5 Ways to Identify the Causes of Low Power Generation Using Thermal Images

When you notice low power generation, the first thing you should check is "where and how the output is dropping." Because power generation can vary greatly depending on solar irradiance, season, weather, system capacity, and the surrounding environment, you cannot determine the cause by looking at the generation numbers alone. However, using thermal images allows you to narrow down locations that may be abnormal by using temperature differences that are difficult to detect with normal visual inspection.

Thermal images are useful for checking the surface temperature of solar panels, heat generation in wiring and connections, the effects of shadows and soiling, and temperature differences between systems. However, a high temperature appearing in an image does not necessarily indicate a fault, since appearance can change depending on shooting conditions and the surrounding environment. What is important is not to judge the image on its own, but to interpret it together with power generation data, on-site conditions, and inspection records.

Please translate the following input into English.

Table of contents

• Prerequisites to check before using thermal images to assess a decrease in power generation

• Search for anomaly candidates in high-temperature regions at the panel level

• Detect circuit-level abnormalities from string- or band-shaped temperature differences

• Observe local temperature changes caused by shadows, dirt, and fallen leaves.

• Check for heat generation around connection points, wiring, and equipment.

• Compare with past images and power generation data to observe changes.

• Do not make definitive judgments based solely on thermal images; use them to inform an inspection plan

Preconditions to check before using thermal images to assess a drop in power generation

When using thermographic images to investigate the causes of low power output, it is important to first standardize the imaging conditions. The temperature distribution of a photovoltaic system varies with solar irradiance, wind, ambient air temperature, cloud movement, time of capture, panel angle, and reflections from surrounding surfaces. If images are taken under unstable conditions, temperature differences that are not actually abnormal can appear to be problems.

For example, during periods when clouds pass frequently, the solar irradiance on the panel surface can change over a short time. Thermal images captured under such conditions tend to show a mixture of areas cooled by cloud cover and areas that had been receiving solar irradiance until just before. If this is mistakenly judged as equipment failure, it could lead to unnecessary inspections or replacement decisions. Before viewing thermal images, you need to confirm whether the solar irradiance was stable at the time of capture, whether the wind was not too strong, and whether the images were not taken immediately after rain or condensation.

Also, judging that power generation is low based solely on a simple comparison with the previous day is insufficient. Even if it looks sunny, output can change due to thin clouds, yellow sand, pollen, humidity, temperature, and differences in solar altitude. In practice, you compare the same system’s past performance, nearby systems on the same day, irradiance data, output by each power conditioner, current values for each string, and so on to determine where the decline is occurring. Thermal images are a valuable piece of evidence for advancing that analysis, but they do not by themselves identify the cause.

It is also necessary to clarify the subject of the photography. Depending on whether you want to broadly inspect the entire plant, look only at a specific string of panels, or re-check an area that previously had abnormalities, the shooting method and the points to look for will change. Wide-area shooting makes it easy to grasp overall trends, but it can be difficult to detect heating at small connection points. Close-up shooting makes it easier to confirm localized abnormalities, but you may miss trends across the entire plant.

When viewing thermographic images, it is also important to combine them with visible images and location information. A thermographic image alone can make it difficult to tell which panel and which position were photographed. This is especially true in industrial facilities, where many similar rows of panels are installed; if the images and on-site locations are mismatched, the wrong inspection target may be selected. When photographing, it is advisable to record the row number, rack number, panel position, orientation, and time of capture so that they can be re-verified on site later.

Safety must also be considered. Even if you find a hot spot in a thermal image, do not touch live equipment or wiring carelessly. Because photovoltaic (solar) installations generate power during the day, some voltage may remain even after shutdown procedures. Verification of heated areas and inspection of connection points must be carried out following the procedures of the facility manager or personnel with electrical expertise. Thermal images should be treated as reference material to safely narrow down inspection targets, not as a means to forcibly touch and check hazardous locations.

Search for anomaly candidates in high-temperature areas at the panel level

One straightforward way to read a thermal image is to check whether any areas on the panel surface are hotter than the surrounding areas. Solar panels do experience some temperature rise while generating power, but panels under the same conditions tend not to show extreme temperature differences. If only a portion of a panel is hotter compared with the surrounding panels, it is worth checking as a potential anomaly that could be causing reduced power output.

