Five steps to compare causes of low power generation by generation per kW

When you feel that the power output of a solar PV system is low, the first thing to check is not simply the absolute amount of generation. If system capacities differ, the same amount of generation can be evaluated differently. Comparing a 10 kW system and a 50 kW system by the same absolute generation alone can lead to incorrect conclusions about the cause. In such cases, it is useful to compare using generation per kW of capacity.

Generation per kW is a method of dividing the amount of electricity produced over a given period by the installed capacity. For example, dividing monthly generation by the installed capacity evens out differences in system size and allows comparison. This makes it easier to determine whether a temporary decline is due to weather or is related to shading, soiling, equipment malfunction, or differences in measurement conditions.

This article explains five steps for comparing causes of reduced generation using generation per kW, aimed at practitioners who search for "low power generation". Even when power plant sizes and installation conditions differ, standardizing the order of checks makes it easier to move beyond intuitive judgment and systematically narrow down the cause.

Table of Contents

• Standardize the baseline before comparing generation per kW

• Step 1 Organize generation data and equipment capacity and unify units

• Step 2 Calculate daily and monthly generation per kW

• Step 3 Separate natural declines by examining weather and seasonal differences

• Step 4 Compare differences at the zone, system/string, and power conditioner unit levels

• Step 5 Investigate on-site causes starting from locations with large differences

• Cautions when evaluating based on generation per kW

• Continuously visualize causes of low generation

Standardize the criteria before comparing generation per kW

When determining whether power output is low, the first thing to avoid is comparing systems while ignoring differences in system size. Generation itself varies greatly depending on the installed solar panel capacity, tilt angle, azimuth, solar irradiance conditions, surrounding environment, and equipment configuration. Because plants with larger installed capacity tend to produce greater total generation, simply looking at generation figures alone does not reveal whether the generation efficiency is good or bad.

A convenient metric for that is generation per kW. This divides the generation by the installed capacity to see how much was generated per 1 kW. Viewed daily, it can be compared as kWh/kW per day. Viewed monthly, you can check differences between installations as monthly kWh/kW. By using a metric that standardizes installed capacity, it becomes easier to spot differences between power plants of different sizes and between sections within the same plant.

However, you cannot determine all causes based solely on generation per kW. Even with the same installed capacity, differences in tilt angle and orientation will change generation patterns. Installations arranged neatly on flat ground and installations dispersed across sloped terrain receive sunlight and cast shadows differently. Furthermore, there are site-specific factors such as snow, fallen leaves, dust, bird droppings, vegetation growth, and shadows from nearby structures.

Therefore, energy generated per kW should be used not as a final judgment to determine the cause, but as an entry point to detect whether a decline has occurred and its extent. By quantifying differences and prioritizing on-site inspections for locations with large deviations, you can more easily reduce unnecessary checks. Conversely, if you conclude "this equipment is faulty" based solely on the numbers, you risk overlooking differences caused by weather or measurement conditions.

In practice, it is important not just to look at daily generation but to regularly check the value obtained by dividing it by the installed capacity. If you manage multiple plants, comparing them over the same period, using the same aggregation unit and the same capacity standard helps detect anomalies early. Even when a single plant has multiple systems, listing generation per kW by system makes it easier to tell whether only a specific section is underperforming or the entire plant is.

Step 1 Organize power generation and installed capacity and standardize units

The first step is to organize the generation and the installed capacity used for comparison. To compare the causes of low generation by generation per kW, the units and periods of the source figures must be consistent. Generation is generally measured in kWh, but if aggregation periods differ—daily, monthly, yearly—they cannot be compared. Installed capacity is often expressed in kW, and dividing generation by installed capacity yields the generation per kW.

For example, if the generation in a month is 5,000 kWh and the installed capacity is 50 kW, the generation per kW for that month is 100 kWh/kW. In the same month, if another installation generates 2,000 kWh with a capacity of 20 kW, that is also 100 kWh/kW. In this case, looking only at total generation the former appears larger, but per unit of capacity they can be judged to be similar. The actual validity varies with region, season, equipment conditions, and solar irradiance conditions, so you should not judge good or bad performance based solely on the example figures.

