5 indicators to compare, other than the same month of the previous year, when power generation is low

When you feel that power generation is low during the monthly check of a solar power generation system, the first thing often looked at is a comparison with the same month of the previous year. A comparison with the same month of the previous year is intuitive and easy to use, but it is risky to judge an anomaly based on that alone because weather, equipment condition, operating conditions, and the surrounding environment may not be the same as in the previous year. In practice especially, the mere fact that generation is lower than in the same month last year often makes it difficult to distinguish whether the cause is a fault, insufficient solar irradiance, power curtailment, or the effects of shading or soiling.

To correctly identify the cause of low power generation, it is important to have comparison axes other than the same month of the previous year. By checking multiple indicators together, it becomes easier to calmly determine whether the issue is simply a weather effect, an equipment abnormality, or a change in the site environment.

Table of Contents

• Why it is difficult to assess based solely on the same month last year when power generation is low

• Indicator 1: Relationship between Solar Irradiance and Power Generation

• Indicator 2 Power generation per unit of installed capacity

• Indicator 3 Comparison with nearby or similarly conditioned facilities

• Indicator 4 Variability by power conditioner and by string

• Indicator 5: Recording on-site conditions such as air temperature, shadows, and soiling

• Practical steps for narrowing down causes using five indicators

• Summary: When power output is low, increase the number of comparison axes to make a judgment.

Why it's difficult to assess using only the same month of the previous year when power generation is low

When checking whether power generation is low, comparing the same month with the previous year is a useful starting point. Because the seasonal conditions are similar in the same month and sunshine hours and solar altitude do not change much, looking at how much generation has increased or decreased compared with the previous year gives a rough sense of the trend. However, solar power output can vary from year to year even in the same month. The results can differ just from variations in weather conditions — for example, years with many rainy or cloudy days, years affected by typhoons or prolonged rain, or years with many clear days but high temperatures that suppress output.

A lower year‑on‑year figure for the same month does not necessarily indicate an equipment failure. Conversely, even if the year‑on‑year comparison does not show a significant drop, there may actually be faults in some circuits or pieces of equipment. For example, if the same month last year also experienced poor weather, this year’s power generation may appear comparable to last year’s yet still fall short of the level that could be expected. Also, although there was little shading in the same month last year, this year the effects of surrounding vegetation, temporary structures, or adjacent buildings may have increased.

What practitioners should be aware of is that year-on-year comparison for the same month is a comparison of results and not an indicator that directly reveals causes. Monthly changes in power generation are determined by multiple overlapping factors such as solar irradiance, temperature, installed capacity, equipment operational status, grid-side constraints, soiling, shading, snowfall, vegetation, and missing measurement data. Therefore, judging solely by year-on-year monthly comparisons can lead to misprioritizing inspections, mistaking non-issues for abnormalities, or conversely downplaying signs that should not be overlooked.

Also, when someone reports low generation, they may be looking at different numbers — for example, "low amount sold to the grid," "the generation shown on the monitoring screen has dropped," or "the monthly report figures are lower than expected." Generated energy, sold electricity, self-consumption, power conditioner output, and readings from metering devices each mean different things. Before comparing with the same month of the previous year, it is important to clarify which figure you are comparing.

This is not to say you should avoid year-on-year (same-month) comparisons. On the contrary, they are useful for understanding long-term trends. However, when isolating the causes of low power generation in practice, you should not make the year-on-year (same-month) comparison your final judgment; you need to verify it in combination with other indicators. From here, we will look in order at five indicators you should compare besides the year-on-year (same-month) comparison.

Indicator 1: Relationship between solar irradiance and power generation

When power generation is low, the first thing to check is its relationship with solar irradiance. Because solar power generation is heavily influenced by the amount of sunlight received from the sun, it is difficult to determine the cause by looking at generation output alone. Even if monthly generation is low, if the solar irradiance for that month was similarly low, the main factor may be weather rather than an equipment fault. On the other hand, if solar irradiance was at normal levels or otherwise sufficient but generation alone is low, prioritize checking the equipment and the site conditions.

In practice, it is important to compare power generation and solar irradiance over the same period. When examining monthly generation, check monthly solar irradiance trends; when examining daily generation, check daily irradiance trends; and when looking at drops by time of day, check time-of-day irradiance trends. If the periods are misaligned, an effect that was actually caused by a period of bad weather can appear to be an equipment malfunction. Especially after consecutive rainy or cloudy days, it is easier to make a judgment if you not only look at the monthly totals but also extract and inspect the generation curves from sunny days.

