8 Causes of Discrepancies in Solar Power Generation Simulations

Solar power generation simulations are indispensable for equipment planning, profitability assessments, design comparisons, and preparing explanatory materials for power plants. However, it is not uncommon that a system anticipated to produce sufficient energy in simulations yields less power in actual operation, or that the monthly trends do not match expectations. Solar power generation is determined by many overlapping factors such as solar irradiance, temperature, azimuth, tilt, shading, equipment characteristics, installation conditions, and operations and maintenance. Therefore, instead of narrowing down the cause of a discrepancy to a single factor, it is important to sequentially check differences between input conditions and site conditions, and between design conditions and operational conditions.

Table of Contents

• What to understand before addressing discrepancies in solar power generation simulations

• Cause 1: Solar irradiance data do not match actual weather conditions

• Cause 2: Input azimuth or tilt angle differs from measured values

• Cause 3: Overlooking shading from surrounding buildings or trees

• Cause 4: Underestimating generation loss due to temperature rise

• Cause 5: Assumptions about panel capacity or equipment specifications differ from actual equipment

• Cause 6: Loss settings for wiring, conversion, soiling, etc. are too optimistic

• Cause 7: Construction accuracy and site survey errors are not reflected

• Cause 8: Post-commissioning degradation and maintenance condition are not considered

• Verification procedure to reduce discrepancies

• Summary

What to understand before addressing discrepancies in solar power generation simulations

When considering causes of discrepancies in solar power generation simulations, the first point to keep in mind is that a simulation does not precisely predict future generation but estimates a reasonable generation amount based on the assumed input conditions. Because solar power generation is strongly affected by natural conditions, generation can vary for the same equipment between a year with many sunny days and a year with prolonged cloudiness. In other words, it is not abnormal for simulation results and actual performance not to match perfectly.

However, if the discrepancy is large or consistent in the same direction each month, the issue may lie not in weather variation but in input conditions, design conditions, or site conditions. For example, if the annual generation is roughly similar but actual winter output is low, shading at low solar elevations may be suspected. Conversely, if only summer generation is lacking, temperature-induced output reduction or poor ventilation conditions may be involved.

Practitioners should avoid treating simulation results as a single number. Breaking down annual generation, monthly generation, hourly generation trends, loss items, and generation per unit capacity makes it easier to identify the cause of discrepancies. Looking only at annual values may appear acceptable, but monthly breakdowns can reveal the effects of shading, snow, soiling, and temperature losses.

Discrepancies can stem not only from input errors on the simulation side but also from insufficient site measurements. If roof or ground slopes, heights of surrounding obstacles, panel mounting positions, cable routes, or inverter installation environments are not accurately known, even high-performance simulation tools will deviate from reality. Forecast accuracy depends as much on how accurately site conditions are captured as on the calculation method.

Therefore, in solar power generation simulations, it is more important to confirm the assumptions that support the numbers than to focus solely on the numbers themselves. Understanding why a given generation amount resulted, which losses were assumed and by how much, and how site shading and tilt were reflected allows calm isolation of causes when actual performance diverges.

Cause 1: Solar irradiance data do not match actual weather conditions

Solar irradiance is the most fundamental input for solar power generation simulations. Since photovoltaic systems convert energy received from the sun into electricity, discrepancies in irradiance assumptions lead directly to large differences in predicted generation. If calculations use historical average irradiance data and the operational year has fewer sunny days than average, actual generation will be lower than the simulation. Conversely, a sunnier-than-average year can produce higher actual generation.

In practice, note that there are different types of irradiance data, and results vary depending on which station, period, and statistical conditions the data are based on. When using data from a nearby observation station, the weather at the plant or residential installation may not exactly match the station’s conditions. Even within the same municipality, coastal, mountain, basin, urban, and agricultural-adjacent areas can have different cloud formation, fog occurrence, and snowfall tendencies.

Also, irradiance differs between horizontal-plane irradiance and irradiance incident on the panel surface. Since panels are usually installed at a fixed tilt, horizontal irradiance should be converted to panel-plane irradiance considering azimuth and tilt. If those conversion assumptions are inappropriate, discrepancies may occur only in specific seasons.

Irradiance mismatches are especially noticeable in short-term comparisons. For example, judging a simulation based on just a few months of operation may mistake an unusual weather period for design or equipment problems. When verifying generation, check monthly, seasonal, and annual trends separately to distinguish temporary weather-driven fluctuations from persistent design-related discrepancies.

Be mindful of the vintage of irradiance data used in simulations. Long-term averages are often used in project planning, but recent climatic trends and localized weather variations can influence actual results. Rather than emphasizing only short-term revenue forecasts, explain the variability in irradiance year to year so stakeholders understand the possible upside and downside in generation.