What you need to be careful about here is which capacity you use as the facility capacity. Depending on whether you look at the total capacity of the solar panels or the rated capacity of the power conditioner, the meaning of the calculation results changes. Generally, when comparing generation performance, the installed capacity of the solar panels is often used as the benchmark, but if you want to assess the effects of an oversizing design or output control, you also need to confirm the capacity on the power conditioner side. If standards are mixed within the company or on site, the same facility can end up with different evaluations.

Also, you need to align the sources of generation data. There are several types of data that can be treated as generation—values on monitoring screens, readings displayed by the power conditioner, readings from the export meter, outputs from aggregation software, and so on. However, the measurement locations and aggregation timing for each may differ. Due to power consumed within the facility, communication losses, aggregation cut-off times, and differences in rounding, they may not match exactly.

Before making comparisons, it is important to decide which power generation figures to use. In practice, separating monitoring data used to observe daily trends from meter-reading data used for final performance verification reduces confusion. Monitoring data is well suited to early detection of anomalies, but it may include communication failures or missing data. On the other hand, meter-reading data is suitable for performance verification, but it can make it difficult to see detailed changes by time of day.

It's also a good idea to cross-check equipment capacity against ledgers, as-built drawings, and equipment configuration lists. In facilities that have undergone expansions, partial shutdowns, or equipment replacements, capacity may still be recorded using the previous values. If part of the solar panels has been removed, a specific system has been offline for an extended period, or power conditioners have been replaced, the observed generation per kW can be significantly skewed.

The purpose of this procedure is to standardize the period for generation, the source of generation data, the criteria for equipment capacity, and the scope of coverage. If you move on to the next analysis while these remain ambiguous, you will not be able to tell whether the cause lies with the equipment or with the aggregation conditions. When you feel that generation is low, it is especially important to first align the numerical foundations.

Step 2 Calculate daily and monthly generation per kW

Next, using the organized generation data and the installed capacity, calculate the generation per kW on a daily or monthly basis. Viewing it daily makes it easier to detect sudden drops or anomalies on specific days. Viewing it monthly makes it easier to grasp seasonal trends and long-term declines. In practice, it is effective to examine both, depending on the purpose, rather than only one.

Daily generation per kW is calculated by dividing that day's generation by the system capacity. If it is significantly lower on a particular day despite continued sunny weather, that should prompt suspicion of output curtailment, equipment shutdown, communication loss, breaker operation or tripping, sudden shading, snowfall, soiling, or similar causes. On the other hand, if it is only lower on rainy or cloudy days, it may be within the range of natural variation.

Monthly generation per kW is calculated by dividing the monthly generation by the installed capacity. Viewing by month smooths out day-to-day weather variations to some extent, making it suitable for comparing multiple systems. If a single plant is lower than surrounding facilities in the same month, you should check not only regional differences but also the installation conditions and operational status of the equipment. Comparing with the same month in the previous year is also useful, but because of year-to-year weather differences, you should avoid concluding that a lower value than the previous year is necessarily abnormal.

It is important to view the calculated values side by side under the same conditions. For example, align the comparison as much as possible by using multiple facilities in the same region, multiple power conditioners within the same power plant, the same row of mounting racks, or sections with the same orientation; when there is a difference in generation per kW among facilities with similar conditions, it becomes easier to narrow down the cause.

On the other hand, simply comparing equipment in regions that are far apart can be greatly affected by differences in solar radiation and weather. Even within Japan, sunshine hours, snowfall, weather during the rainy season, and the impact of typhoons vary by region. When comparing equipment in distant regions, it is safer to treat lower values not as abnormalities but as reference values for trend checking and prioritization.

When reviewing calculation results, pay attention not only to the average but also to the variability. Even if the average does not look that bad, performance can drop sharply on specific days. Conversely, although day-to-day variability may be large, monthly figures can often be roughly the same as those of surrounding equipment. By examining both daily and monthly data, it becomes easier to determine whether a decline is temporary or sustained.

