6-Step Inspection Procedure to Prevent Power Output Decline from Being Overlooked

Even when they notice low power output from a solar power generation system, people may assume “it’s just the weather” or “it’s just a seasonal drop” and leave it alone. Indeed, output is affected by irradiance, temperature, season, shading, snowfall, dirt, and so on, so you cannot determine an anomaly based solely on day-to-day fluctuations. However, if a declining trend is left unchecked, minor soiling or partial equipment malfunctions can be overlooked, leading to a prolonged decrease in the amount of electricity sold and the amount used for self-consumption.

When responding to a drop in power generation, what’s important is not to immediately settle on a single cause, but to organize and inspect in the order of recording, comparison, on-site verification, equipment checks, and recurrence prevention. This article explains, in six steps, an inspection procedure to ensure reduced power generation is not left unaddressed at solar power plants and commercial solar photovoltaic installations, aimed at operational personnel searching for causes with the query "発電量　低い".

Table of Contents

• The first perspective to adopt to prevent declines in power generation from being left unaddressed

• Step 1 Check daily and monthly power generation

• Step 2 Separate weather and seasonal factors

• Step 3: Compare differences per power conditioner unit

• Step 4: Check for shadows and dirt around the panels on site

• Step 5 Check for abnormal indications on wiring and equipment

• Step 6: Keep inspection records to help prevent recurrence

• Operations for continuous monitoring of low power generation

• Summary

The First Perspective to Adopt to Prevent Leaving Power Output Declines Unaddressed

When you feel that power generation is low, the first thing to be aware of is not to judge the state of "low" based solely on a feeling. The power output of a solar PV system can vary greatly from day to day even with the same equipment. There is a difference between sunny and cloudy days, and even on equally sunny days the results can change depending on the solar altitude, ambient temperature, the presence or absence of wind, morning and evening clouds, and how shadows from the surroundings fall. For that reason, concluding that there is an abnormality based on the output of a single day can lead to unnecessary inspections or incorrect judgments.

On the other hand, you should avoid simply assuming it’s due to the weather and not checking anything. In particular, if the drop in power generation continues for several days to several weeks, if only certain sections within the same plant are noticeably lower, or if there is a large difference compared with previous days under the same conditions, some form of inspection or verification is necessary. Equipment faults do not always manifest as a complete shutdown. They can be hard to detect, appearing as reduced output in only some circuits, as impacts that occur only during specific time periods, or as alarms that temporarily trigger and then recover.

In practice, it is easier to treat a drop in power generation not as a task of immediately identifying the cause, but as a task of distinguishing whether it is within the range of normal variation or a decline that requires inspection. If you check, in order, the power output, irradiance conditions, differences between equipment, on-site conditions, and alarm history, you can more easily narrow down the possible causes. Conversely, if you go to the site without a set order or only look at the equipment, oversights and the need for rechecks tend to increase.

Also, to avoid leaving low power generation unaddressed, it is important in daily management to decide in advance how large a drop should prompt an inspection. Without a standard, judgments will vary by person. One person may immediately regard it as abnormal, while another may attribute it to seasonal factors and let it pass. In practice, decisions become more consistent when multiple comparison benchmarks are used, such as the same month in the previous year, the most recent sunny day, other units within the same plant, and similar nearby installations.

When inspecting a drop in power output, attention to safety is also essential. Solar power generation equipment is installed outdoors and carries risks such as working at heights, on slopes, in mud, among vegetation, and around electrical equipment. Even if you are eager to identify the cause of the drop, on-site checks must be conducted following safe procedures. Do not forcefully handle things like footing after rain, areas around panels during strong winds, or the doors and terminals of electrical equipment; when necessary, decide to consult specialist contractors or maintenance personnel.

As such, when power generation is low, it is important not to hastily assume a single cause, but to verify items sequentially from the records. From the next chapter, we will organize in detail the items to be checked, following six inspection procedures that are easy to apply in practice.

Step 1 Check daily and monthly power generation

The first thing to do is check the power generation records on a daily and monthly basis. When you feel the power generation is low, organizing the generation data before going to the site makes it easier to identify the scope and timing of inspections. If you only look at daily generation figures, they are heavily affected by the weather, but looking at monthly or weekly data makes it easier to detect a continuous downward trend.

