Causes of Lower-than-Expected Power Output and 7 Checks to Do Before Making Improvements

When power generation is lower than expected, it’s easy to immediately suspect equipment failure or poor installation. However, solar power generation is affected by weather, season, solar irradiance, temperature, shading, soiling, system conditions, and measurement methods, so you cannot simply conclude that “less means abnormal.” In practice, before narrowing down the cause of low generation, it is important to verify whether the comparison conditions are aligned, whether the data you are looking at is correct, and whether the order of checks to perform on site is organized.

This article outlines seven checks that operations personnel investigating "low power generation" should perform before undertaking improvement work or deciding on repairs. The aim is to calmly assess the likelihood of an anomaly by combining on-site inspections with data verification, without jumping to conclusions about a drop in power generation.

Table of Contents

• If power generation is lower than expected, first ensure that the comparison conditions are consistent.

• Check 1: Are solar radiation and weather variations included in the assumptions?

• Check 2: Are you overlooking seasonal variations and temperature effects?

• Check 3: Is there any degradation caused by shadows or surrounding obstructions?

• Check 4: Have you checked the panel surface for dirt or deposits?

• Check 5: Have you checked the records of equipment shutdowns or output restrictions?

• Check 6 Are there any discrepancies in measured values or aggregation methods

• Check 7: Have you retained on-site records and comparison materials from before the improvement?

• Organize the causes of low power generation before proceeding to decisions about improvements.

When power output is lower than expected, first ensure the comparison conditions are consistent

When you feel that power generation is lower than expected, the first thing to do is not to suspect the equipment but to align the comparison conditions. Solar power generation output can vary greatly from day to day even with the same installation. It is natural for output to differ if conditions differ — for example, between a clear day and a cloudy day, summer and winter, morning and afternoon, times when shadows are more likely, or the presence or absence of snow or yellow dust. When comparing expected generation with actual generation, your judgment will also change depending on which period you target, whether you take solar irradiance into account, and whether you exclude equipment downtime.

In practice, saying that power generation is low is not enough to identify the cause. You need to distinguish whether it is low on a daily basis, a monthly basis, only during specific hours, or only in specific sections. For example, if the entire plant is uniformly low, you would suspect weather or solar irradiance conditions, output limitations, or measurement conditions. On the other hand, if only a particular circuit or a particular string/row is low, you would proceed to check local factors such as shading, soiling, connection conditions, equipment outages, or panel condition.

It is also important to understand under which assumptions the expected power generation itself was produced. Design-stage simulations are generally created based on assumptions such as past meteorological data, installation tilt, orientation, assumed losses, and equipment specifications. Therefore, if the actual local weather or surrounding environment differs from those assumptions, the power generation may appear lower than expected. Rather than treating the expected value as an absolute standard, it is important to view it as a reference for comparison with actual results.

When investigating the cause of low power generation, first clarify “compared to what.” Whether it is compared to the same month of the previous year, to nearby facilities, to the design value, or to the forecast shown on the monitoring screen will change the points you need to check. If you begin remediation work while the comparison target is ambiguous, you may perform unnecessary cleaning or inspections, or conversely overlook abnormalities that truly need to be checked.

Therefore, when power generation is lower than expected, it is safer to organize generation performance, the relevant period, weather, solar irradiance, outage history, equipment capacity, and measurement locations before proceeding to identify the cause. By establishing the basis for judgment before making improvements, it becomes easier to prioritize on-site responses and to explain the situation to stakeholders.

Check 1: Are solar radiation and weather variations included in the assumptions?

The first things to check when generation is low are solar irradiance and weather variations. Because solar power generation produces electricity by receiving sunlight, if sunlight is weakened by clouds, rain, fog, haze, snowfall, smoke haze, yellow dust, or similar conditions, generation will also decrease. This does not immediately indicate a malfunction; it is a fluctuation that can occur due to the way generation works. In particular, on days when the sky looks bright but thin clouds are spread, generation can be lower than under clear-sky conditions. It is important not to judge by the visible weather alone, but to compare solar irradiance and weather records with the generation output.

A common situation in practice is directly comparing generation from past clear, sunny days or the values assumed at the design stage with actuals from periods that were mostly cloudy. Even if monthly generation is lower than expected, if that month had many rainy or cloudy days, the cause may be weather conditions rather than an equipment problem. Conversely, if solar irradiance was sufficient yet only generation was low, it is necessary to proceed to check the equipment or the measurement system. If this distinction is not made, the order of investigating causes becomes confused.

