Five Measures to Reduce Errors in Solar Power Generation Simulations

Solar power generation simulations are indispensable for planning power generation projects, designing equipment, making investment decisions, and explaining plans to financial institutions and internal stakeholders. By grasping in advance how much annual generation can be expected, the seasonal generation trends, and the expected revenue from sales and effects of self-consumption, it becomes easier to assess the validity of a business plan.

On the other hand, simulation results are forecasts and do not exactly match actual generation. Many factors affect generation—insolation, temperature, shading, installation angle, equipment performance, soiling, snow accumulation, surrounding environment, and the state of operations and maintenance. If calculations are performed with coarse input conditions, the numbers may look neat on paper but can differ substantially after actual operation.

What matters is not making the error zero, but understanding the causes of errors and carefully eliminating the parts that can be reduced in advance. Especially for practitioners, it is important not only to look at the simulation numbers themselves but also to be able to explain the assumptions under which those numbers were calculated and where uncertainties remain.

Table of Contents

• Why do errors occur in solar power generation simulations?

• Measure 1: Bring insolation data and meteorological assumptions closer to the site

• Measure 2: Accurately reflect installation angle, orientation, and shading conditions

• Measure 3: Set equipment performance and degradation realistically

• Measure 4: Reduce input discrepancies with on-site surveys and surrounding environment investigations

• Measure 5: Revise simulations using operational performance data

• How to make revenue decisions and explain internally assuming errors

• Summary: Generation forecasts change greatly with input accuracy and site understanding

Why do errors occur in solar power generation simulations?

Errors in solar power generation simulations do not arise simply because the calculation method is wrong. In many cases they accumulate from differences between input assumptions and actual site conditions. Simulations calculate generation based on assumptions such as insolation, temperature, panel capacity, installation angle, orientation, shading, and loss rates. Therefore, if assumptions diverge from the reality at the site, the results will also diverge from reality.

For example, even within the same region, coastal areas, mountainous regions, urban areas, industrial zones, and farmland have different meteorological conditions and surrounding environments. Even if two locations are close on a map, differences like a tendency to fog, longer-lasting snow cover, mountain shadows in the morning and evening, or reflections and obstructions from nearby buildings can affect generation. Using only average meteorological data may fail to capture such local differences.

Installation conditions of solar panels are also a major source of error. Although drawings may show a south-facing orientation at an optimal angle, in reality roof pitch, mounting errors, ground level differences, and connections with adjacent equipment can slightly alter angle and orientation. For large ground-mounted systems, row spacing, terrain slope, and post-development ground shape can change the times when front rows cast shadows on rear rows.

Assumptions about equipment performance are important too. Solar panels and inverters have rated performance, but in real generation environments temperature rise, wiring losses, conversion losses, soiling, degradation, and downtime occur. Underestimating these leads to higher simulated generation and wider gaps with actual performance. Conversely, being overly conservative can unduly lower perceived project viability.

To use generation simulations in practice, you must avoid treating calculated results as absolute and instead understand which conditions strongly influence outcomes. The first step to reducing error is not fine-tuning numbers but organizing points where errors tend to occur and improving the reliability of input values through on-site checks and data scrutiny.

Measure 1: Bring insolation data and meteorological assumptions closer to the site

The most fundamental element in solar power generation simulation is insolation data. Since solar power converts energy received from the sun into electricity, deviations in assumed insolation lead to large deviations in predicted generation. In practice, regional average insolation data and historical meteorological data are commonly used, but results can vary depending on which data are chosen.

To reduce errors, first understand the nature of the insolation data being used. Check whether long-term averages are used, whether recent years are reflected, how close the observation point is to the planned site, and what assumptions are used when converting horizontal-plane insolation to tilted-plane insolation. In particular, in mountainous or coastal areas the weather patterns at the nearest observation point may differ from those at the project site. If an observation point is on flat ground but the planned site is on a mountain slope, treating them as the same region can lead to large errors.

Temperature assumptions should not be overlooked. Solar panels do not always perform better simply because insolation is strong; panel efficiency decreases as panel temperature rises. In regions with high summer insolation, temperature losses can be significant depending on ambient temperature and ventilation conditions at the installation surface. Panel temperature behavior differs between roof-mounted systems close to the roof and ground-mounted systems with good airflow.