If a hot spot appears as a dot, part of the cell may be under load. Localized high temperatures can be caused by several factors, including dirt, bird droppings, fallen leaves, small shadows, cell damage, or internal connection faults. The important thing here is not to immediately conclude that the panel needs to be replaced upon finding a hot spot on a thermal image. If the issue is simply something adhered to the surface, cleaning or improving the surrounding environment may change the condition, and if a temporary shadow is the cause, changing the time of imaging can alter the temperature distribution.

If a high-temperature area is concentrated within a portion of the panel’s cell region, you need to consider what that area signifies for the power-generation circuit. A solar panel generates electricity from multiple cells combined, and if some cells do not function adequately, that part can become hot. The fact that the hot area is small does not necessarily mean its impact is small. Depending on the panel’s internal connection structure, a localized malfunction can affect the output of the entire panel or the whole string.

On the other hand, a panel may appear hotter than its surroundings. In this case, consider the possibility that that panel alone is not producing output or that its electrical condition differs from the other panels. A generating panel converts part of the light energy it receives into electricity. If it is in a state that makes it difficult to extract that electricity, even when receiving the same solar irradiance a larger proportion may appear as heat. Therefore, when a particular panel looks hotter despite receiving the same solar irradiance as its surroundings, it is a prompt to check the wiring, connections, internal circuitry, shading, soiling, and so on.

However, caution is required when interpreting temperature differences. The edges of panels, areas near frames, parts shaded by the mounting structure, and areas receiving reflections can appear to have temperatures different from their surroundings. Depending on the shooting angle, reflections of the sky and nearby objects can affect the thermal image, causing apparent temperature differences that do not reflect the actual temperatures. Checking not only from straight on but also from different angles and distances, and cross-referencing with visible images, can reduce misjudgments.

What on-site personnel should look for is not just whether there are hot spots. It is important to check together the shape and position of the hot spot, the difference from the surroundings, repetition in the same column, and differences from past images. For example, if similar hot patterns appear simultaneously on multiple panels, the cause may be shadows, dirt, installation conditions, or imaging conditions rather than a failure of individual panels. Conversely, if only one panel in the same column shows a clearly different temperature, that can be used as a basis for prioritizing an individual inspection.

Detecting circuit-level faults from string- or band-shaped temperature differences

On thermal images, attention is paid not only to localized hot spots but also to temperature differences that appear as string-like or band-like patterns. Low power generation can be caused not only by a local defect in a single panel but also occur at the circuit level, such as in strings, junction boxes, breakers, or power conditioner input circuits. If power generation data shows that only a particular system has low output, thermal images may also show temperature differences in the rows or areas of panels corresponding to that system.

For example, among a group of panels installed with the same orientation and angle, an entire column may sometimes show higher temperatures. In such cases, consider the possibility that that column is not generating enough power or is electrically isolated. Of course, wind exposure and solar irradiance can also differ between columns, so temperature differences alone cannot be conclusive. However, if panels belonging to the same column consecutively show different temperatures, it is worth following up with a circuit-level inspection.

Banded temperature differences can also occur due to shadows. When narrow shadows from mounting racks, railings, power lines, nearby structures, vegetation, or the like fall on panels, they may appear as banded temperature variations on thermal images. The shaded areas receive less light available for power generation, and under certain conditions can be accompanied by localized temperature increases. Even a shadow that looks slight can have a large impact on output depending on its position in the circuit.

Also, temperature differences related to certain circuits inside the panel may appear as bands or blocks. If only a section of the panel shows a temperature different from its surroundings in a thermal image, there may be possible faults in the section associated with a bypass diode, part of a group of cells, or the wiring route. However, such assessments depend on equipment specifications and measurement conditions, so do not draw conclusions from the image alone; verify together with current and voltage values and inspection history.

When inspecting temperature differences along strings, cross-checking with the power plant's wiring diagrams and system diagrams is indispensable. If you don't know which panel belongs to which string and which input it is connected to, even if you detect potential anomalies in a thermal image you won't be able to translate them into actual inspection targets. This is especially true for facilities that have undergone multiple expansions or refurbishments, where the drawings and the on-site wiring may not fully match. Matching the positions shown in thermal images to the electrical system is important for narrowing down the causes of reduced power output.