When checking solar irradiance, we examine a combination of values from pyranometers installed at the plant, meteorological data from surrounding areas, and local weather notes. Even when using an on-site pyranometer, readings can be skewed by sensor contamination, installation angle, or shadows from nearby objects. It is important not to treat pyranometer values as absolute; cross-check them against the plant’s power generation curve and on-site conditions. At sites without a pyranometer, nearby meteorological data may be used as a reference, but it may not fully reflect localized clouds, mountain shadows, or weather differences caused by sea breezes.

When looking at the relationship between power generation and solar irradiance, it's easier to judge if you consider not only the absolute generated power but also how much is being generated relative to the irradiance. It's natural for generation to be low on days with low irradiance, but if generation does not increase on days with high irradiance, other factors may be suspected. By checking daily generation curves and seeing whether the ramp-up on clear days, the midday peak, and the evening decline are not severely distorted, it's easier to distinguish between weather-driven variations and anomalies.

Rather than looking only at days when you feel power generation is low, check the several days before and after and past data from the same season as well to reduce bias in your assessment. A decrease confined to a single day can often be explained by weather or temporary controls, but if output is consistently low even on sunny days, the priority for an on-site inspection increases. Before making improvements, confirming whether output is low even after accounting for weather differences is the first major checkpoint.

Check 2: Are you overlooking seasonal variations and temperature effects?

When actual power generation is lower than expected, overlooking seasonal variations and temperature effects can cause normal changes to be mistaken for anomalies. Solar power generation changes seasonally due to the sun’s altitude, hours of sunlight, ambient temperature, snowfall, and changes in the surrounding environment. In summer, longer daylight hours tend to increase generation, but high temperatures can raise panel temperature and reduce generation efficiency. In winter, lower temperatures can sometimes be advantageous for efficiency, but shorter daylight hours and a lower solar altitude make the system more susceptible to shading.

What practitioners should be careful about is not to simply assume “more in summer” or “less in winter.” Power generation varies depending on the region and installation conditions. In snowy areas, output can drop significantly in winter, and in mountainous areas there are times when long shadows occur in the morning and evening. In coastal areas and around farmland, surfaces can become more prone to deposits such as salt, dust, pollen, and farm-related particulates depending on the season. Seasonal variations need to be considered not only in terms of sunlight hours but also in conjunction with local environmental conditions.

Air temperature also plays an important role. Solar panels tend to generate more power when solar irradiance is strong, but their output tends to decrease as panel temperature rises. For that reason, the peak on a clear midsummer day may not reach expectations. This does not necessarily indicate equipment failure. Under conditions with light wind and increased heating of the panels from reflected heat off the roof or ground, the peak of the generation curve can be lower than expected. When assessing generation, it is effective to check not only solar irradiance but also air temperature and the installation environment.

On the other hand, there are decreases that cannot be explained by seasonal variation alone. If power generation has fallen significantly compared with the same month of the previous year despite similar weather, or if the shape of the generation curve has clearly changed on sunny days with similar conditions, it is necessary to check for shading, soiling, equipment shutdowns, circuit abnormalities, measurement faults, and the like. Seasonal factors can serve as material to explain a drop in generation, but attributing everything to the seasons can delay the detection of anomalies.

To analyze seasonal variations, it's easier to understand the actual situation by comparing not only monthly generation totals but also typical generation curves for clear-sky days. Comparing past data from the same month, the same weekday or the same weather conditions, and days with similar temperatures makes it easier to detect changes in equipment condition. Before implementing improvements, it is important to organize not only the finding that generation is low but also whether it falls within a reasonable seasonal range or is abnormal even after accounting for temperature effects.

Check 3: Is there any degradation caused by shadows or surrounding obstacles?

One cause of low power generation that is easy to overlook on-site is the effect of shadows. In solar power generation, even a partial shadow on a panel can affect output depending on the circuit configuration and equipment control conditions. Causes of shading include buildings, trees, utility poles, fences, adjacent equipment, mountains, slopes, residual snow, temporary materials, weeds, and other factors. In particular, sites that had no issues at installation can later experience shading due to tree growth, nearby construction, or changes in material storage locations.

When checking for shadows, visiting the site just once may be insufficient. Because shadow positions change with the time of day and season, a site that appears problem-free at midday may be heavily shaded in the morning or evening. In winter, when the sun’s altitude is lower, shadows that didn’t reach the rows of panels in summer can fall on them. If you can identify time periods with reduced power generation, it is effective to check the shadow conditions during those times. If the generation curve shows a drop only in the morning, only in the afternoon, or a sudden change at a specific time, it is worth suspecting shadow effects.

