5 perspectives to distinguish increased self-consumption from system anomalies when power generation is low

When managing a solar power generation system, if you feel that "generation is low," the first thing to check is whether there is an abnormality in the generation equipment itself or whether the amount of electricity sold to the grid or the apparent surplus has decreased because more of the generated electricity is being consumed within the building or facilities. If you look only at the amount sold to the grid, generation may appear to have declined even though in reality increased self-consumption has merely reduced the amount flowing outward. Conversely, if you assume a decline in generation is due to increased self-consumption, you risk overlooking abnormalities such as dirty panels, shading, equipment shutdowns, or measurement faults.

This article organizes five perspectives for distinguishing increased self-consumption from equipment faults when power generation is low, ordered by how easy they are for on-site personnel to check.

Table of Contents

• First checks to make when you feel the power generation is low

• Perspective 1: Observe changes in self-consumption and the amount of electricity sold simultaneously

• Perspective 2: Read the time-of-day electricity consumption and the power generation curve overlaid.

• Perspective 3: Isolate changes caused by weather, seasons, and load equipment

• Viewpoint 4 Check for deviations in measured values and signs of equipment anomalies

• Perspective 5: Organizing Record-Keeping and Consultation Decisions

• Summary to Avoid Overlooking Low Power Generation Conditions

What to check first when you notice low power generation

When you think the power generation is low, if the first figure you check is only the amount of electricity sold, it's easy to draw the wrong conclusion. In photovoltaic systems, not all generated electricity is necessarily sold. If electricity is being used within the building or facility, part or most of the generated electricity may be used for self-consumption, and only the surplus is recorded as the amount sold or as reverse power flow energy. Therefore, even if the amount sold decreases, it does not necessarily mean that the generation itself has decreased.

In particular, for self-consumption systems and systems that sell surplus power, you need to consider "generated power", "self-consumption", "sold power", and "purchased power" separately. Even if the amount of generation is the same, if daytime use of air conditioning, manufacturing equipment, pumps, charging equipment, lighting, ventilation, and refrigeration/freezing equipment increases, self-consumption will increase and sold power will decrease. In this case, looking only at the sold power statement may make generation appear low, but the generated electricity may be being used within the facility.

On the other hand, if the power generation itself has declined, it is necessary to check the condition of the generation equipment. Causes are not limited to a single factor: dirt on the panel surface, shading from vegetation or buildings, stoppage of the power conditioner, grid-side restrictions, faults in junction boxes or wiring, or communication failures of measuring devices. Depending on the type of anomaly, you need to look at a combination of not only the amount of electricity sold but also monitored generation values, output by each piece of equipment, alert history, and the on-site external condition.

In practice, it is important first to clarify what the figure you feel is "low" refers to. Whether the amount of electricity sold is low, the monitored generation value is low, it is lower than the same month last year, lower than the simulated value, or lower than adjacent facilities will determine where you should look. Rather than lumping the phrase "generation is low" together, organizing which metric is low, since when, and by how much makes it easier to judge whether the cause is increased self-consumption or an abnormality.

Also, it is important not to judge based on a decrease on a single day. Solar power generation is affected by weather, so output changes due to cloudy skies, rain, yellow sand, snowfall, temperature, humidity, and changes in solar irradiance. With data from a single day alone, it can be difficult to distinguish whether it is an abnormality in the power generation equipment or a fluctuation caused by natural conditions. By comparing data across multiple days, the same weekday, the same time of day, and the same season, it becomes easier to determine whether the decrease exceeds the normal range of variation.

When troubleshooting low power generation, it is more efficient to first organize the relationships in the data rather than immediately going to the site and suspecting the equipment. Confirm the relationships between generated power, self-consumption, electricity sold, and electricity purchased; then examine changes by time of day; and finally consider abnormalities in equipment or measurement. Following this flow makes it easier to reduce unnecessary inspections and oversights.

Perspective 1: View changes in self-consumption and electricity sales simultaneously

When you feel the power generation is low, the first thing to check is the relationship between self-consumption and electricity sold. If electricity sold has decreased while self-consumption over the same period has increased, it may not be a fault in the generation equipment; rather, increased electricity use within the facility has reduced the amount available for sale. Conversely, if self-consumption has not increased but both electricity sold and power generation have decreased, you need to check the generation side and the metering.