When power output is low, it is easier to organize the investigation if you first distinguish whether the decline is occurring across the entire system or is concentrated in part of the array. If it is low overall, candidates include weather, solar irradiance, soiling, temperature, and system-side control. If only some parts are low, local issues such as panels, strings, connections, circuit breakers, and input circuits are likely candidates. Observing string- or band-shaped temperature differences in thermal images is a way to visually assist this differentiation.

However, in thermal images captured over a wide area, panels that are farther away suffer from insufficient resolution, making fine anomalies harder to see. It is practical to use an overview image to locate areas with temperature differences and, if necessary, confirm details with close-up images. Trying to determine the cause using only wide-area images increases the risk of oversights and misjudgments. Thermal images are more effective when used by combining overall inspection and detailed inspection rather than relying on a single image.

Observe localized temperature changes caused by shadows, dirt, and fallen leaves

Shading and soiling are often overlooked causes of reduced power generation. Even if the panels themselves are not faulty, areas of the surface that light cannot easily reach will lower power output. In particular, when only part of a panel is shaded, or when there are fallen leaves, bird droppings, soil dust, or contact with grass, thermal images may show localized temperature differences.

The effects of shading are often underestimated by visual inspection alone. On-site at a power plant, a small shadow may seem like no big problem. However, because solar panels consist of multiple cells electrically connected, the position and extent of a shadow can affect output more than it appears. In thermal images, temperature differences appear at the shaded areas and their surroundings, providing clues to reduced power generation.

The same applies to dirt. Dirt thinly spread across the panel surface can be difficult to see in regular photographs. When dust or pollen is thinly deposited over a wide area, it may not appear as a distinct hot spot. On the other hand, things that locally block light—such as bird droppings or fallen leaves—can show up as noticeable temperature differences in thermal images. Comparing visible-light images and thermal images side by side makes it easier to determine whether a temperature difference is caused by surface deposits or by a possible internal abnormality.

Shadows from weeds and nearby trees can also cause a decrease in power output. In particular, for rows of panels located low to the ground, seasonal changes in grass height and solar altitude can produce shadows that were not previously a problem. Because the position of shadows changes between morning and afternoon, a single imaging session may not fully capture the effects of shading. If there are specific times when power output is low, it is effective to review thermal images taken under conditions close to that time.

Be aware that fallen leaves and foreign objects can change condition over a short period. Even if a foreign object was present at the time of imaging, it may have been moved by the wind by the time of inspection. Conversely, something not visible at the time of imaging may have accumulated later. Therefore, if you suspect a shadow or dirt when looking at a thermal image, record the imaging date and time and the on-site inspection date and time, and confirm that conditions have not changed. Treating past conditions as if they are the same as current ones based only on images can lead to incorrect judgments.

When considering shading or soiling as the cause, it is also important to check the relationship with power generation data before taking measures such as cleaning or tree trimming. For example, if output is low only during certain seasons or times of day, shading is suspected. If output tends to recover after rain, the effect of surface soiling can be considered. If only specific rows of panels continue to show low output, you should include electrical faults as candidates in addition to shading or soiling.

The purpose of checking for shadows and dirt in thermal images is not simply to locate dirty areas. It is to assess how likely they are to be contributing to reduced power generation and to set priorities. The approach differs between widespread light soiling and localized obstructions. It is important to use thermal images to increase the information available for deciding whether to address the issue with a site-wide cleaning plan or by prioritizing removal of specific spots and improvements to the surrounding environment.

Inspect for heat buildup at connections, wiring, and around equipment

The cause of low power generation is not necessarily limited to the panel surface. Problems can also be hidden in the paths through which electricity flows, such as connection points, wiring, junction boxes, circuit breakers, terminal areas, and around the power conditioner. Thermal images are also used to check the heat generation of these electrical components.

Heating at connection points can occur due to increased contact resistance, loose or improperly tightened connections, aging, moisture, or load on the cable. When resistance increases in a part carrying current, that spot tends to heat up. The presence of heating does not necessarily indicate a serious fault, but if it is clearly hotter than other connection points under the same conditions, it should be considered a candidate for higher inspection priority.