In regions affected by snowfall, consider not only annual insolation but also winter stoppages and snow persistence. Simulations may assume some generation in winter, but in reality snow remaining on panels can cause near-zero generation for several days. Whether panels are installed at an angle that allows snow to slide off, whether nearby structures trap snow, and whether snow removal will be performed all affect actual performance.

To bring insolation and meteorological assumptions closer to the site, it is effective to compare multiple data sources where possible. Rather than using a single dataset, cross-reference trends in neighboring areas, recent years’ meteorological variations, local terrain conditions, and performance trends of nearby operating systems. If multiple lines of evidence show no major contradictions, the reliability of simulation results tends to increase.

In practice, it is also important to check monthly generation, not just annual totals. Annual figures can look reasonable while seasonal distributions are inconsistent with local climate. For example, overestimated generation during the rainy season, unreflected winter snow impacts, or failure to account for autumn typhoons and cloudy tendencies can indicate hidden error sources behind annual values.

Measure 2: Accurately reflect installation angle, orientation, and shading conditions

Installation angle, orientation, and shading are major factors affecting solar generation. Even with appropriate insolation data, simulation accuracy falls if panel orientation and angle are not accurately reflected. This is especially important for roof-mounted systems and ground-mounted systems on complex terrain.

Installation orientation indicates which direction panels face. Generally, south-facing is considered favorable for generation, but optimal placement depends on site shape, roof geometry, power demand timing, and nearby shading. East- or west-facing installations shift peak generation times. When prioritizing self-consumption, it is necessary to consider the match between generation timing and power usage, not just maximizing annual generation.

Installation angle is equally important. A shallow angle receives more summer insolation but may retain soiling; a steeper angle receives more winter insolation and may shed snow more easily, but wind load and mounting structure considerations arise. Simulations input angle as a numeric value, but on-site roof pitch and mounting constraints may cause actual angles to differ from planned values.

Shading conditions are a representative source of error. In solar systems, even partial shading of a panel can significantly reduce generation depending on circuit configuration. Shading comes from buildings, trees, utility poles, signs, mountains, adjacent equipment, rooftop HVAC, railings, lightning protection, and more. Shading changes by time of day and season. Conducting a single daytime site visit can miss shadows occurring in the morning, evening, or winter.

For ground-mounted systems, row-to-row shading is also important. If row spacing is insufficient, front-row shadows can affect rear rows during periods of low solar altitude. Even if design appears acceptable, actual ground level differences can alter shadow behavior. Judging only from pre-development terrain data or schematic maps may fail to reflect final panel heights and slopes, resulting in prediction errors.

Reducing shading errors requires three-dimensional understanding of the site. A plan view alone cannot sufficiently evaluate adjacent building heights, tree growth, slope inclination, and rooftop equipment rises. Combining site photos, survey data, elevation differences, and positions and heights of surrounding structures to verify seasonal shading is important. For large projects or complex environments, overlooking shading can greatly affect annual generation.

Also, shading conditions can change after planning. New buildings, growing trees, or added nearby equipment may reduce generation after operation starts. While impossible to predict perfectly, checking local land use and future development potential helps assess such risks.

Measure 3: Set equipment performance and degradation realistically

Simulations require input of equipment performance for panels, inverters, and other components. Using overly ideal values here can produce simulated generation higher than reality. To reduce errors, realistically reflect not only rated performance but also operational losses and long-term degradation.

Panel output is based on values measured under standard test conditions. However, actual outdoor conditions rarely match those standards. Generation efficiency varies with angle of incidence, panel temperature, wind speed, soiling, diffuse light on cloudy days, and wiring conditions. Temperature losses are often overlooked: panels heat up more on clear, high-insolation days, which can reduce efficiency.

Inverter conversion efficiency is also important. DC power from panels is converted to AC for use or sale, and losses occur during conversion. Inverters have input ranges and output limits; depending on the capacity configuration, output may plateau during high-insolation periods. If simulations do not reflect these limits, peak generation may be overestimated.

Wiring losses cannot be ignored. Losses arise from cable length, thickness, connection methods, and distances to combiner or distribution boxes. While the impact may seem limited in small systems, in large systems with long wiring distances these accumulate and affect annual generation. Early-design estimates may not have finalized wiring routes, so reconfirmation is needed once design is concrete.

Degradation assumptions directly affect business planning. Panels are used over long periods and gradually lose output. Focusing only on first-year generation in simulations risks overestimating long-term profitability. When evaluating generation over the project life, set annual degradation rates and assess cumulative generation and revenue impacts.