Also, when investigating low power output, it’s important to look at trends over multiple months rather than just a single month. If only one month is low, weather or a temporary shutdown may be the cause, but if generation per kW has been declining for several consecutive months, factors such as dirt buildup, vegetation growth, equipment degradation, poor connections, or increased shading may be involved. The suspected causes differ between a sudden drop and a gradual decline.

What's important here is not to overcomplicate the calculation itself. First, divide the energy generated by the installed capacity and compare them over the same period—even that alone will yield many insights. When you feel generation is low, making a habit of converting total generation into generation per kW of capacity, rather than judging by total generation alone, is the first step in cause analysis.

Step 3: Consider weather and seasonal variations to distinguish natural declines

After calculating generation per kW, next check weather and seasonal variations. Because solar power generation is highly dependent on solar irradiance, days or months with low generation alone do not necessarily indicate equipment failure. Periods of continuous cloudy skies, rain, snowfall, yellow sand (Asian dust), typhoons, or prolonged rain tend to reduce generation. First, it is important to separate drops caused by natural conditions from those caused by equipment problems.

When examining daily generation per kW, check it together with that day's weather records. Whether it is significantly lower on sunny days or on cloudy/rainy days changes the interpretation. If generation on clear days is noticeably lower than that of nearby installations, equipment-related causes are more likely. On the other hand, if multiple installations in the same area are simultaneously low, the influence of weather or solar irradiance conditions is likely to be large.

Seasonal differences are also important. Solar altitude changes with the seasons, and even with the same installation the way shadows fall and how sunlight enters differs between summer and winter. In winter, because the solar altitude is lower, shadows from surrounding buildings, trees, mounting racks, utility poles, fences, and so on tend to lengthen. Shadows that did not cause problems in summer can become the cause of reduced power generation only in winter. Conversely, in summer rising temperatures can lower the output of solar cells, so even with strong solar irradiance the power generation may not increase as much as expected.

For month-by-month comparisons, comparing with the same month of the previous year and with nearby facilities is useful. However, when comparing with the same month of the previous year, if the previous year was unusually sunny, this year may appear low. Conversely, if the previous year had prolonged bad weather, this year may appear high. Therefore, it is important not to rely solely on a simple interpretation such as “a certain percent lower than the previous year,” but to view the results together with other facilities in the same region and with solar radiation trends.

Also, in areas where output curtailment or grid-side constraints occur, power generation can be suppressed even when the weather is good. In such cases, the actual energy produced will be low even if the equipment itself is operating normally. A low generation per kW should not be immediately attributed to panel failure or installation defects; operational control logs and shutdown histories should also be checked.

In snowy regions, the effects of snow need to be examined with particular care. Depending on the amount of snow and the state of snowmelt, the timing for power generation to resume can vary by installation even within the same area. The tilt angle, racking height, surrounding airflow, and how easily snow slides affect how snow remains on the solar panel surface. If snow remains only in parts, output can drop at the string level. If daily generation per kW does not recover even after sunny weather, it is worth checking the site for remaining snow or soiling.

When distinguishing weather effects from seasonal differences, be aware of whether the decrease is overall or partial. If all equipment in the same area is performing poorly, weather-related factors become more likely. If the entire power plant is performing poorly, solar irradiance conditions, output control, or shutdowns of common equipment might be involved. If only a specific section or only certain power conditioners are performing poorly, site-specific shading, soiling, connection issues, or equipment faults become more likely.

The purpose of this procedure is not to immediately assume a drop in generation per kW is an anomaly, but to first separate out natural variations. If you can exclude declines that can be explained by weather or seasonality, the areas that truly need checking become easier to see. Rather than rushing into an on-site response when generation is low, first combine comparisons with weather, season, and nearby equipment to clarify the likely direction of the cause.