One thing to be careful about is not to regard a decrease in the amount of electricity sold as a decrease in power generation. For example, if changes occur such as putting new load equipment into operation during the daytime, longer air-conditioning operating hours, an increase in operating days for factories or warehouses, a shift from weekend to weekday operation, or greater use of charging equipment, more of the generated electricity will be consumed within the facility. As a result, the amount of electricity sold may decrease even though the system is generating normally.

To confirm an increase in self-consumption, it becomes easier to judge by looking not only at generation and electricity sold but also at the movement of purchased electricity. If daytime purchased electricity decreases and electricity sold also decreases, the generated electricity may be being redirected to self-consumption. Conversely, if daytime purchased electricity increases while generation falls, it may indicate that the photovoltaic side is not supplying sufficient power, and you should check for equipment abnormalities or a drop in output.

However, self-consumption may, depending on the system configuration, be measured directly or estimated from generation, sold electricity, purchased electricity, and so on. Because the figures seen vary depending on where monitoring systems and electricity meters are installed, it is important to confirm what each meter is measuring before making comparisons.

The relationship between power generation, self-consumption, and electricity sold becomes easier to understand when compared not to a single number but to the previous month, the same month last year, the most recent sunny day, or days with the same operating conditions. In particular, when there is a concern that "the amount of electricity sold is lower than last year," the way the facility was used may have changed between last year and this year; if operating hours, number of employees, equipment additions, air-conditioning settings, number of operating days, or whether the facility operated on holidays have changed, comparing only the amount of electricity sold will not provide an accurate assessment.

Also, it is important to know whether the increase in self-consumption is temporary or persistent. If self-consumption has risen for only a few days, it may be caused by temporary work or a transient increase in loads. If it has increased throughout the month, operational changes or added equipment may be the cause. If it rises seasonally, seasonal loads such as heating and cooling, ventilation, pumps, dehumidification, and refrigeration/freezing may be influencing it.

What operational staff often overlook is that it may be the "surplus power," not the generated power, that has decreased. Surplus power is the result of subtracting the electricity used within the facility from the electricity generated. Therefore, even if the generation equipment is operating normally, an increase in electricity consumption will reduce surplus power. When the amount of electricity sold is low, before assuming there is a fault in the generation equipment, it is essential to check where the generated electricity was used.

However, it is dangerous to completely rule out the possibility of equipment malfunction just because self-consumption has increased. There are cases where some equipment is simultaneously out of service during periods of increased self-consumption. For example, if on-site facility power usage rises and reduces exported power, and at the same time output drops occur in some generation systems, you cannot determine whether there is an abnormality based solely on exported power. An increase in self-consumption is a plausible explanation, but you also need to check measured generation output and the output of each piece of equipment.

Perspective 2: Overlaying Electricity Consumption by Time of Day with the Power Generation Curve

Checking by time of day is effective for distinguishing increased self-consumption from anomalies. Solar power generation generally ramps up in the morning, peaks during the day, and falls in the evening, forming a characteristic generation curve. The exact shape varies with weather, orientation, tilt, and shading conditions, but there is a consistent pattern on clear days. By overlaying this generation curve with the facility’s electricity consumption by time of day, it becomes easier to interpret why the amount of electricity sold has decreased.

For example, if the facility’s electricity consumption also increases during the generation peak period, the electricity generated may be used on-site, reducing the amount sold to the grid. When daytime air-conditioning operation, production line operation, pump operation, use of charging equipment, or the use of dryers and heating equipment coincide, self-consumption increases. In such cases, if the generation curve itself has not been significantly distorted, the reduction in sold electricity may be attributed to increased loads rather than an equipment malfunction.

On the other hand, if the generation curve’s peak is low despite clear skies, drops off suddenly partway through, shows no output only during specific time periods, fails to ramp up in the morning, or fluctuates greatly around midday, inspection of the equipment is necessary. If shadows fall during particular times, output may drop at similar times every day. If the power conditioner has shut down or is being controlled, output may decline in steps or level off at a fixed value. In the case of communication failures or faults in measuring devices, values may be missing on the monitoring screen even though generation is occurring.