Cause 2: Input azimuth or tilt angle differs from measured values

In solar power generation simulations, the panels’ facing direction and tilt angle are critical. Even slight differences in azimuth or tilt affect not only annual generation but also morning/evening and seasonal generation patterns. Especially for roof-mounted systems, drawings may indicate a near-southern orientation, but the actual orientation can be slightly southeast or southwest. For ground-mounted systems, the grading of the site or as-built support structure can cause subtle differences between assumed and actual angles.

Azimuth errors tend to appear as biases in hourly generation. An east-leaning orientation increases morning generation, while a west-leaning orientation increases afternoon generation. If the simulation predicted a midday peak but actual data show a morning or afternoon bias, the input azimuth may not match reality. Residential roofs and complex facility roofs often have different orientations on each surface, so simplifying the entire system to a single azimuth can produce large discrepancies.

Tilt also affects generation. Lower tilt tends to favor summer performance while higher tilt may favor winter insolation, though optimal trends vary by region and orientation, so judging only by general angles is risky. Leaving tilt as an estimated input can result in monthly generation deviations from actual performance.

In practice, verify that design drawings and site-measured angles match. Even if roof pitch is shown on drawings, renovations, construction tolerances, or on-site detailing can change the actual condition. For ground-mounted systems, row-to-row tilt variations and uneven ground can impact generation. Rather than viewing only overall trends, check conditions by installation area, circuit, and roof surface to reduce discrepancies.

Azimuth and tilt also affect shading evaluation. The same obstacle casts different shadows depending on panel orientation and angle. When investigating discrepancies, validate azimuth and tilt in conjunction with surrounding obstacles, solar elevation, and seasonal changes.

Cause 3: Overlooking shading from surrounding buildings or trees

Overlooking shading is a very common reason why simulation results are higher than actual production. Many things can cast shadows on panels: nearby buildings, utility poles, trees, rooftop equipment, railings, chimneys, and adjacent panel rows. Shadows often occur only during certain hours or seasons, so a single site visit can miss them.

Winter shadows require particular attention. With lower solar elevation, buildings and trees that were not problematic in summer can cast long shadows and affect panels. Even if annual generation looks similar in a simulation, monthly checks may reveal low winter performance. In such cases, the issue may not be irradiance data but shading occurring at low solar elevations.

Shading effects are not simply proportional to the shaded area. Panels are electrically connected in strings, and shading some panels or cells can reduce the output of an entire circuit. Therefore, dismissing small shadows as insignificant can be dangerous. Thin shadows crossing panel rows at morning or evening, or persistent partial shading of panels, can reduce generation more than expected.

Tree shading also changes over time. Small trees at planning may grow and cast shadows years later; deciduous species change shadow density and extent between summer and winter. Nearby building construction or expansion can alter conditions from what was assumed during planning. When examining discrepancies, recheck current site surroundings as well as the conditions at planning time.

Accurately assessing shading requires measuring obstacle positions, heights, and distances on-site and confirming shadow movement by season and time of day. Low obstacles that are easily overlooked—such as rooftop equipment, fences, or parapets—can cast significant shadows at low solar elevations. If shading conditions are not sufficiently reflected in the simulation, actual generation shortfalls may persist.

Cause 4: Underestimating generation loss due to temperature rise

While panels generate more when irradiance increases, their output declines as module temperature rises. Underestimating temperature-related output reduction can cause summer generation to be lower than simulated. Many assume summer yields the highest generation, but rising ambient and module temperatures can prevent output from increasing proportionally with irradiance.

Temperature losses are not determined solely by ambient temperature. Module temperature depends on whether air can flow behind the panel, the gap between the module and roof, whether the mounting surface is metal or concrete, and whether the surrounding structure traps heat. Near–roof-integrated installations or poorly ventilated setups tend to yield higher module temperatures and lower generation under the same irradiance.

Ground-mounted and roof-mounted systems experience different thermal environments. Ground-mounted arrays generally have better ventilation, while rooftops or metal sheet roofs can reflect and retain heat. If a simulation uses standard temperature conditions without capturing site-specific ventilation or roof material impacts, summer generation can be overestimated.

To check temperature losses, examining not only monthly generation but also daily peak output patterns is useful. If a sunny day with ample irradiance shows flattened midday output or lower-than-expected peaks, temperature rise or equipment limits may be involved. If underperformance occurs mainly during high-temperature periods, review the temperature assumptions.

When simulating, reflect module temperature characteristics, mounting method, ventilation conditions, roof materials, and surrounding environment as realistically as possible. Temperature losses are less visible than shading or orientation but are a major factor causing discrepancies with actual generation.