First, check the power generation not only for the past few days but for at least several weeks. If generation is low for only a single day, it may be due to the weather or temporary factors. However, if low generation continues even when sunny days are included, you should suspect dirt, shading, equipment outages or shutdowns, or malfunction in some circuits. When looking at daily data, be aware not only of the absolute level of generation but also the day the decline began, whether it dropped suddenly or gradually, and whether it is biased toward specific time periods.

When checking by month, compare with the same month in the previous year and with the same season in past years. Solar power generation trends change with the seasons. In winter, shorter sunlight hours and a lower solar altitude make the effects of shading more likely to appear. In summer, although the amount of solar radiation is greater, rising temperatures can increase panel temperature and lead to days when output is less likely to rise. Therefore, rather than simply thinking “it’s low even though it’s summer” or “it can’t be helped because it’s winter,” it is important to check whether there is an unusual difference compared with the same period in past years.

When checking power generation, it's easier to compare if you look at generation per unit of installed capacity. If you manage multiple power plants or several sites, differences in facility size make it difficult to judge based on absolute generation values alone. Converting to generation per unit of installed capacity normalizes scale differences to some extent and allows for comparison. However, because orientation, tilt angle, shading conditions, and installation region differ and will affect the results, treat this only as a guideline for identifying potential anomalies.

Also, check the shape of the power generation graph. On clear days, the generation curve generally rises from the morning, peaks around midday, and falls in the evening. On cloudy days it fluctuates up and down, but if you see patterns such as a sudden drop during a clear day, an unusually low output for part of the midday, or output being reduced at the same time every day, consider possibilities such as shading, equipment control, grid-side constraints, or equipment shutdowns. Abnormalities that cannot be identified from simple totals alone are often easier to find by examining the breakdown by time of day.

At this stage, it is not necessary to determine the cause. What matters is clarifying whether the drop is temporary or persistent, whether it is occurring system-wide or only in part, and when it began. What you confirm here will form the basis for subsequent weather checks, equipment comparisons, and on-site inspections. By grasping data trends beforehand, rather than proceeding to on-site checks without reviewing the records, you can reduce wasted inspection effort and make oversights less likely.

Step 2: Separate weather and seasonal factors

When checking the causes of low power generation, distinguishing between weather and seasonal factors is essential. Solar power generation depends on solar radiation, so output fluctuates due to cloud cover, rain, snowfall, yellow sand, fog, and morning or evening clouds. Before determining that a reading is abnormal based only on the power output, you need to confirm what the weather conditions were like on that day.

First, check the sunlight conditions on the days when power generation was low. Even if it looked sunny near the plant, generation can drop if thin clouds lingered for a long time or if it was cloudy only in the morning. Conversely, when the manager is located away from the site, the weather around the office can differ from the weather at the plant. In particular, cloud cover can vary within the same area in mountainous regions, coastal areas, basins, and on large sites. For this reason, it is important to check weather information close to the site together with fluctuations in the generation curve.

Next, consider the differences in daylight hours and solar altitude by season. In winter, daylight hours are shorter and the sun’s angle is lower, so shadows from nearby trees, buildings, utility poles, and between rows of mounting racks tend to extend farther. Shadows that were not a problem in summer can cause reduced power generation in winter. Spring and autumn are periods when the sun’s angle is changing, and morning and evening shadows that previously went unnoticed may become more noticeable. If the period when you felt the power generation was low coincides with a change of season, it is worth focusing on checking how shadows fall.

The effect of temperature should not be overlooked. Solar panels generate more easily when solar irradiance is strong, but when panel temperature is high the power output can become harder to increase. As a result, even on summer days with strong sunlight you may not see as much generation as you would expect compared with clear days in spring or autumn. To avoid misidentifying this as an anomaly, it is important to understand seasonal generation trends as well as simple solar irradiance strength.

On the other hand, there are declines that are difficult to explain by weather or season alone. For example, if a particular power conditioner shows lower power output on the same sunny day, or there is a clear difference compared with nearby installations of a similar scale, you should not dismiss it as merely weather-related. Also, after rain, dirt may be temporarily washed away and power output may recover, while in other cases splashed mud or fallen leaves remain and only some parts continue to show reduced output. Weather can help explain the cause, but it must not be used as an excuse to overlook declines.