When reviewing time-of-day data, check not only power generation but also the trends in electricity sold and electricity purchased. If, during sunny daytime, generation is sufficient, purchased electricity is suppressed, and only electricity sold is low, it may indicate increased self-consumption. Conversely, if generation is low, purchased electricity is high, and electricity sold is also low, this is difficult to explain by increased on-site demand alone and suggests a decline in output or a shutdown of the generation equipment.

Also, the judgment changes depending on whether the time period when you feel the power generation is low is in the morning and evening or during the daytime. In the morning and evening the solar altitude is low, and the system is more susceptible to the effects of shading, orientation, tilt, and surrounding structures, so output can appear lower than during the daytime. If it is low only in the morning or only in the evening, check shading and orientation conditions and any changes in the surroundings. If it is low even during the central daytime hours, a wider range of checks is necessary, such as dirt, equipment shutdowns, grid restrictions, or measurement errors.

When comparing by time of day, it is important to choose days with weather conditions as similar as possible. Comparing a cloudy day with a sunny day makes it difficult to judge equipment anomalies because the shape of the generation curve differs greatly. By comparing the most recent sunny day, a sunny day from the same month of the previous year, and a day on the same weekday with similar operating conditions, you can more easily separate the effects of self-consumption from changes on the generation side. If possible, also check records of solar irradiance and on-site weather notes to improve the accuracy of your assessment.

When looking at a power generation curve, it's important to consider not only the "total" but also the "shape." Even if daily generation is low, it may be a natural variation caused by bad weather. However, if the curve's shape is missing on sunny days, if it drops at the same time every time, or if only some of multiple systems are lower, those characteristics can be clues to an abnormality. Viewing an increase in self-consumption as likely to show up as a decrease in the amount of electricity sold, and equipment faults as likely to show up as disturbances in the generation curve, makes it easier to organize the order of checks.

Perspective 3: Isolating Changes in Weather, Seasonal Variation, and Load Equipment

When power generation is low, check the weather, the season, and changes in facility load before assuming an equipment fault. Because solar power generation is strongly affected by solar irradiance, output can temporarily drop during prolonged cloudy or rainy periods, seasons with a lot of yellow sand or pollen, or in regions affected by snow or frost. Also, during hot periods, rising panel temperatures can make it harder for output to increase. If these natural condition–driven variations are judged as abnormalities without first being understood, it can lead to unnecessary inspections.

Seasonal effects are reflected in both power generation and on-site self-consumption. In summer, longer hours of sunlight tend to increase generation, but increased use of air conditioning and ventilation can raise self-consumption. In winter, shorter hours of sunlight and a lower solar altitude make generation more likely to decline. At the same time, if the use of heating or frost-prevention equipment increases, the amount of electricity sold can appear even smaller. Because there are seasons in which reduced generation and increased self-consumption occur simultaneously, simple month-to-month comparisons alone cannot provide a definitive assessment.

Changes in the facility’s load equipment are also important. If new equipment is introduced, operating hours are extended, operations start on holidays, production processes are changed, or charging and heating are carried out during the daytime, on-site self-consumption increases. In particular, when more equipment operates continuously during the daytime, the peak period of solar power generation can coincide with the peak in power usage, causing the amount of electricity sold to the grid to drop significantly. In such cases, looking only at the amount sold can make generation appear low, but in reality the generated electricity may be being used effectively.

On the other hand, there are declines that are difficult to explain by weather or load changes. If power generation is clearly lower than the same period in the previous year despite continued sunny weather, if a specific section is lower compared with other installations at the same site, if there is a large difference between systems of the same capacity, if alerts are triggered, or if the generation curve is unnaturally interrupted, an inspection of the equipment is necessary. It is important to determine whether the decrease exceeds what can be explained by natural conditions or increased self-consumption.

To improve the accuracy of fault isolation, check facility operations during periods of low power generation. In practice, generation monitoring data and on-site operational information are often managed separately, and generation personnel may not be aware of operational changes on the facility side. When consulted about low power output, verify whether, during the relevant period, there were equipment additions, increased work activity, air-conditioning setting changes, holiday operation, temporary construction, or changes to battery charge/discharge operations.

Also, when assessing the impact of weather, it's safer not to judge solely by "sunny" or "cloudy." Even thin cloud cover reduces solar irradiance, and even on a sunny day output can fluctuate due to the temporary passage of clouds. After rain, dirt on the panel surface may be washed away, but depending on the surrounding environment, effects such as mud splashes, puddles, or humidity can also occur. It is important not to attribute low power generation to a single cause, but to verify by considering weather, season, load, and on-site conditions together.