When inspecting heat on wiring, distinguish whether the entire conductor is warm or only parts are hot. Energized wiring may show a uniform temperature rise, but if only a localized area is hot, you should check connection points, bends, damage, insulation condition, and how it is secured. The surrounding environment also affects temperature, especially where cables are bundled, exposed to direct sunlight, or in poorly ventilated locations. On thermal images, you need to carefully differentiate between electrical heating and temperature increases caused by the environment.

When inspecting junction boxes or panels, safety procedures must always be the prerequisite. It is dangerous to open an energized panel carelessly. Even if a thermal image shows abnormal heating on the outside of a panel, any internal inspection must be carried out by personnel with the appropriate qualifications and authority following proper procedures. Do not determine internal components based solely on temperature differences visible from the exterior; verify findings together with inspection records, current values, voltage values, alarm history, and other relevant data.

The temperature around the power conditioner is also important when checking for a drop in power generation. Because equipment heats up during operation, a high temperature in itself can sometimes be natural. However, if the intake or exhaust is blocked, heat is trapped around the unit, there are problems with filters or ventilation paths, or only certain input circuits show different conditions, these can be related to output curtailment or reduced operating efficiency. By checking how heat accumulates around the equipment with thermal images, you can consider the possibility that the cause of reduced power generation lies on the equipment side rather than the panel side.

When inspecting heat generation in electrical equipment, it is important to have a reference for comparison. Comparing components of the same model, under the same load conditions, and in the same installation environment makes it easier to identify potential abnormalities. Looking at a single temperature alone can make it difficult to judge whether it is high or low. For example, outdoor panels in summer tend to become hot overall, and there can be temperature differences between surfaces exposed to direct sunlight and shaded sides. Rather than judging solely by the colors of a thermal image, it is important to consider the surrounding conditions and relative differences.

Abnormalities in connections and wiring, if left unaddressed, can lead not only to reduced power generation but also to safety risks. Therefore, when thermal images reveal potential hotspots, they should be handled not only with regard to their impact on power generation but also from the perspective of equipment maintenance. It is important to organize what to inspect, how to check it, whether a shutdown is necessary, and how to record the findings, and to establish a process for requesting confirmation from specialized contractors as needed.

Compare changes with past images and power generation data

Thermal images are not something you take once and finish with; their value increases when compared with past images and power generation data. When investigating the cause of low power output, looking only at the current thermal image does not tell you whether a temperature difference was originally present or has recently developed. If you have past images, you can confirm whether a hot spot existed at the same location before or has newly appeared this time.

When comparing with past images, pay attention to differences in shooting conditions. If season, time, ambient temperature, solar radiation, wind, shooting distance, or shooting angle differ, the same equipment can appear differently. Therefore, rather than simply looking at color differences alone, it is important to compare relative differences within the same column and to select images taken on days with similar conditions. If the color display range is set to change automatically, the meaning of the colors may vary from image to image. Do not rely solely on the apparent colors; check the temperature values and the differences from the surroundings.

Cross-checking with power generation data is also essential. Even if a hot spot is found in a thermal image, you cannot tell how much that location is contributing to a drop in power output unless you look at it together with the data. For example, even if a panel has a hot spot, the impact on overall power generation may be small. Conversely, a temperature difference that does not appear significant in a thermal image may still require electrical verification if the output of a particular string is persistently low.

Not only should you check the days when you felt the power output was low, but it is also important to confirm when the decline began. Whether it dropped suddenly or fell gradually changes the suspected causes. If the drop was sudden, candidates include breaker operation, poor connections, equipment shutdown, falling objects, and sudden shading. If it is decreasing gradually, consider accumulation of dirt, growth of vegetation, component degradation, and poor equipment cooling. Thermal images serve as visual documentation to confirm those changes.

Also, comparisons within the same power plant are effective. When there are multiple installations with the same capacity, the same orientation, the same tilt angle, and similar surroundings, comparing output and temperature distribution makes it easier to identify candidates for anomalies. If only one section has lower power generation and that same section shows concentrated temperature differences, the inspection priority for that section should be higher. Conversely, if the entire plant is declining similarly, you need to consider solar irradiance conditions, ambient temperature, grid-side control, overall soiling, and so on.