Also, the effects of shading are not necessarily uniform across the entire plant. If only certain rows, certain sections, or certain circuits have reduced output, focus on checking obstacles and the terrain around those areas. For example, if only the rows beside a fence are low, only the rows below a slope are low, or only a section near trees is low, this pattern may indicate an influence from the external environment rather than an equipment failure. When monitoring data allows comparison by section or circuit, it is important to examine local differences as well as the overall average.

Weeds can also cause shading. At ground-mounted solar installations, grass that grows in front of the panels can cast shadows on the lower edges and lead to reduced power generation. Even grass that appears insignificant in on-site photos can, depending on the sun’s angle, cast longer shadows and have a large impact. Power output may recover after weeding, but care must be taken during weed removal to avoid damaging wiring, racking, and panels. Keeping before-and-after photos and generation data makes it easier to explain the impact of shading.

When checking for shadows, record not only the object causing them but also the time of day, the area affected, and the extent of the shading. If you record photos, compass direction, time, the affected plot, and the weather together, you can later cross-check them with power generation data. Even when considering corrective measures such as felling, pruning, moving materials, or weeding, proceeding after recording the shadow conditions makes it easier to verify the effectiveness of the work.

Check 4: Have you checked the panel surface for dirt and deposits?

When power output is lower than expected, dirt and deposits on the panel surface should also be inspected. Because solar panels are installed outdoors, they can become coated with dust, yellow sand, pollen, bird droppings, fallen leaves, mud splashes, salt, volcanic ash, or lingering snow. If the soiling is temporary and can be washed away by rain, it may not be a major issue, but soiling that remains for long periods or localized buildup can lead to reduced power output.

However, the presence of dirt does not necessarily mean that immediate cleaning is appropriate. Cleaning requires consideration of work safety, the risk of electric shock, the risk of falling, scratches to the panel surface, the handling of wash water, and impacts on the surrounding environment. Inspections on roofs or at heights and work around electrical equipment should be carried out by personnel who have appropriate knowledge and safety measures in place. Even when dirt is suspected as the cause of reduced power generation, it is safer to first perform a visual inspection, take photographic records, and cross-check with generation data to determine whether cleaning is necessary.

The effects of soiling appear differently depending on whether it is thinly deposited across the entire surface or heavily deposited in localized areas. Widespread dust or yellow sand can gradually reduce the power output of the entire plant. In contrast, localized obstructions such as bird droppings or fallen leaves can affect specific panels or circuits. When monitoring data shows differences between sections, check whether the locations of the soiling match the areas with reduced output. Distinguishing whether the drop is system-wide or limited to certain parts also makes it easier to rule out causes other than soiling.

When checking panel soiling, it's easier to assess if you look not only at days with low power output but also at the periods before and after rain. If there is a tendency for power output to improve after rain, surface soiling may have been a factor. However, because weather conditions may have changed after the rain, simple comparisons should be avoided. It's important to compare sunny days and observe changes under conditions with similar irradiance and temperature.

Also inspect not only the panel surfaces but the drainage and the condition of the surrounding ground. If the ground is prone to mud splashing, the surrounding environment tends to accumulate fallen leaves, there are structures that attract birds, or there are nearby unpaved roads, soiling can recur. Even if you clean once, if the same causes remain, a reduction in power generation may occur again. Before implementing improvements, confirming not only the soiling itself but also the environmental factors that cause it will make it easier to consider measures to prevent recurrence.

Check 5 Have you checked the records of equipment shutdowns and output restrictions?

When investigating the causes of low power output, check the equipment shutdown history and whether any output restrictions are in place. In solar power generation systems, the condition of devices such as the power conditioner, connection boxes, collection equipment, protective devices, and communication equipment can make the power output appear low. If equipment has completely stopped, abnormalities are easy to detect; however, if only some equipment has stopped or short intermittent stoppages are occurring, you may first notice the problem as a decline in monthly power generation.

First, you should check whether the monitoring screens or records show any errors, alarms, stoppages, recoveries, communication interruptions, control states, etc. If stoppages or alarms on specific equipment coincided with the period of low power generation, prioritize those as candidate causes. However, the absence of displayed alarms does not necessarily mean there is no abnormality. Communications may have been interrupted and records not acquired, or minor state changes may appear as reduced power output. When reviewing records, it is important to look at power output, equipment status, and communication status together.