Accumulated snow, ashfall, dust, pollen, bird droppings, and fallen leaves can also have effects depending on the season and region. In particular, soiling that covers part of the panel surface can sometimes affect power output. Even small-looking dirt can lead to a localized drop in output depending on the arrangement of cells and circuits. Because such soiling is difficult to identify from power generation data alone, it should be checked during subsequent on-site inspections.

The purpose of separating weather and seasonal factors is not to skip inspections but to determine their priority. If the day with reduced output was clearly due to bad weather and generation recovers on the following sunny day, the situation may not be urgent. However, if output does not recover even on sunny days, if the difference is large compared with days under the same conditions, or if only specific equipment shows lower output, then it is necessary to proceed to equipment-level comparisons as the next step.

Step 3: Compare differences at the power conditioner unit level

Comparing power conditioners on a unit-by-unit basis is effective for narrowing down the causes of reduced power generation. If you only look at the plant’s total power generation, it can be difficult to tell where the decline is occurring. Even if the overall output appears slightly low, in reality only some power conditioners or certain circuits may have experienced a large drop. Conversely, if the whole system drops similarly, you are more likely to suspect weather, solar irradiance, grid-side constraints, or overall soiling.

First, at installations with multiple power conditioners, compare the generation of each unit for the same day. If, despite being installed on the same site under similar conditions, only one unit shows lower generation, you should check the panels, wiring, input circuits, and the condition of the unit itself connected to that unit. Differences that are easy to overlook when looking only at the simple total generation can be detected at an earlier stage by comparing each unit.

When making comparisons, take differences in equipment capacity into account. If the number of panels or their capacities connected to each power conditioner differ, comparing absolute generation values as they are will not lead to a correct judgment. Equipment with larger capacity will produce more power, and equipment with smaller capacity will produce less. Therefore, it is important to compare generation per unit of capacity and to compare with the historical generation trends of the same equipment. If a device that had shown similar trends until yesterday suddenly drops, it should be prioritized for inspection.

Comparisons by time of day are also helpful. If only one power conditioner has a slow ramp-up in the morning, drops output around midday, or is low only in the evening, you can consider possibilities such as shading, temperature, input circuits, or equipment control. If it drops at the same time every day, the relationship between the sun’s position and shadows may be the cause. If the timing is irregular and the output suddenly falls or recovers, also check for poor contacts, equipment protection operations, communication failures, or changes in external conditions.

When comparing differences in power generation, attention must also be paid to missing communication data. Even if generation is actually occurring, faults in monitoring devices or communication can prevent data from being collected. In such cases, the output may appear low, but it might not be due to an actual drop in the equipment’s output. Do not judge based solely on the values on the monitoring screen; as needed, cross-check multiple sources of information such as on-site displays, data loggers, and export meters to help avoid misinterpretation.

Also, even if there is no difference at the power conditioner unit level, you should not be completely reassured. If the whole system is uniformly underperforming, a common cause should be considered. For example: dust or pollen deposited across the entire site, surrounding vegetation having grown and creating overall shading, adverse weather conditions, or output being curtailed due to grid-side constraints. In other words, whether there are differences between individual devices is an important factor in deciding whether to investigate causes locally or system-wide.

What’s important in this step is to narrow down the suspected area as much as possible. Rather than conducting a blanket inspection of the entire power plant, preselecting the equipment, time periods, circuits, and zones where performance declines are observed will make on-site verification more efficient. In the next step, based on the suspicious areas identified in the data, you will check for shading and soiling around the panels.

Step 4: Confirm shadows and dirt around the panels on site

After checking the power generation data and differences among equipment, the next step is to inspect the site for shadows and dirt around the panels. While shadows and dirt are among the causes of low output that are easy to detect in the field, they are difficult to judge from data alone. In particular, if part of a solar panel is shaded or if dirt remains on part of a panel, it can affect power generation.

During on-site inspections, first observe the overall condition of the panel surfaces from both a distance and up close. From a distance, it is easier to recognize differences in color by row, bands of dirt, accumulations of fallen leaves, concentrations of bird droppings, soil dust adhesion, and the extent of weed growth. Up close, check whether mud has accumulated on the lower edges of the panels, whether there are localized deposits on the surface, and whether moisture or deposits remain around the frames. However, avoid stepping onto the panel surfaces or touching electrical equipment inadvertently; checks should be carried out within a safe range.