When examining the relationship between solar irradiance and power generation, check not only whether there is more or less irradiance but also whether the power output generally follows changes in solar irradiance. On clear days, if the generation rises in a smooth, bell-shaped curve, peaks around midday, and declines toward the evening, the overall system may have no major abnormalities. Conversely, if, despite sufficient irradiance, generation levels off midway, drops suddenly, is unnaturally low only during specific time periods, or the curve is erratic even on clear days, check for curtailment, equipment shutdowns, shading, communication loss, or measurement anomalies.

When checking solar irradiance, it’s easier to compare if you have measured values from the installation site. However, not all facilities have a pyranometer. In such cases, combine local meteorological data, weather notes from monitoring data, site inspection records, and daily weather logs to see whether a drop in power generation is correlated with weather. Rather than assuming you can’t verify anything because there is no pyranometer, layering available records makes it easier to at least determine whether the major cause is weather-related or equipment-related.

Also, even with the same solar irradiance, higher temperatures tend to reduce the output of solar panels. If generation in summer does not increase as much as expected, it may be that, despite abundant irradiance, high temperatures are suppressing output. Therefore, it is important not to evaluate generation based on irradiance alone, but to also consider the temperature and site conditions described below. In particular, when you receive inquiries such as “it was sunny but generation was low,” check one by one which factor—irradiance, temperature, output control, shading, or equipment shutdown—is affecting performance.

Comparison with solar irradiance is an entry point for broadly dividing the causes of low power generation into two categories. If solar irradiance is low, suspect weather impacts; if only the power generation is low relative to irradiance, suspect equipment or site conditions. Because it allows comparisons that are closer to the cause than year-on-year same-month comparisons, it is a metric that is highly worth checking first.

Indicator 2 Power generation per unit of installed capacity

Next, what we want to compare is power generation per unit of installed capacity. If you only look at the total generation, you cannot correctly compare sites with different system sizes. Larger systems will appear to have higher total generation and smaller systems lower, so it is difficult to judge performance based solely on simple monthly generation. Therefore, divide the generation by the installed capacity to check how much is being generated per unit of capacity.

Viewing generation per unit of installed capacity makes it easier to compare facilities of different sizes. For example, if there are multiple solar power installations in the same area, converting to generation per unit of capacity rather than total monthly generation makes it easier to identify which installations are relatively low. Even if installed capacities differ, if installation conditions and weather conditions are similar, generation per unit of capacity tends to be roughly similar. Any installations that deviate significantly from that should be prioritized for inspection.

This indicator is particularly useful when your company manages multiple facilities. Instead of viewing a low-producing facility in isolation, you compare side-by-side how much other facilities generated in the same month. If generation per unit of capacity is low overall, that may indicate regional weather impacts or the possibility of wide-area output curtailment. Conversely, if only a specific facility is underperforming among facilities in the same region or under similar conditions, it is more likely to point to an issue specific to that facility.

When looking at power generation per unit of installed capacity, it is important to standardize the definition of installed capacity. Whether you use the capacity of the solar panels or the capacity of the power conditioner (inverter) changes the meaning of the figures. For monthly management and equipment comparisons, you should record which capacity you are using and compare using the same standard each time. If standards are mixed, generation may appear low or, conversely, appear better, leading to incorrect judgments.

Generation per unit of installed capacity is a useful metric, but it cannot fully correct for differences in installation conditions. If orientation, tilt, shading, panel type, mounting height, surrounding environment, the effects of snow or soiling, or the timing of commissioning differ, output can vary even for systems with the same capacity. Therefore, it is safer to treat this indicator not as implying “the same capacity will produce the same generation” but as a way to normalize scale differences for comparison.

In practical investigations to find causes of low power generation, after identifying installations with low generation per unit capacity, you proceed to check within those installations by equipment, by circuit, and by time of day. Distinguishing whether the low output is overall or limited to certain parts narrows the inspection scope. Generation per unit capacity is better suited for cross-sectional comparisons than for comparisons with the same month of the previous year, and it is a highly practical metric for personnel managing multiple installations.