Output restrictions and grid-side conditions should not be overlooked. Even when generation equipment is capable of producing power normally, output may be curtailed due to grid-side circumstances or for equipment protection. In such cases, a lower-than-expected generation output does not necessarily indicate equipment failure. On the generation curve, this can appear as a plateau at a constant output despite clear weather, curtailment only during specific time periods, or similar trends across multiple facilities. In these situations, check notifications of output restrictions, operation logs, monitoring data, and voltage conditions.

Power conditioners may also have their output reduced due to temperature rise or poor ventilation. If the installation site has poor airflow, objects are placed around the unit, dust has accumulated on filters or intake and exhaust vents, or the unit is exposed to direct sunlight or a high-temperature environment, the output may be lowered to protect the equipment. During on-site inspections, check the temperature environment around the equipment, ventilation space, any unusual noises, display status, and warning lights. However, internal inspection of electrical equipment or recovery operations involve danger and must be performed only by personnel with the appropriate qualifications and authority.

If only some circuits or devices are offline, the drop in total power generation can appear small. However, if left unaddressed the losses will accumulate, so comparisons by section and by device are important. Do not judge only by the plant’s overall total; check whether the power output of a specific device is low relative to its installed capacity, and whether there are differences compared with devices under the same conditions. Before taking corrective action, rather than assuming a failure, it is important to narrow down the targets by sequentially checking shutdown history, control history, communication status, and device-level power generation.

Check 6 Are there any discrepancies in measured values or aggregation methods?

When power generation is lower than expected, it may not be that actual generation is low but that there are discrepancies in measurements or aggregation methods. The display on monitoring devices, the value on the feed-in meter, generation per power conditioner, and aggregated values in daily and monthly reports may not match. If you make a judgment without confirming which value is being used as the reference, you may reach a conclusion that differs from reality.

First, verify what the figures you are looking at as "generated power" actually represent. Whether it is instantaneous output, daily generation, monthly generation, the AC-side value, an estimated DC-side value, the amount sold to the grid, or the surplus after self-consumption changes the meaning. In systems that include self-consumption in particular, generated power and sold power may not match. If part of the generated electricity is used within the facility, looking only at the amount sold can make generation appear lower. Before judging a decline in generation, you need to align the definitions of the figures being checked.

Also check for shifts in the aggregation period. Even if the monitoring screen aggregates by calendar date, meter readings may be aggregated using a different date boundary. The time at the start or end of the month, communication delays, data gaps, or transcription errors from manual input can make monthly generation look lower. When comparing daily reports, monthly reports, monitoring device records, and billing-related records, confirm the target period and the time cutoffs. Even a one-day shift can change the figures depending on whether it includes a day with significantly different weather.

Be aware of data loss caused by communication failures. Even if the power generation displayed on the monitoring device is low, the system may actually have been generating power but the communication was interrupted and the data not recorded. If you do not check the duration of the communication outage, whether data interpolation was performed, and how the recovery was reflected, you may mistakenly interpret it as a drop in power generation. If the monitoring data contains unnatural gaps, flat lines, or sudden zero readings, suspect not only a fault in the generation equipment but also issues on the communication or recording side.

Also pay attention to how equipment capacity is handled. If there has been an expansion, equipment replacement, circuit change, panel removal, or change in measurement points, the conditions for comparison with past data will change. If you compare using the same criteria even though the equipment capacity has changed, you may mistakenly conclude that power generation is lower or higher. Before making improvements, confirm that the equipment register, single-line wiring diagram, equipment configuration, measurement points, and monitoring settings match the current state.

Checking measured values and aggregation methods may be unremarkable, but they have a major impact on the accuracy of root-cause investigations. Before starting work on site, confirming the source of the data, the period covered, the units, the aggregation conditions, and the communication status can reduce unnecessary on-site interventions. When addressing the issue of low power generation, it is important to proceed on the premise that the cause may lie not only in the equipment itself but also in how the numbers are interpreted.

Check 7: Have on-site records and comparison materials been retained prior to improvement?

When you identify the cause of lower-than-expected power generation, it is important to leave on-site records and comparison materials before beginning corrective work. When you find something that appears to be the cause, you will be tempted to immediately carry out cleaning, weeding, equipment recovery, settings checks, and component replacement. However, if you have not recorded the pre-improvement condition, it becomes difficult to determine later whether the work actually had an effect. Because improvements in power generation are also influenced by the weather, even if generation increases after the work, you may not be able to distinguish whether that was the result of the corrective work or simply better weather.