When checking for shadows, it is important to be aware of the time of day. Even if there are no shadows at the time you visit the site, significant shading can occur in the morning or evening. In winter, because the sun’s altitude is low, shadows from surrounding trees, fences, buildings, utility poles, mounting racks, and adjacent equipment extend long. If power generation data shows a drop during a specific time period, inspecting the site under conditions close to that time makes it easier to find the cause. If the time of the site inspection differs from the time when the generation drop occurs, do not make judgments based only on photos or records; recheck as necessary.

Vegetation growth can also cause a reduction in power generation. Even in locations that posed no problems at installation, weeds and trees can grow over months to years and cast shadows on the lower parts of panels or on portions of rows. Slopes, site boundaries, drainage channels, and areas around fences are particularly prone to vegetation growth, and if maintenance is delayed the shaded area can expand. Even when mowing is performed, it is important to check after the work, because missed patches or grass that falls over after cutting can end up on the panels.

Regarding soiling, we distinguish between uniform light soiling across the entire surface and localized heavy soiling. If sand dust or pollen is deposited across the whole site, the plant’s overall power output may gradually decline. On the other hand, if bird droppings, fallen leaves, mud splatter, or the like remain in certain areas, they can appear as reduced output from specific panels or circuits. Some soiling is naturally washed away by rain, but depending on how the rain strikes and the panel angle, dirt can remain on the lower edges. On site, we verify which areas and to what extent are soiled, and whether that corresponds with the areas of reduced generation in the performance data.

Checking drainage and ground conditions can help you understand the causes of reduced power output. In locations prone to puddles, mud splatter increases and moisture makes weeds more likely to grow. On sites where strong winds lift sand, panel surfaces can become dirty in a short time. Nearby construction, farming activities, or traffic on unpaved roads also make dust more likely to accumulate. To prevent recurring low power output, it is important not only to identify dirt but also to understand the environment that causes it.

Shadows and soiling found during on-site inspections should be recorded together with photos, the time, the location, and the relevant equipment. Ensuring these can later be cross-referenced with power generation data increases the accuracy of cause determination. If records are vague, it becomes difficult to confirm the effect after cleaning or mowing. In operations that do not tolerate reduced power output, it is important not to let on-site observations remain one-off but to retain them linked to the data.

Step 5: Check Wiring and Equipment for Abnormal Indications

If inspecting for shadows or dirt does not reveal a clear cause, or if there are discrepancies at the power conditioner unit level, check the wiring and equipment for abnormal indications. In solar power generation systems, not only the panels but multiple elements are involved, such as junction boxes, combiner boxes, switches, power conditioners, monitoring devices, and communication equipment. Even if only part of the system is malfunctioning, it may manifest as a reduction in power generation rather than a complete shutdown.

The first thing to check is whether the power conditioner or monitoring device shows or logs any alarms, shutdowns, curtailments, or communication errors. Even if no alarm is currently displayed, one may have occurred in the past and then automatically cleared. If the times when power generation was low coincide with timestamps in the alarm history, that can help narrow down the cause. When checking the display contents, do not rely solely on terminology; confirm the meanings according to the equipment manual and maintenance standards.

Next, check for differences between input circuits. The information available varies by device, but if there are large differences in voltage or current by string, input branch, or circuit, you should inspect the panels, wiring, connections, and switches connected to that circuit. If only some of multiple circuits are low, this may correspond to the extent of shading or soiling, or it may be related to poor connections or equipment-side problems. Electrical checks involve risk, so they must be performed by qualified and authorized personnel following safety procedures.

Around wiring, visual abnormalities are also subject to inspection. Check for sagging cables, damaged protective conduits, traces of water around junction boxes, suspected animal damage, discoloration around terminals, unusual odors, unusual noises, and the like. However, avoid opening electrical equipment or touching terminals unnecessarily. If an abnormality is suspected during the visual inspection, hand it over to specialized maintenance personnel, who will carry out the necessary measurements and repairs.

Consider grid-side impacts and output curtailment. Even if the equipment itself shows no major abnormalities, output may be curtailed by grid conditions or controls. In such cases, generation may not increase even when there is available solar irradiance. Check whether there is a history of curtailment or control, which time periods it occurs in, and whether it matches the shape of the generation curve. Be careful not to confuse equipment faults with decreases caused by control, as that can lead to unnecessary on-site work or incorrect countermeasures.