To separate increased self-consumption from anomalies, it is essential to simultaneously consider both the generation-side conditions and the consumption-side conditions. Looking only at the generation side can cause you to overlook increases in self-consumption, while looking only at electricity usage can cause you to miss equipment abnormalities. Solar power generation must be evaluated not as an isolated piece of equipment but within the overall flow of power throughout the facility.

Viewpoint 4 Check for deviations in measurements and signs of equipment malfunction

If checking self-consumption, weather, and seasonal load still does not explain the low power generation, check for measurement drift or signs of equipment abnormalities. The important point here is that the values displayed on the monitoring screen do not necessarily reflect the actual on-site conditions. Even if the generation equipment is operating normally, problems with measuring instruments, communication equipment, electricity meters, or recording settings can make the reported generation appear low.

First, you should check whether there are any anomalies in the relationship between the measured power generation and the amounts sold, purchased, and self-consumed. If generation suddenly falls to near zero while purchased power and facility operation show no significant change during the same period, this could indicate not only an actual stoppage of generation but also measurement or communication failures. Conversely, if generation is occurring but the amount sold is extremely small, you need to distinguish between increased self-consumption, battery charging, and faults on the sales metering side.

When an equipment fault is suspected, check the output differences for each power conditioner, generation differences by system, alert history, and shutdown history. In facilities with multiple units, looking only at total generation can make localized problems less noticeable. Even if the facility as a whole is generating power, situations can occur where only a specific unit is stopped, only a particular string has low output, or part of a junction box is producing no current. When generation is low relative to plant capacity, it is important to examine not only the overall value but, as far as possible, the differences at the equipment and circuit levels.

During on-site inspections, some abnormalities can be identified by visual observation. Dirt on the panel surface, bird damage, fallen leaves, shadows from vegetation, shadows from surrounding structures, cable damage, abnormalities around junction boxes, equipment indicators, and ventilation conditions can all provide clues to reduced power generation. However, because there is a risk of electric shock and accidents when checking distribution boards, junction boxes, equipment interiors, and wiring, these should not be opened or worked on without the necessary knowledge and qualifications.

Site personnel should limit their work to visual inspections and data organization, and leave electrical checks and internal equipment inspections to qualified personnel or maintenance companies for safety.

A commonly overlooked source of discrepancy in measurements is the conditions used for period aggregation. If the daily, monthly, meter-reading periods, the monitoring system’s aggregation period, and the electricity sales statement’s applicable period are misaligned, reported generation may appear lower. For example, directly comparing generation from the first to the last day of a month with electricity sales based on meter-reading dates will produce a difference because the number of days covered differs. When comparing data, it is necessary to align the period, units, target equipment, and aggregation method.

Also, checking the installed capacity and the history of expansions is essential. When comparing with past data, if there were periods when part of the facility was shut down, periods when equipment was replaced, periods when operations were changed, or periods when measurement settings were altered, a simple comparison can be misleading. Before concluding that power generation is low, confirm whether the conditions being compared are the same.

Signs of anomalies often appear not as a single number but as multiple irregularities. If gaps in the power generation curve, differences in output between devices, alert history, on-site shading, missing measurements, and inconsistencies between sold and purchased electricity coincide, the priority for checking the equipment on-site should increase. Conversely, if the generation curve looks natural, there are no alerts, and daytime self-consumption has increased, a decrease in sold electricity is more likely due to higher self-consumption.

Perspective 5: Organizing Methods for Record-Keeping and Consultation Decisions

When power output is low, it is important to record the results of your checks. Even if you cannot determine on the spot whether the cause is increased self-consumption or an equipment malfunction, keeping data in a form that allows later comparison makes it easier to explain the situation when consulting maintenance providers or internal stakeholders. If records are insufficient, the same checks will be repeated multiple times, which delays identifying the root cause.

The items you should record are: the period when you felt power generation was low, the comparison reference, the weather, the facility’s operating status, generated power, self-consumption, power sold, power purchased, presence or absence of alerts, differences in output between devices, and the shadows and dirt observed on site. Organizing these in chronological order allows you to explain which figures changed under which conditions, rather than relying on mere impressions. In particular, it is important to record separately whether only the power sold is low, the generation itself is low, or self-consumption has increased.