Indicator 3 Comparison with Nearby or Similarly Conditioned Facilities

When power generation is low, it is also important to compare with nearby or similarly conditioned installations. Comparing with the same month of the previous year is a comparison along the time axis, while comparing with nearby installations is a cross-sectional comparison for the same period. If they are in the same region, weather and solar radiation trends will be similar, making it easier to determine whether only a particular installation is underperforming or whether the entire region is experiencing lower output.

In comparisons with nearby facilities, the closer they are, the more likely the weather conditions will be similar. However, even at nearby locations, mountain shading, sea breezes, fog, snowfall, localized clouds, or shadows from surrounding buildings can alter power generation conditions. Therefore, rather than simply assuming that proximity means identical conditions, it is important to use such comparisons while confirming whether they are appropriate as reference targets. If possible, comparing facilities with similar orientation, tilt, system size, commissioning date, and surrounding environment will improve the accuracy of the assessment.

For example, if several facilities in the same area are all experiencing a similar drop in power generation, consider factors such as adverse weather, wide-area output control, or grid-side issues. On the other hand, if neighboring facilities are generating normally but only your facility is underperforming, prioritize checking your facility. Making this distinction lets you organize where to start looking before suddenly performing detailed inspections of all equipment on site.

When comparing with facilities under the same conditions, daily trends are also useful. If you only look at the monthly total, anomalies lasting a few days can be masked in the overall data. When viewed by day, if only your facility shows a drop starting on a certain date, check what happened before and after that date. This makes it easier to identify triggers for reduced power generation, such as inspection work, power outages, equipment shutdowns, setting changes, communication failures, nearby construction, effects around mowing, or conditions after strong winds or heavy rain.

Comparing by time of day is also helpful. If, compared with nearby installations, output is lower only in the morning, only in the evening, or appears to hit a ceiling only around midday, those patterns can provide clues to consider shading, orientation, tilt, temperature, output control, or equipment capacity limits. In particular, morning and evening declines may be related to shadows from surrounding structures or vegetation. If output is low only during the daytime, check for high temperatures, output curtailment, equipment-side limitations, and the operating condition of the power conditioner.

When performing comparisons with neighboring facilities, you must also pay attention to the data quality of the comparison targets. If a facility you are comparing to has measurement gaps or communication failures, it cannot serve as a valid baseline. Also, if the comparison target has similar anomalies, you may overlook abnormalities in your own facility. Therefore, when multiple comparison targets are available, it is safer to look at several facilities rather than just one and confirm the overall trend.

Comparisons with nearby facilities or facilities under the same conditions are effective for separating the causes of low power generation into site-specific issues and region-wide problems. By adopting the perspective of "how does this month compare to other facilities"—which is hard to see from year-on-year monthly comparisons—you can reduce bias in judgment.

Indicator 4: Variability by Power Conditioner and by String

When power generation is low, looking only at the facility’s monthly totals can cause some anomalies to be overlooked. Even if the overall generation appears only slightly reduced, one or more power conditioners, circuits, or strings may actually be experiencing a significant drop. Therefore, in addition to comparing the facility as a whole, it is important to check the variability by power conditioner and by string.

When comparing power generation by power conditioner, operational differences between units become apparent. Even when installation conditions are similar within the same facility, if only a particular unit shows lower generation, it is necessary to check that unit for shutdowns, errors, abnormalities in the input circuit, poor connections, the status of circuit breakers, communication loss, and so on. Even if the decline looks small when considering only the total generation, examining by unit can often reveal the cause.

String-level checks allow you to observe generation imbalances at a finer scale. If, among the strings connected to the same power conditioner, a particular string shows lower current or generation, on-site factors such as panel soiling, shading, wiring, connection points, module defects, contact with vegetation, bird damage, or falling debris may be involved. Of course, because a simple comparison cannot be made if the string configuration or number of modules differs, comparisons should be made between circuits under the same conditions.

When checking variability, confirm not only absolute values but also the relative differences at the same time of day. In solar power generation, differences between circuits tend to appear in the morning and evening due to solar elevation and shadows, while the period around midday can be easier to compare. Choosing stable periods on clear days for comparison makes it easier to find abnormal circuits. On cloudy days, output fluctuates finely because of cloud movement, so do not judge based only on short-term differences; it is important to observe trends over multiple days and multiple time periods.

Variations by power conditioner or by string help narrow down whether the cause of reduced power generation is a system-wide issue or a problem confined to a specific area. If the whole system is uniformly low, consider weather, output control, grid conditions, overall soiling, measurement settings, and so on. Conversely, if only part of the system is low, focus onsite inspections and electrical checks on that range. Because the inspection scope can be narrowed, the efficiency of the response improves.