Communication failures can sometimes look like a drop in power generation. If monitoring data is missing, the displayed value may be lower than the actual power generated. If data is missing only during certain time periods, multiple devices show anomalous values at the same time, or on-site displays and remote monitoring values do not match, it is necessary to check the communication equipment and recording devices. Determining whether the issue is a malfunction of the power generation equipment itself or a problem with data acquisition will change the priority of the response.

When checking equipment, it is important not to assume that a drop in power generation is a "panel problem." Even if no abnormalities are visible on the panel surface, the cause may lie in the equipment, wiring, communications, or controls. Conversely, it is premature to conclude equipment failure based solely on an alarm display. Because the alarm may have been triggered by the surrounding environment or temporary conditions, decisions should be made by combining generation data, on-site conditions, and historical records. While prioritizing safety, it is important to distinguish between what can be checked and what requires specialist intervention.

Step 6: Keep inspection records to help prevent recurrence

At the end of an inspection for a decline in power generation, the most important thing is to keep inspection records and use them to prevent recurrence. Even if you identify the cause and take temporary measures, if no record remains you will have to investigate from scratch again when a similar decline occurs. In practice, not only the inspection itself but also organizing information in a form that can be checked later determines the quality of power generation management.

Inspection records should include the date when low power generation was noticed, the period of data that was checked, the weather, the equipment or sections with reduced output, shadows or dirt found on site, alarm history, actions taken, and changes in generation after those actions. In addition to written descriptions, saving photos and screen captures together makes it easier to recreate the situation later. Add information to photos that identifies the shooting date, shooting time, shooting location, and the equipment depicted, as this makes it easier to cross-check with generation data.

When recording, it is important to separate causes that have been confirmed from those that remain possible. For example, if a panel is dirty and power generation recovers after cleaning, it can be regarded as likely that the dirt was the main factor. On the other hand, if output does not fully return after cleaning and differences between units remain, continued checks for causes other than dirt are necessary. If you record something that is not confirmed as if it were certain, it can lead to incorrect judgments at the next inspection.

To prevent recurrence, inspection results are incorporated into operational rules. If shadows from vegetation were the cause, the timing and frequency of mowing and the priority areas will be reviewed. If fallen leaves or mud splashes were the cause, checks of the surrounding environment and drainage will be added to regular inspections. If alarms are repeatedly triggered on a particular device, it will be subjected to detailed checks separate from routine inspections. If missing communication data was the cause, checks of monitoring equipment and the communication environment will be incorporated into operations.

Confirming the effect after corrective actions is also important. After cleaning, mowing, equipment restoration, wiring repairs, and similar work, check whether power generation has recovered. If the weather conditions before and after the actions differ greatly, a simple comparison can be difficult. Compare as much as possible on days with similar weather, at the same time of day, and for the same equipment unit to see whether there is a trend of improvement. If no improvement is seen despite taking action, another cause may remain.

By accumulating inspection records, the weaknesses of each power plant become apparent. At one plant, shadows tend to appear on winter mornings; at another, the effects of grass are more likely to appear after the rainy season; at another facility, communication anomalies are more likely to occur—these kinds of trends can be identified. Once such trends are understood, inspections and countermeasures can be carried out before the periods when power generation is likely to decline, rather than scrambling to respond after output has already fallen.

To avoid allowing declines in power generation to go unchecked, it is important not to treat inspections as one-off tasks. By reflecting inspection results in subsequent site patrols, cleaning, grass cutting, equipment checks, and monitoring settings, power generation management will gradually stabilize. If records are well organized, decisions are easier to hand over when personnel change, making it easier to move away from person-dependent management.

Operations for Continuous Monitoring of Low Power Output

Even if you carry out the six inspection procedures, responding to a drop in power generation does not end there. Solar photovoltaic systems are installed outdoors, so shadows, dirt, vegetation, weather, and equipment condition are constantly changing. Even if an inspection found no problems at one time, the situation can change after a few months. To avoid leaving low generation unaddressed, operations need to combine daily monitoring with regular checks.

In continuous monitoring, you first establish the baseline power generation trends. For each installation, observing generation on sunny days, seasonal trends, differences between power conditioners, and how generation ramps up in the morning and evening makes it easier to detect anomalies. Without a baseline, it is difficult to determine whether generation is low or simply normal for the season. It is important not just to store daily generation data, but to review it in a form that makes comparisons easy.