Step 4 Observe differences at the area, system, and power conditioner unit levels

After checking the weather and seasonal differences, next look at where within the plant the discrepancies are occurring. To identify the cause of low generation, it is important to compare not only the plant-wide generation per kW but also by section, system, string, and power conditioner unit. Relying on overall values alone can cause you to miss anomalies that are concentrated in parts of the plant.

The first thing to check is variations within the same power plant. Even if the overall generation per kW is low, the cause differs depending on whether all sections are equally low or only some are low. If the entire plant is underperforming, check the weather, output control, shared incoming power equipment, missing monitoring data, and shutdowns on the main feeder side. If only some sections are low, focus on shading, dirt, vegetation, panel damage, poor connections, power conditioner shutdowns, and string abnormalities.

When comparing on a power conditioner unit basis, it is important to check the solar panel capacity connected to each unit. Not all power conditioners necessarily have the same capacity connected. If you compare only the generated output while the connected capacities differ, units with smaller capacities will appear to perform worse. Dividing each power conditioner's generated output by the corresponding solar panel capacity and viewing it as generation per kW allows a fairer comparison.

Even with the same power conditioner, the connected orientations and tilts can differ. In installations where east-, west-, and south-facing arrays are mixed, generation patterns by time of day differ. Because there are sections that perform strongly in the morning, sections that perform strongly at midday, and sections that ramp up in the afternoon, there are differences that cannot be seen from daily totals alone. In addition to daily generation per kW, looking at time-of-day output curves makes it easier to detect the effects of shading and orientation.

For example, if a section is significantly lower only in the morning, shading on the east side or shading from obstacles in the morning may be the cause. If it is lower only in the afternoon, shading on the west side or the influence of surrounding trees, buildings, or slopes may be involved. If it is low around midday as well, you need to consider broader causes such as soiling of the solar cell surface, connection problems, power conditioner operating limits, overheating, or equipment faults.

When comparing sections, it is important to link the numerical values to the on-site drawings and layout plans. Even if you can identify a system with low readings in the data, inspections will take longer if you don't know which row, which mounting rack, or which power conditioner it corresponds to on site. Organizing system numbers, combiner box numbers, power conditioner numbers, racking rows, and the range of solar panels, and correlating them with the generation data will streamline cause identification.

One thing to pay particular attention to is cases where the names on the monitoring screen do not match the on-site labels. If the as-built name, the on-site label, the asset register, and the name on the monitoring screen are out of sync, there is a risk of inspecting a different location than the one judged to be underperforming. Before tracing the cause of low power generation, confirming that the system(s) represented in the data correspond correctly to the equipment on site will make it easier to avoid incorrect judgments.

The purpose of this procedure is to determine whether the decline is widespread across the system or occurring only in specific areas. By comparing generation per kW by section or by equipment unit, you can narrow down the areas that need inspection. Breaking a major problem of low generation down into smaller, verifiable units makes it easier to carry out practical responses.

Step 5 Verify on-site causes starting from locations with large discrepancies

When low-performing areas become apparent in a comparison of generation per kW, finally investigate the on-site causes. Because numbers alone cannot determine the cause, it is necessary to cross-check with the actual condition of the equipment. As a priority, it is efficient to first inspect locations that show large differences compared with surrounding equipment or areas under the same conditions, locations that have dropped suddenly, and locations where the decline is ongoing.

On-site checks should first look for the presence or absence of shading. Tree growth, weeds, buildings, utility poles, fences, adjacent equipment, slopes, and the relative positions of the mounting structures can cast shadows on the solar module surface. Even partial shading can affect power generation, and the occurrence of shading changes with time of day and season. Even if no shading is visible during inspection, shadows may appear in the morning or evening or in winter, so confirm them by comparing low-generation time periods with the site’s orientation.

Next, check for dirt and deposits on the solar panel surface. If sand and dust, bird droppings, fallen leaves, pollen, yellow sand, mud splashes, or rain streaks are widely attached, power generation may decrease. In particular, low-tilt installations and sites with exposed soil, farmland, or many trees nearby can more easily retain dirt. However, avoid concluding that a large drop has occurred based solely on light soiling. It is important to see whether the measured decrease in output corresponds to the extent and area of the dirt.