Cause 5: Assumptions about panel capacity or equipment specifications differ from actual equipment

Another cause of simulation discrepancies is a mismatch between assumed equipment specifications and the equipment actually installed. Differences in nominal maximum output per panel, panel count, circuit configuration, inverter capacity, conversion efficiency, input range, and assumptions about oversizing (DC/AC ratio) will naturally shift generation forecasts. Specifications may change during procurement or construction, but simulation inputs are not always updated accordingly.

Panel capacity cannot be judged only by total capacity. If panels face multiple orientations, generation tendencies differ by orientation and tilt even with the same total capacity. East–west split roofs, multiple roof surfaces, or arrays with varying tilt mean that inputting capacity as a single aggregate condition can cause hourly and monthly generation to diverge from reality. In practice, organize capacity and conditions by installation surface and reflect them in the simulation.

Inverter specifications also affect generation. Even if the DC-side generation is sufficient, AC-side output capacity imposes limits, and output clipping can occur on sunny days. If this is not properly accounted for, simulations may show high generation while in practice peak clipping reduces actual output. Especially when designing with high capacity to increase annual generation, it is necessary to check the impact of output limits.

Equipment efficiency is not constant. Inverter conversion efficiency varies with load ratio, and behavior at low or high outputs may differ from simulation settings. If multiple circuits are connected under improper conditions or voltage ranges are unsuitable, expected performance may not be realized. When checking discrepancies, look beyond catalog representative values and evaluate expected efficiency under actual operating conditions.

If post-construction generation is lower than expected, reconcile as-built drawings, equipment specifications, circuit diagrams, site photos, and monitoring data to confirm that installed equipment matches simulation inputs. If design changes or substitute equipment were used, recalculate the impact of those changes on generation.

Cause 6: Loss settings for wiring, conversion, soiling, etc. are too optimistic

Irradiance incident on the panels does not translate directly into generation. Various losses occur: wiring loss, conversion loss, mismatch loss, soiling loss, reflection loss, standby loss, and equipment aging. Underestimating these losses makes simulated generation tend to exceed actual results.

Wiring losses depend on cable length, cross-sectional area, current, and connection methods. Even in small residential systems they may be negligible, but for large roofs or ground-mounted systems with long cable runs they are nontrivial. Design-phase assumptions of ideal cable routing can be invalidated by as-built detours that increase cable length; if such differences are not reflected in simulations, they contribute to discrepancies.

Soiling losses are also important. Panel surfaces accumulate dust, pollen, yellow sand, bird droppings, leaves, and exhaust-related grime. Some soiling is washed away by rain, but shallow-tilt panels or sites near trees, roads, factories, or agricultural fields retain dirt more. Soiling can be thinly spread across the surface or concentrated in spots; the latter can significantly reduce output.

Reflection losses and angle-of-incidence effects also cause monthly deviations. When sunlight strikes the panel at an angle—especially in the morning, evening, or winter—incident irradiance on the panel surface is reduced. Consider incidence angle effects in addition to simple irradiance.

Also pay attention to mismatch losses. Individual panels vary slightly in power characteristics. Temperature differences across an installation, soiling, shading, or manufacturing variance can cause output differences within the same circuit. If simulations only apply standard loss rates, actual site-specific losses may exceed those assumptions.

Loss settings should not be set pessimistically without basis, but setting them unrealistically low will make simulated generation look good while diverging from operational performance. For internal approvals and customer explanations, be prepared to justify which losses you assumed and by how much.

Cause 7: Construction accuracy and site survey errors are not reflected

Even with correct simulation design conditions, if the as-built construction does not match those conditions, generation will deviate. Small differences arising during construction—rack angles, spacing between panel rows, installation height, foundation positions, roof detailing, cable routing, and equipment placement—can influence generation. For ground-mounted systems, land elevation differences, grading precision, row shading, and drainage conditions also affect generation.

Site survey errors cannot be overlooked. If site contours, roof dimensions, or obstacle positions obtained before design are inaccurate, simulation inputs will be incorrect. Small differences in obstacle distance or height change shading timing. Narrower-than-drawn spacing between rows can lead to front-row shadows on rear rows in winter, reducing generation.

On roofs, survey accuracy is critical. If panel surface azimuth, slope, steps, ridges, valleys, or rooftop equipment positions are not accurately recorded, panel layouts used in simulation may differ from reality. Complex roof shapes or buildings with many existing rooftop items are especially prone to divergence if judged from drawings alone.

Discrepancies due to construction accuracy may not be obvious from post-construction photos or drawings. Confirm as-built angles and positions and reconcile them with the design model and simulation inputs. When generation is lower than expected, inspect whether the installed system matches the simulation assumptions rather than focusing only on equipment and weather.