To link thermal images with power generation data, the granularity of the records is important. If the images lack location information and the generation data are only overall values, it becomes difficult to narrow down the cause. If you can separate and check the data by power conditioner unit, input unit, string unit, etc., as far as possible, it becomes easier to correlate the temperature differences found in the thermal images. Managing site photos, inspection memos, work histories, cleaning histories, and mowing histories together also makes it easier to explain the background of reduced power generation.

Comparing with past images also helps with early detection of abnormalities. Even at a stage when power generation has not yet been greatly affected, if the temperature difference at the same spot gradually increases, it can be considered a sign of future trouble. In equipment maintenance, rather than responding after an abnormality becomes obvious, monitoring changes and conducting planned inspections makes it easier to reduce downtime and workload. Thermal images can be used not only to investigate causes after a drop in power generation but also for ongoing management to prevent declines.

Do not make definitive conclusions based solely on thermal images; use them to inform inspection planning.

Thermal images are a useful means of investigating the causes of low power generation, but you should avoid concluding the cause from images alone. Temperature differences can be influenced by various factors such as faults, shadows, soiling, wind, reflections, installation conditions, and imaging conditions. If you find a suspicious area in a thermal image, it is important to treat it not as a conclusion but as an entry point for developing an inspection plan.

In practice, we organize the locations identified in thermal images by urgency and their impact on power generation. Items that require safety verification, such as abnormal heating at connection points, should be referred promptly for professional inspection. Those that could potentially be remedied by on-site actions, such as localized dirt or fallen leaves, are prioritized for on-site inspection and cleaning. Suspected internal panel malfunctions are investigated with electrical measurements and history checks, and the need for replacement is determined carefully.

Thermal images are also useful as materials for explaining the causes of low power generation. When sharing the situation among on-site personnel, managers, maintenance staff, and clients, they can visually present potential anomalies that are difficult to convey with numbers alone. However, rather than raising concern based solely on the image’s appearance, it is necessary to organize and explain the imaging conditions, the data that were checked, the suspected causes, and the inspections to be carried out next. Thermal images are supporting material for decision-making and should not be used alone to draw definitive, responsible conclusions.

When incorporating findings into an inspection plan, checking reproducibility is also important. Verify whether the same temperature difference appears when the same location is imaged on a different day or at a different time, whether it improves after cleaning, and whether the hot area remains during periods without shadows. Overreacting to a temperature difference observed only once can lead to unnecessary work. Conversely, if the same tendency persists after multiple checks, it provides a basis for raising the inspection priority.

When investigating reduced power output, it is also important to keep a record of what was confirmed on site. If thermal images, visible-light images, capture locations, capture date and time, weather, solar irradiance conditions, power generation data, inspection results, and actions taken can be managed in a single workflow, it becomes easier to review the cause later. Especially when multiple people manage the equipment, insufficient records can lead to repeatedly checking the same location or treating previously addressed anomalies as new problems.

When power output is low, the important thing is not to assume a single cause. In actual field conditions, dirt, shading, equipment controls, wiring faults, temperature rise, and seasonal factors may overlap. Thermographic images are a means to identify among these the factors that appear as temperature anomalies. You should combine generation data with on-site inspections and take a methodical approach to sequentially narrow down which factors are most likely related to the decline in power output.

In summary, the key points to look for in thermographic images are hot spots at the panel level, temperature differences at the string or circuit level, localized changes caused by shading or soiling, heating at connection points and wiring, and discrepancies compared with past images or generation data. By combining these, it becomes easier to identify the causes of reduced power generation based on records rather than intuition. This also helps clarify inspection priorities while maintaining on-site safety and avoid unnecessary replacement decisions.

To conduct surveys using thermal images on an ongoing basis, a system is required that can efficiently handle image capture, location organization, image comparison, and correlation with power generation data. Recording the condition of the power plant on site and keeping it in a form that is easy to review later makes it easier to explain the causes of low power output and to make decisions about improvements. Rather than treating thermal images as a one-off check, managing them linked with inspection records and generation data contributes to stable equipment operation.