Don't overlook the effects of vegetation. In ground-mounted solar PV installations, grass can grow and reach the lower edge of the solar panels. Growth is rapid in summer, and even if there was no problem at the previous inspection, it can become a shading source in a short period. Vegetation shading often appears only on specific rows or along the outer perimeter, so it becomes easier to identify when paired with differences in generation per kW by section.

On the equipment side, check the power conditioner's operating status, error history, stop history, temperature rise, and cooling condition. Clogged filters or intake/exhaust vents, insufficient surrounding ventilation, or high-temperature environments can cause the output to be reduced. Even if no error is displayed, if there are signs such as output not increasing during certain time periods, repeated reboots, or delayed start of operation, a detailed inspection is necessary.

Junction boxes and the areas around cables should also be checked. Poor connections, loose terminals, broken wires, damaged insulation, water ingress, corrosion, blown fuses, and similar issues can cause reduced power output in specific circuits. Because these matters involve safety, inspections should be carried out only after the necessary qualifications and procedures have been met. Some problems may not be apparent on visual inspection, so avoid unnecessary disassembly or contact with live parts, and hand the work over to specialized personnel as needed.

Missing monitoring data and communication failures are also subjects for on-site inspection. Although generation may actually be occurring, gaps in communications can make the recorded generation appear low. Compare the power conditioner's display values and meter readings with the monitoring data to determine whether the generation itself is low or whether there is a problem with data acquisition. In particular, on days when the communications device was restarted or there were line faults, the daily power generation can drop unnaturally.

After on-site inspections, record the findings tied to quantitative data. Note which section, from when, and by how much generation per kW was reduced, and what was observed on site, so that future decisions are easier. After countermeasures, check the recovery status using the same indicators. If cleaning, weeding, equipment restoration, or connection corrections were performed, verify the effectiveness of the measures by checking whether subsequent generation per kW has returned to the same level as surrounding equipment.

In this procedure, the objective is to corroborate differences found in the numerical data with facts observed on site. Energy generation per kW is a tool for narrowing down candidate causes, and only when combined with on-site verification does it lead to practical decision-making. To efficiently identify the causes of low power generation, it is important to treat numerical comparison, time-of-day checks, on-site inspection, and rechecking after countermeasures as a continuous flow.

Points to note when judging generation per kW

Energy output per kW is a convenient metric, but if used incorrectly it can lead to incorrect judgments. The most important caveat is to avoid simply comparing systems that are not being evaluated under the same conditions. Dividing by installed capacity does not make the comparison fair. If installation angle, orientation, region, surrounding environment, installation timing, equipment configuration, or the presence of output control differ, the energy output per kW will also vary.

For example, a system oriented mainly to the south and a system oriented east–west will have different daytime output curves. Annual and monthly generation may also differ. If the tilt angle is different, seasonal generation trends will change. Some conditions tend to boost output in winter, others in summer, and a comparison of a single month alone cannot fully evaluate this. Therefore, when looking at differences in generation per kW, you must always make your assessment with the differences in system conditions included.

Next, it is also important not to judge based only on short-term data. If values are low for just one day or only for a few hours, factors such as weather, passing clouds, output control, communication loss, or temporary stoppages may be influencing the results. Of course, a sudden drop can be a sign of an abnormality, but it is dangerous to conclude a failure or degradation based solely on an isolated decline. It is important to combine daily, monthly, and time-of-day analyses to assess the persistence of the decrease.

Also, when investigating the causes of low power generation, you should be careful when comparing actual output with expected values. Simulation figures and the assumed generation used at the design stage are based on specific conditions. If actual weather, soiling, surrounding environment, or equipment operating conditions differ, actual performance may fall short of those assumptions. Rather than immediately concluding construction defects or equipment failures because output is lower than expected, you need to check the differences between the assumed conditions and the actual conditions.

