential, not just desktop settings. Recording the surrounding environment, ground conditions, wind patterns, rainwater runoff, places where birds tend to gather, and the locations of dust sources can give PVSyst input values more credibility. In particular, it is important to consciously reflect on-site observations in the settings for low-slope roofs, areas near unpaved roads, around agricultural land, coastal areas, industrial zones, and around newly developed land.

In the design and maintenance of solar power generation, connecting simulation figures with on-site realities is important for improving the accuracy of power generation forecasts. LRTK, a high-precision GNSS positioning device that can be attached to an iPhone, can be used to record on-site confirmed soiling factors, inspection locations, photo records, and surrounding conditions together with high-precision location information. When you want to make more practical preparations for setting soiling losses in PVSyst, inspection records after the start of operation, or identification of cleaning targets, the ability to accumulate on-site conditions as accurately geolocated data is a major advantage. To improve simulation accuracy, it is important not only what is entered into the software but how correctly on-site information can be reflected. By utilizing LRTK, you can consistently support on-site confirmation, inspection, recording, and reporting at the plant, making power generation assessments in PVSyst and site management more reliable.