In on-site records, keep as complete a set as possible of photos, the time the photos were taken, the direction, the target section, the weather, the inspector, and the inspection details. If verifying shadows, photos that show the area shaded and the time are required. If verifying soiling, keep photos that show the type of soiling, its extent, the panel row, and the condition before and after cleaning. For weeds or obstructions, photograph so that the positional relationship to the panels is clear. When checking device displays or alarms, record not only the displayed content but also information that identifies the device.

As comparative materials, it is easier to make a judgment if you align the pre- and post-improvement power generation data, solar irradiance, weather, temperature, outage history, and work dates/times. When comparing power generation before and after the work, it is desirable to choose sunny days with conditions as similar as possible. Comparing days with very different irradiance conditions can lead to overestimating or underestimating the effect of the improvement. To explain the effect of the work, it is helpful to look not only at the energy output but also at the shape of the generation curve and the differences among the affected sections.

Records from before the improvements are also useful for reporting to stakeholders. When explaining the causes of low power generation, there can be discrepancies in understanding if only verbal explanations are used. By presenting on-site photos together with data, you can share concrete information on the effects of shadows, the extent of soiling, the timing of equipment shutdowns, and discrepancies in measurements. When multiple stakeholders are involved—such as the plant owner, managers, construction personnel, and maintenance personnel—having common materials makes decision-making about responses smoother.

Recording conditions before making improvements also helps prevent recurrence. For example, trends such as shadows appearing at the same time every year, increased soiling in certain seasons, mud splashing after rain, and equipment temperatures tending to rise in summer become apparent when records are accumulated. Rather than treating an issue as a one-off, reflecting it in the next inspection and in the annual maintenance plan can shift responses to declines in power generation from reactive to preventive.

When power generation is low, corrective work should be carried out safely in the sequence of hypothesizing the cause, conducting an on-site inspection, recording the findings, performing the work, and confirming the effect. If work is carried out without keeping records, the cause remains ambiguous, and when the same problem recurs you will have to investigate again from the beginning. Although keeping records before making improvements may seem like a hassle, it is an important step for improving management quality in the long term.

Clarify the causes of low power output before deciding on improvements

When power generation is lower than expected, multiple causes may be involved, such as faults, soiling, shading, weather, seasonal effects, output limits, and measurement errors. Therefore, it is important not to draw conclusions based on a single factor but to check them in sequence. First, ensure comparable conditions, check solar irradiance and weather differences, and take seasonal variations and temperature effects into account. Then, by examining shading, soiling, equipment stoppages, output limits, and measurement discrepancies, you can systematically narrow down the cause.

In practice, the low power output is often the first thing that stands out. However, decreases in generation can be either normal fluctuations or abnormalities that warrant prompt action. If a decline that can be explained by weather or seasonal factors is judged to be a fault, unnecessary interventions will increase. Conversely, if equipment stoppages or shading effects are overlooked and blamed on the weather, losses may continue. That is why it is important to take an approach of combining data with on-site conditions and making judgments through the 7 pre-improvement checks.

It is especially important not to treat power generation, solar irradiance, onsite photos, equipment condition, and work records separately. The power generation numbers alone can make the cause hard to see, and onsite photos alone can make it difficult to judge their impact on generation. By organizing multiple pieces of information in the same chronological order, it becomes easier to understand when, where, and what happened. This also makes it easier to prioritize improvement work.

Also, safety must not be overlooked when making improvement decisions. Photovoltaic (PV) systems are electrical equipment and may involve working at heights and outdoor work. Even when checking the causes of low power output, approaching too closely, performing unwarranted internal inspections of equipment, touching live parts, or photographing from unstable footing should be avoided. On-site inspections should be carried out only after the scope of work and authority have been clearly defined and by safe methods.

To continuously manage the causes of low power generation, it is essential to record information obtained on site and keep it organized so it can be used for subsequent inspections and improvement decisions. Organizing photos, location information, timestamps, inspection notes, and generation data makes it easier to track trends in shading and soiling, changes in terrain and the surrounding environment, and differences between individual assets. Rather than treating a drop in generation as a one-off incident, the ideal is to detect it early within operations management and maintain a condition in which it can be explained.

If you feel the power generation is lower than expected, rather than rushing into corrective work, first try organizing potential causes according to the 7 checks presented here. If you can link the data with on-site conditions and verify them, you can reduce unnecessary work and more easily determine the necessary actions. If you want to more efficiently grasp the condition of a solar power plant and manage inspection records, it is effective to consolidate generation data, on-site photos, and inspection notes, and to establish an operational system that allows stakeholders to review them.