Soiling and maintenance states should be considered realistically. Dust, pollen, bird droppings, fallen leaves, and exhaust-related deposits on panel surfaces reduce generation. Rain may wash some soiling away, but shallow installation angles or surrounding conditions can cause soiling to persist. Industrial areas, farmland, roadsides, and coastal locations are more prone to soiling and salt effects.

Assumptions about downtime are also important. In practice, periods without generation occur due to inspections, failures, communication issues, grid constraints, or protective device operations. Assuming continuous normal operation in simulations will create gaps with actual performance. It is not necessary to be overly pessimistic, but considering a realistic level of downtime risks according to system size and operational setup is prudent.

To set equipment performance realistically, do more than input specification values: estimate losses according to installation environment and operation conditions. Cross-check manufacturer specs, design assumptions, past project data, and operational data from similar facilities to ensure no overestimation. Simulations should improve decision-making accuracy, not inflate numbers.

Measure 4: Reduce input discrepancies with on-site surveys and surrounding environment investigations

On-site surveys and surrounding environment investigations are extremely important for reducing simulation errors. Simulations consider site area, shape, elevation differences, slope, surrounding structures, and installable area when planning equipment layout. If these data are inaccurate, layout plans can mismatch the site and affect generation forecasts.

Relying only on drawings and map data can miss subtle on-site conditions. For example, plot boundaries may be unclear and the actual installable area smaller than assumed; there may be greater elevation differences than expected; drainage channels or slopes may prevent placing mounts; nearby trees or structures may exist; vehicle access and maintenance space may need to be secured. These conditions affect equipment layout, number of panels, tilt, and row spacing.

The same applies to roof installations. Judging by roof area on drawings alone can overlook rooftop equipment, inspection walkways, lightning protection, upstands, load limits, and areas where waterproofing must be preserved, all of which limit installable area. Roof pitch and orientation may also differ slightly from drawings. Calculating without confirming on-site dimensions and heights leads to discrepancies in panel layout and shading conditions.

For ground-mounted systems, understanding elevation differences is especially important. Plan views alone cannot fully capture site undulations and slopes. Even small elevation differences impact row shading, rack height, earthworks volume, and drainage planning. Whether terrain is a former forest, agricultural conversion, fill site, or sloped land, caution is needed. Treating topography as a flat plane in simulations versus accounting for actual slope can change outcomes.

Surrounding environment surveys should consider not only current obstacles but also future changes. Trees grow and neighboring land use may change. A site with little shading at the planning stage may see increased shading in a few years as trees grow. Expected modifications or new construction nearby should be noted as long-term risks to generation.

Higher precision in on-site surveys benefits not only simulations but also design, construction, and maintenance. Accurate position and elevation data facilitate panel layout studies, shading checks, cable route planning, maintenance access, and as-built verification. Conversely, vague initial site information often leads to design changes later, requiring revisions to generation forecasts and business plans.

Practitioners should focus on information gathering before inputting data into simulation software. No matter how advanced the calculation, if the site condition inputs are coarse, output reliability will not improve. Accurately capture site location, shape, elevation differences, obstacles, and surrounding environment as much as possible, and reflect that information in simulation conditions—this is the basics of reducing errors.

Measure 5: Revise simulations using operational performance data

Simulations are not finished after being produced at the planning stage. Reviewing them against operational performance after commissioning improves forecast accuracy and helps quickly identify equipment anomalies and areas for improvement. This is especially important when managing multiple plants or planning new projects—operational data should inform future simulations.

After operation begins, check more than just annual generation. Monthly, daily, and hourly generation data reveal when deviations from forecasts occur. Annual totals may appear acceptable while showing seasonal biases, such as generation being lower than predicted in summer and higher in winter. Such trends point to causes like temperature losses, snow, shading, soiling, output control, or equipment downtime.

When comparing to actuals, consider weather in that period rather than simply comparing generation numbers. If insolation is lower than average that year, lower generation than predicted is natural. Conversely, if weather conditions were favorable but generation is low, the cause may be equipment or environmental. If possible, evaluate actual generation together with actual insolation to better isolate causes.

Typical issues revealed by operational data include output declines due to soiling, shading effects at specific times, inverter downtime, communication failures, output limits, and string-level anomalies. For example, low generation only in the morning on clear days suggests an eastern shadow; low generation only in the evening suggests a western shadow; if only certain equipment performs poorly, suspect device or wiring faults. Differences between simulation and actuals can serve as leads for equipment management.