Also, changes during construction may not have been reflected in simulations. Even if the panel count remains unchanged, changes in layout, rack height, panel offsets to avoid rooftop equipment, or altered cable routes require re-evaluation of generation. Simulations should not be a one-time exercise for conceptual design only; update assumptions at the stages of finalized construction conditions and post-completion verification to improve accuracy.

Cause 8: Post-commissioning degradation and maintenance condition are not considered

Solar power systems do not maintain identical performance after commissioning. Panel output gradually declines with long-term use, and generation varies with equipment condition, soiling, connection degradation, changes in the surrounding environment, and maintenance frequency. Comparing multi-year actual performance to a simulation that only considered first-year generation naturally shows discrepancies.

Degradation progresses slowly and may be hard to detect in the short term. However, correcting yearly generation for irradiance and comparing year-by-year can reveal a gradual decline. While degradation itself is unavoidable, faster-than-expected losses may indicate equipment defects, poor connections, accumulated soiling, or increased shading.

Maintenance condition is also important. Systems with inadequate inspections may have unnoticed soiling or damage, cable faults, junction box anomalies, or inverter downtime. Without detailed monitoring data, generation declines can persist unnoticed for long periods. Comparing simulation to actuals is not only to evaluate forecast accuracy but also to detect operational issues early.

Changes in the surrounding environment also cause post-commissioning discrepancies. Nearby construction, tree growth, changes in adjacent land use, or increased bird or leaf activity can affect generation. Locations that were fine during planning may have different shading or soiling conditions after several years.

When using simulations for long-term financial analysis, consider not only first-year generation but also degradation rates and maintenance practices. Regularly compare actual generation to simulation assumptions and current equipment condition to use the simulation as an operational management tool rather than solely a forecasting document.

Verification procedure to reduce discrepancies

To reduce discrepancies, first check each input condition one by one. Organize solar irradiance data, installation location, azimuth, tilt, panel capacity, equipment specifications, shading conditions, loss rates, and degradation assumptions, and clearly state the basis for each. Rather than judging performance solely by generation numbers, understand how much each assumption affects generation.

Next, reconcile assumptions with site conditions. Confirm roofs, site layout, obstacles, surrounding environment, and equipment placement on-site rather than relying only on drawings or desk-based information. Accurately capturing roof surface orientation and slope, ground elevation differences, panel-row spacing, and positions of nearby buildings and trees improves simulation input accuracy. Insufficient site verification before calculating generation makes it harder to trace causes when large discrepancies arise later.

Also, check not only annual generation but monthly and hourly results. Even if annual totals align, spring vs. autumn, summer vs. winter, and morning vs. afternoon discrepancies can differ. If only winter is low, consider shading or snow; if only summer is low, consider temperature losses; if mornings/evenings are low, consider azimuth or shading; if midday output flattens, consider output limits or temperature rise. The pattern of discrepancy helps infer causes.

After commissioning, continuously monitor actual results. Comparing generation to irradiance and seasonal conditions makes it easier to determine whether differences are due to weather or equipment. Sudden drops in generation may indicate equipment stoppage, soiling, or connection issues; gradual declines suggest degradation or maintenance problems.

In practice, do not treat simulation as a one-off deliverable. Update assumptions and, if necessary, recalculate at basic design, detailed design, during construction when conditions are finalized, and after commissioning. Especially if panel layout or equipment specs change during construction, re-evaluate the simulation under the revised conditions.

Summary

The causes of discrepancies between solar power generation simulations and actual results are not limited to a single factor. Solar irradiance data, azimuth, tilt, shading, temperature, equipment specifications, various losses, construction accuracy, and post-commissioning degradation and maintenance conditions all combine to determine generation. When actuals and forecasts do not match, check monthly, hourly, and equipment-area breakdowns rather than only annual totals.

Improving simulation accuracy requires not only refined calculation methods but also precise capture of site conditions. If roof or site geometry, obstacle positions, panel tilt, racking, and cabling are not properly reflected, even detailed calculations will deviate from reality. Reliable generation forecasts require both desk-based input work and accurate on-site verification.

In particular, discrepancies from shading, tilt, and installation position can be improved by accurate site measurements. In planning and verification of solar power systems, recording high-precision coordinates of installation locations, roof surfaces, terrain conditions, and surrounding obstacles makes it easier to confirm simulation assumptions and analyze causes after commissioning.

If you want to improve site survey accuracy, using LRTK—a GNSS high-precision positioning device that can be attached to an iPhone—allows easy recording of site location information and helps reconcile design and simulation conditions. To reduce simulation discrepancies, enhancing the accuracy of the site data that underpins calculations is as important as the generation calculations themselves.