To detect a drop in power generation early, it is also effective to set specific times for checks. Even if it is difficult to inspect in detail every day, deciding on weekly checks, monthly checks, and checks at seasonal transitions makes it easier to prevent long periods of neglect. In particular, during the rainy season, after typhoons, after snowfall, in periods with heavy yellow dust or pollen, when vegetation is growing, or when there is nearby construction, conditions that tend to reduce power generation are more likely to coincide, so it is advisable to raise the priority of inspections.

It is also important to distinguish between on-site inspections and remote monitoring. Remote monitoring allows you to check power generation and alarm histories, but it may not be sufficient to assess panel surface soiling, vegetation overgrowth, nearby shading, drainage conditions, or the situation around fences. Conversely, relying solely on on-site inspections can make it easy to overlook time-of-day output variations and comparisons with past data. Combining both makes it easier to link anomalies in the data with actual conditions on site.

When there are multiple operational staff, it is also important to share the decision criteria. If you define who looks at which data, how far they should investigate, and at what stage cases should be escalated to specialists when a drop in generation is suspected, you can reduce delays and duplicated work. If each person applies different criteria, minor declines may be overlooked or, conversely, unnecessary on-site checks may increase. By keeping operating rules concise, you can improve responsiveness to declines in power generation while reducing the burden of daily management.

Furthermore, information about a reduction in power generation can be used not only for troubleshooting but also for equipment improvement. If locations prone to shading can be identified, that information can be used for vegetation management and reviewing the surrounding environment. If sections prone to soiling can be identified, they can serve as material for considering cleaning schedules and drainage measures. If differences between devices are consistently observable, they can be used as a basis for maintenance and replacement planning. It is important to treat the phenomenon of low power generation as an opportunity to learn the weaknesses of the equipment.

When conducting continuous monitoring, be careful not to focus too much on power generation alone. Power generation is an important metric, but it is affected by weather and seasonal factors, so it cannot always be evaluated on its own. By looking at power generation together with solar irradiance, differences among equipment, alarm history, on-site photos, and maintenance records, you can get closer to the root cause. Establishing processes that link multiple sources of information leads to earlier detection of declines in power generation and helps prevent recurrence.

Summary

When you notice low power generation, it is important not to immediately attribute it to the weather but to carry out an inspection in a systematic order. First, check daily and monthly generation to see whether the decline is temporary or persistent. Next, isolate weather and seasonal factors to determine whether the decrease should be regarded as abnormal. Then check for differences between power conditioner units to judge whether the decline is occurring across the entire system or is localized to certain equipment or zones.

On site, check for shadows, dirt, vegetation, fallen leaves, mud splashes, and the surrounding environment around the panels. By paying attention to suspicious time periods or areas in the data, it becomes easier to identify the cause. If shadows or dirt alone cannot explain the issue, check wiring, abnormal equipment indicators, alarm history, communication status, and whether output control is in effect. When inspecting anything related to electrical equipment, prioritize safety and, for the necessary scope, hand the work over to specialized maintenance personnel.

Furthermore, it is essential to record inspection results and use them to prevent recurrence. If you document the causes of power generation declines, the actions taken, and the changes observed afterward, decision-making will be quicker in future inspections. As trends such as shadows from vegetation, soiling, equipment variation, and communication anomalies accumulate, you will be able not only to respond after power output has fallen but also to carry out preventive management before periods when declines are likely.

An operation that does not leave declines in power generation unaddressed is not sustained by special tasks alone. By accumulating daily data checks, regular on-site inspections, understanding differences between pieces of equipment, and organizing inspection records, you will be able to grasp the condition of the power plant more quickly. When you feel that generation output is low, verifying each item one by one based on records and on-site conditions—rather than relying on intuition—is the quickest route to stable generation management.

If you want to detect declines in power generation at a solar power plant early and manage them by linking on-site conditions with data, it is effective to establish a system that makes inspection records and location information easy to use. By implementing operations that centrally handle changes in power generation, on-site photographs, anomaly locations, and response histories, you can reduce oversights and delayed responses. For site management that avoids leaving declines in power generation unaddressed, building a system that continuously connects data with on-site records is important.