The impact of output control and suppression of voltage rise is another point that is easy to overlook. If there is sufficient solar irradiance but power generation does not increase, not only equipment faults but also grid-side conditions and operational controls may be involved. If you suspect only the solar cells or wiring without checking the power conditioner's operation history, records of output suppression, and voltage-related histories, identifying the cause will take a roundabout path.

Furthermore, note that the generation per kW also changes depending on which capacity standard is used. Whether you base it on solar panel capacity, power conditioner capacity, only the capacity currently in operation, or the total installed capacity, the results will differ. When preparing comparison materials, you must clearly state which capacity was used as the divisor.

Missing data and corrections are also important. If generation during periods when communications were interrupted is aggregated as zero, the generation per kW will appear lower than it actually is. Conversely, when data is later supplemented, preliminary figures may differ from finalized figures. Before judging that generation is low, checking for data gaps, duplicates, and misalignments in aggregation timing can prevent incorrect conclusions.

Generation per kW is not a metric that can, on its own, definitively determine whether an anomaly exists. However, it is useful as a common indicator for organizing the causes of low power generation. The important thing is that, after identifying a discrepancy in the values, you check in order whether that discrepancy can be explained by weather, season, installation conditions, operational conditions, or on-site anomalies.

Continuously visualize the causes of low power generation

Rather than checking the cause of low power generation once and stopping there, continuously visualizing it has great practical significance. Because solar power generation systems are installed outdoors, the surrounding environment and equipment conditions change over time. Vegetation grows, dirt accumulates, components deteriorate, and the communication environment can also change. Equipment that was normal at one point can show noticeable declines in power generation a few months later.

In ongoing management, it is important to regularly record daily and monthly generation per kW and keep them in a state where they can be compared with past values and nearby equipment. This allows you not only to detect sudden drops but also to catch trends that are gradually decreasing. Gradual declines are difficult to notice in daily monitoring, but they can become clear when viewed as monthly trends.

Also, linking the history of countermeasures with numerical data improves management accuracy. If you record the dates when you performed weeding, cleaning, equipment restoration, wiring corrections, or pruning of surrounding trees, you can check how the generation per kW changed before and after those actions. If the numbers improve after measures, it becomes easier to explain the relationship between cause and countermeasure. If there is no improvement, it provides an opportunity to re-examine other causes.

If you manage multiple power plants, it's useful to prepare management spreadsheets or dashboards that allow comparison in the same format. If the aggregation units or capacity criteria differ for each plant, checking them each time takes longer. Organizing daily and monthly data, installed capacity, generation per kW, degradation rate, outage history, and on-site response history makes it possible to quickly identify units with low generation.

It is also important for on-site personnel and managers to share decision criteria. Even if shadows or dirt are visible in the field, if they are not linked to the figures on the management side, it can be difficult to convey their priority. Conversely, even if management notices a drop in the numbers, inspections become inefficient if they do not know where on site to check. Using generation per kW as a common metric makes it easier to have conversations that connect the numbers with on-site conditions.

Low power generation can be caused not by a single factor but by multiple overlapping factors. For example, a month of poor weather combined with shading in certain sections, or an accumulation of dirt together with communication outages. Even in such cases, separating generation per kW at the overall, section, and equipment levels makes it easier to identify which factors are affecting which areas.

Finally, visualizing power generation helps not only with anomaly response but also with improving everyday management. If you can identify which equipment is stable, which equipment experiences large seasonal drops, and which sections repeatedly have issues, that information can inform inspection schedules, weed-control plans, cleaning decisions, and consideration of equipment replacement. Rather than scrambling to check only when generation is low, routinely monitoring generation per kW will enable earlier detection of and response to abnormalities.

To correctly identify the causes of low power generation, it is effective to compare not only total generation but also generation normalized by installed capacity (generation per kW). First, align the standards for generation and installed capacity, calculate on a daily and monthly basis, separate out weather and seasonal differences, look for differences by site block and by individual equipment, and finally verify the causes on site. By institutionalizing this workflow, it becomes easier to manage generation without relying on intuition. If you want to continuously monitor the plant’s condition and detect early signs of decline, it is important to put in place a system that links on-site information with generation data.