When making consultation judgments, separate high-urgency cases from cases that can be monitored. If an alert has been triggered, equipment has stopped, there is a burning smell, you hear unusual noises, visible damage is present, or there are concerns about electric shock or leakage, do not attempt to handle it on site; promptly contact a qualified specialist. If there are signs related to safety, prioritize preventing accidents over whether power generation is high or low.

On the other hand, if the amount of electricity sold is low but the generation curve looks natural, daytime self-consumption has increased, and changes in facility operation can be confirmed, one option is to first look at several days to several weeks of data to check the trend. However, monitoring and leaving it alone are not the same. If you don't decide when, under what conditions, and which figures to recheck, detection of abnormalities can be delayed. Sharing scheduled check dates and the items to check among staff can reduce oversights.

When explaining things internally, rather than using only the expression "low power generation," it's easier to convey the situation if you make it concrete with phrases such as "electricity sold has decreased but self-consumption has increased," "generation itself is lower than the same period last year," "only the output of specific equipment is low," and "the generation curve on sunny days drops off partway." Using concrete expressions makes it easier to align the understanding of facility operators, maintenance staff, and management.

Also, for future reference, it is useful to have a baseline for normal conditions. Whether power generation is low cannot be determined without knowing the usual generation trends. By regularly keeping track of the standard generation curve on sunny days, monthly generation by season, self-consumption patterns on weekdays versus holidays, output differences between installations, and the relationship between electricity sold and purchased, you can notice anomalies sooner.

The way you keep records doesn't need to be complicated. What's important is that you can compare under the same conditions later. Recording the date, time period, weather, power generation, electricity sold, self-consumption, electricity purchased, facility operating status, and any observations in the same format will be helpful for monthly checks and maintenance inspections. Rather than recording only when power generation is low, keeping records from normal conditions as well makes it easier to see the difference from abnormalities.

Summary to Avoid Overlooking Low Power Output Conditions

When generation is low, it is important not to immediately assume that a drop in electricity sold indicates an equipment fault, and also not to completely rule out an equipment fault simply because self-consumption has increased. In solar power systems, when the electricity generated is used within the facility, the amount sold decreases. Therefore, the amount sold alone cannot accurately reflect the condition of the generation equipment. Looking at generated power, self-consumption, sold power, and purchased power together makes it easier to understand the flow of electricity.

To separate increased self-consumption from anomalies, first check changes in self-consumption and exported power over the same period. Next, overlay the time-of-day generation curve with electricity consumption to see whether an increase in daytime load is having an effect. Then check weather, season, facility operating conditions, and changes in load equipment; if the situation still cannot be explained, suspect measurement drift or signs of equipment malfunction. Verifying in this order makes it easier to narrow down the cause.

In practice, consultations about low power generation often begin with confusion between sold electricity, generated electricity, and self-consumption. Simply separating whether the amount sold is low, the generation itself is low, the amount of purchased electricity has increased, or daytime consumption has increased can greatly change the accuracy of the assessment. In particular, for self-consumption and surplus-feed-in systems, it is important to verify the facility’s overall electricity use together with the generation equipment, rather than checking the generation equipment alone.

However, if there are irregularities in the power generation curve, differences in output among equipment, alerts, missing measurements, or on-site soiling or shading, this may indicate an equipment fault. If there are signs related to safety or electrical verification is required, do not attempt to work on-site; consult qualified personnel or the maintenance company. Confirming a drop in power generation is essential not only to identify the cause quickly but also to operate the equipment safely.

To avoid overlooking periods of low power generation, it is effective to retain normal-condition data so you can compare it during abnormalities. By continuously checking the daily generation curve, monthly generation, self-consumption, electricity sold, electricity purchased, and facility operating status, you can calmly determine whether a drop in electricity sold is due to increased self-consumption or an equipment fault.

In managing solar power generation, it is important not to take the visible numbers at face value but to interpret which flow of electricity those numbers represent. When you feel generation is low, checking not only the amount sold to the grid but also self-consumption and equipment condition can reduce unnecessary worries and oversights. Keeping generation data and on-site operational information aligned on a daily basis leads to early detection of generation declines and safer equipment management.