Accumulating operational experience allows you to improve simulation assumptions next time. Insights such as larger-than-expected winter snow losses in a region, certain tilt angles retaining more soiling, need for salt-impact maintenance on coastal sites, or more influential morning/evening shading in mountainous areas become concrete only with actual data. Reflecting your company’s or facility’s operational experience rather than only desk-based standard values increases prediction accuracy.

Reviewing simulations after operation also helps explain results to stakeholders. If you cannot explain deviations between planned simulations and actuals, trust in the simulation declines. But if you can categorize deviations into weather, equipment, environment, and operation factors, you can treat them as actionable issues rather than mere forecast misses. Managing generation encompasses not only creating numbers at the planning stage but also verification after operation.

Revising simulations need not be done excessively often, but it is worthwhile at least after the first year of operation, once seasonal trends are visible, and after equipment changes or surrounding environment changes. Updating assumptions based on actuals and feeding them into future decisions raises the overall management accuracy of generation projects.

How to make revenue decisions and explain internally assuming errors

While striving to reduce errors in simulations, it is also important to make revenue decisions assuming that errors will remain. Because weather cannot be perfectly predicted, do not rely on a single generation forecast—practically it is useful to assume multiple scenarios and check profitability.

For example, set a standard case, a conservative case, and a favorable case, and assess the impact on annual generation and revenue for each. The conservative case assumes lower-than-expected insolation years, larger temperature losses, and somewhat larger soiling and downtime. The favorable case assumes good insolation years and lower losses. This approach shows how much variability the business plan can tolerate.

When explaining internally, clarify not just the final generation numbers but also the underlying assumptions. Be ready to explain which region and period the insolation data are based on, which drawings or survey results the installation angles and orientations use, and how shading and losses are estimated. If the basis for the figures is vague, explaining changes after planning or deviations in actuals becomes difficult.

Different stakeholders emphasize different points—financial institutions, investors, management, and facility managers each have different concerns. For those focused on profitability, annual generation and financial impacts matter. Facility managers care about maintenance, downtime risk, shading, and soiling management. Designers and constructors focus on site conditions, layout constraints, wiring, racking, and ground conditions. When sharing simulation results, tailor explanations to the stakeholder’s decision perspective.

Assuming errors is not a pessimistic stance but a realistic approach to understanding risks and improving plan reliability. If the goal is only to make generation look high, discrepancies after operation will cause problems. On the other hand, organizing risks and producing reasonable forecasts helps stakeholders make comfortable decisions.

Generation forecast errors affect not only revenue but also equipment sizing, contracted power, self-consumption rates, need for storage, and maintenance planning. For self-consumption purposes, not only total annual generation but also whether demand and generation times align is important. Even if generation exceeds forecasts, unused excess reduces expected benefits. Conversely, a slight shortfall in generation might still be acceptable if the system reliably supplies during peak demand times.

Summary: Generation forecasts change greatly with input accuracy and site understanding

To reduce errors in solar power generation simulations, verify the validity of input assumptions carefully rather than relying solely on calculation results. The basic practical measures are: bring insolation and meteorological assumptions closer to the site; accurately reflect installation angle, orientation, and shading; set equipment performance and degradation realistically; reduce input discrepancies through on-site surveys and surrounding environment investigations; and revise simulations using operational data.

Simulations are not meant to predict the future perfectly. They exist to identify risks at the planning stage and provide decision-making material for profitability and equipment design. Therefore, rather than avoiding errors, decompose the causes of errors and clarify which aspects have been confirmed and where uncertainty remains.

Site understanding in particular strongly affects forecast accuracy. Insolation and meteorological data can be checked by desk work, but elevation differences, installable areas, surrounding structures, trees, shading, and rooftop obstacles are easily overlooked if the site is not measured correctly. To improve simulation accuracy, do not only refine input values in the software but also enhance the quality of the on-site data that underlies those inputs.

To efficiently collect position checks, surveys, and records of the surrounding environment in the field, a system that can easily obtain high-precision position information is helpful. Using LRTK, a GNSS high-precision positioning device that can be attached to an iPhone, makes it easier to conduct project site surveys for solar equipment, confirm installation ranges, record surrounding obstacles, and perform post-construction verification on site. Bridging desk calculations and site data is essential to reduce simulation errors even slightly and make generation forecasts usable as practical decision-making material.