However, when judging variations, it is essential to confirm design differences. If you compare circuits with different orientations, surfaces with different tilts, rows that are prone to shading, or strings with different numbers of panels as if they were under the same conditions, you may mistake normal differences for abnormalities. It is important to review the system diagram, single-line wiring diagram, string table, layout plan, and past inspection records in advance and decide which units are appropriate for comparison.

When you suspect that power output is low, it is effective to first look at the overall figures and then break them down by equipment and by circuit. Abnormalities that were not evident in the total values can become apparent when you examine the variability. Because year-on-year comparisons for the same month tend to only reveal increases or decreases for the entire facility, indicators that show internal variability are very important in practice.

Indicator 5 Recording of on-site conditions such as air temperature, shadows, and dirt

The causes of low power generation may not be discernible from the data screen alone. Even if there are no major anomalies in solar irradiance or equipment-specific data, site conditions—such as ambient temperature, shading, soiling, weeds, snow accumulation, fallen leaves, bird droppings, dust, nearby construction work, and temporary construction materials—can affect power generation. Therefore, it is important to compare power generation data with records of site conditions.

When assessing power output, ambient temperature is an aspect that is easy to overlook. Solar panels tend to generate more electricity the stronger the solar irradiance, but their output is less likely to increase when temperatures are high. On a clear midsummer day, if there is ample sunlight yet the power output does not rise as expected, you should consider the impact of high temperatures. In particular, on rooftops or in poorly ventilated locations panel temperatures can rise easily, causing differences in output even under the same irradiance conditions. You cannot conclude from ambient temperature alone, but checking it together with solar irradiance makes it easier to distinguish whether the issue is an anomaly or simply a seasonal characteristic.

Recording shadow patterns is also important. Shadows change with the time of day, the season, and the solar altitude. Even in the same month of the previous year, changes such as surrounding trees growing, structures being added on adjacent land, material storage areas being rearranged, or grass growing around slopes or fences can affect power generation. If generation is low only in the morning and evening, or if only a particular string is showing low output, check for possible shading. Keeping site photos from the same position and the same time of day makes later comparisons easier.

Dirt can also be a factor in reduced power generation. The causes of soiling vary by site and include soil dust, pollen, yellow sand, bird droppings, fallen leaves, mud splatter, and the effects of exhaust and airborne particulates. Sometimes these are washed away naturally by rain, but in areas where panels have a gentle tilt, places where dirt tends to accumulate at the lower edge, or locations with a lot of exposed soil nearby, soiling can remain. When power generation is low, checking data before and after cleaning or changes after rainfall makes it easier to estimate the impact of soiling.

Changes in weeds and vegetation should also be recorded as site conditions. On ground-mounted installations, grass can reach the lower edge of panels or cast shadows along fences or around the mounting racks. If overall generation or per-string variability changes before and after mowing, vegetation may have been affecting performance. Because weeds can grow suddenly with the seasons, photographic records taken during site visits are useful in addition to monthly data.

When comparing site conditions, it is important to standardize the format of records. If you consistently record the photo location, shooting direction, time of day, weather, inspector, and observations each time, it becomes easier to judge changes when reviewing later. Rather than increasing records only in months with low power generation, recording during normal times makes the difference from abnormal conditions visible. Without photos and inspection notes from normal conditions, it is difficult to tell whether there has been a change at the site or if it was the original condition.

On-site conditions such as temperature, shading, and soiling are indicators that help account for causes that are hard to see from numerical data alone. By capturing site changes that same-month year-on-year comparisons cannot detect, you can assess the reasons for low power generation in a way that is closer to reality. In practice, it is important not to separate monitoring data and site records, but to review them together on the same time series.

Practical procedure for narrowing down causes using five indicators

The five indicators we have reviewed so far are each useful on their own, but in practice their effectiveness increases when used in combination. When power generation is low, what’s important is not to decide from the outset on a single cause. If you make snap judgments—assuming a fault because the year-on-year comparison is low, assuming an anomaly because it was sunny, or assuming equipment failure because output is lower than neighboring sites—you can end up overlooking issues or carrying out unnecessary inspections.

In practical workflow, first clarify which figure you are comparing. Confirm whether it is generated electricity, the amount sold, the surplus after self-consumption, or the reading on the monitoring device. If this remains ambiguous, changes on the consumption side or in measurement can be misinterpreted as a decline in generation output. Especially for systems that include self-consumption, generated electricity and sold electricity do not match, so first organize which figure you are looking at.

Next, examine the relationship with solar irradiance. If irradiance is low in a given month, lower power generation itself may be a natural outcome. Conversely, if irradiance is sufficient but generation is low, proceed to check the equipment and site conditions. At this stage, checking the power generation curves on clear-sky days makes it easier to detect anomalies that are not apparent from monthly totals alone.

After that, normalize for capacity and compare using generation per unit of capacity. If you manage multiple facilities, check whether generation per unit of capacity is significantly lower than at other facilities. If it is low overall, suspect weather or regional conditions; if only certain facilities are low, suspect site-specific problems. If the conditions for the comparison differ, take those differences into account when making your assessment.

Next, compare with nearby installations or ones under the same conditions. Determining whether installations in the same area are similarly low, or whether only your installation is low, will change the inspection priorities. If only your installation is low, proceed to checks by equipment or by string. If only particular power conditioners or strings within the same installation are low, you can narrow down the scope of the cause.

Finally, cross-check against the records of on-site conditions. Check what was happening at the site at the same times as the data changes — shadows, soiling, weeds, ambient temperature, snowfall, nearby construction, inspection work, power outages, setting changes, etc. If days or time periods with low power generation overlap with changes in on-site conditions, it becomes easier to formulate hypotheses about the cause.

By checking in this sequence, you can broadly classify decreases in power generation. If the entire output is low in correlation with solar irradiance, weather effects are likely. If solar irradiance is sufficient but the entire output is low, check output control, overall soiling, measurement settings, and grid-side conditions. If only some equipment or strings are low, suspect equipment shutdown, wiring, connections, shading, or localized soiling. If only your facility is lower compared to neighboring facilities, prioritize site-specific factors.

What matters is keeping a record of the comparison results. If you record the day you noticed low power generation, the metrics you compared, the data you checked, what you observed on site, the hypotheses you formed, and the actions you took, it will be easier to judge when a similar decline occurs next time. In monthly reports and inspection reports, instead of simply writing “power generation is low,” leaving the results of comparisons — for example, “solar irradiance was also low,” “neighboring facilities showed the same trend,” “only a specific string had declined,” or “an improving trend was seen after mowing” — will make the documents more useful in practice.

Isolating the causes of low power output may not yield a definitive answer in a single attempt. Weather, equipment, on-site conditions, and measurement factors can overlap. Therefore, it is important to narrow down the possibilities using multiple indicators and to take an approach that leads to the necessary inspections and corrective actions.

Summary: When power output is low, increase the number of comparison axes when making judgments

When power output is low, comparing with the same month of the previous year is an easy-to-understand method, but that alone is not sufficient to determine the cause. Weather varies from year to year even in the same month, and site conditions and equipment condition also change. If you conclude that there is an anomaly based solely on a lower result than the same month last year, you may increase unnecessary inspections or, conversely, overlook anomalies that should actually be checked.

Indicators to compare in practical work include the relationship between solar irradiance and power output, power output per unit of installed capacity, comparisons with neighboring or similarly conditioned facilities, variations by power conditioner and by string, and records of site conditions such as temperature, shading, and soiling. By combining these, it becomes easier to determine whether a decline is due to natural weather-related decreases, an equipment-specific anomaly, or changes in the site environment.

What is especially important is not to view power generation as a standalone number. Power generation is the result of many factors, such as solar irradiance, temperature, equipment operating conditions, shading, soiling, vegetation, and measurement conditions. When you see a low result, that is precisely when you should calmly verify it using multiple comparison axes.

Also, it is important to store comparison results as records. If you retain normal power generation trends, generation curves for sunny days, site photos, inspection records, equipment-specific data, and string-level data, it becomes easier to judge deviations during abnormalities. Rather than rushing to check only when generation is low, maintaining the ability to make routine comparisons leads to more stable facility management.

To streamline identifying the causes of reduced power generation, linking on-site records with generation data and having a system that allows immediate comparison when needed is helpful. If you want to continuously monitor the condition of equipment with low power output, rather than relying on a specific product name, it is important to establish operations for monitoring data, inspection records, photo management, and report creation that fit your company’s equipment scale and management structure.