How to calculate the annual forecast of solar power generation using 6 items

Table of Contents

• Basics to grasp before calculating an annual forecast

• Item 1: Determine the installed capacity

• Item 2: Establish region-specific reference generation

• Item 3: Reflect azimuth and tilt angle

• Item 4: Estimate the impact of shading and surrounding obstructions

• Item 5: Subtract system losses

• Item 6: Incorporate degradation and operating conditions

• Example calculation of an annual forecast using the six items

• Common pitfalls in annual forecasting

• How practitioners can proceed to improve forecast accuracy

• Summary

Basics to Keep in Mind Before Calculating Annual Forecasts

When you want to produce an annual forecast of solar power generation, the first thing to understand is that annual kWh is not determined by system capacity alone. Solar power systems are expressed in scales such as 10 kW or 20 kW, but those figures only indicate the magnitude of the system’s output. How much electricity a system will actually produce over the course of a year is determined by a combination of factors, including solar irradiance conditions, installation orientation, tilt angle, shading, losses, and degradation over time.

Many practitioners who search for "solar power generation calculation" are likely looking for the annual kWh they can use for decision-making — for pre-installation comparisons, rough estimates for internal presentations, checking the appropriateness of system size, and organizing expectations for self-consumption. A common mistake is to end the prediction by simply multiplying the system capacity by a rough factor. That can be useful for initial assessments, of course, but for figures used in practice it's worth clarifying one more layer of assumptions to improve both accuracy and explanatory power.

To make the annual forecast easier to understand, annual energy production is assembled using the following approach. First, determine the system capacity, then establish a baseline for how much can be generated per kW per year in that region. After that, apply corrections for orientation and tilt, corrections for shading, losses in equipment and wiring, and finally reflect degradation and operational conditions. In other words, an annual forecast is less a single large equation than a process of sequentially stacking six verification items.

The units to keep in mind here are kW and kWh. kW represents the instantaneous output capacity, while kWh is the amount of electrical energy generated over a given period. For example, if an ideal 5 kW system generates power for 1 hour, it produces 5 kWh; for 4 hours, 20 kWh. In annual forecasts, you consider the accumulation of those hours over the course of a year. That's why you need to look not only at the system's capacity but also at the conditions under which it will operate and the extent to which it will do so.

In this article, the annual forecast of solar power generation is organized into six items and the calculation workflow is explained in a form that operational staff can use directly. Rather than assuming complex calculation software from the outset, the explanation follows an order that is easy to work through on site, so start by grasping the framework for assembling the annual forecast.

Item 1 Determine the equipment capacity

The starting point for an annual generation forecast is the installed capacity. If this remains unclear, no matter how carefully you apply corrections, the annual generation forecast cannot be determined. Installed capacity is generally calculated from the number of solar panels and the output per panel, and can be expressed as installed capacity (kW) = number of panels × output per panel (kW).

For example, if you install 25 panels at 0.4 kW each, the installed capacity is 10 kW. If you install 30 panels at 0.42 kW each, it is 12.6 kW. In this way, the installed capacity may appear to be determined mechanically at first, but in practice the first discrepancy often arises here, because the theoretical maximum number of panels and the number that can actually be installed often do not match.

If roof-mounted, setbacks at the roof edge, rooftop equipment, inspection space, distribution across multiple roof planes, building constraints, and so on mean you cannot install as much as the apparent area suggests. If ground-mounted, you must look at usable site area, clearances, maintenance access routes, shading patterns, and so on. In other words, equipment capacity needs to be determined not simply as the "maximum that seems placeable" but as the "capacity that can actually be adopted given the site conditions."

Also, even for the same 10 kW, the annual generation changes depending on how it is distributed across surfaces. A mostly south-facing 10 kW system and a 10 kW system split between east and west have the same capacity but do not have the same predicted generation. Therefore, it is important not only to treat installed capacity as a single number, but also to be aware of the layout information — how much is mounted on each surface. Subsequent orientation and shading corrections only make sense with this information.

When a practitioner prepares an annual forecast, they should first confirm the installed capacity and document the assumptions under which that capacity was decided; doing so makes it much easier to revise the figures later. A 10% difference in installed capacity will, if other conditions are the same, result in roughly a 10% difference in annual power generation. That is why, as the first item in an annual forecast, it is important not to leave the installed capacity ambiguous.

Item 2 Establish baseline power generation amounts by region

Once the installed capacity has been determined, the next required figure is the regional reference generation. This is an indication of how much a 1 kW installation can generate in a year in that region, and it serves as the baseline number for annual generation forecasts. The approach is to first set Annual generation (kWh) = Installed capacity (kW) × Regional reference generation (kWh/kW·year), and then apply adjustments for individual conditions.

The reason for setting this reference generation value is that solar radiation conditions vary by region. Even with the same 10 kW system, the annual power generation differs between regions with good solar radiation and regions that are more susceptible to cloudy weather or snowfall. Nonetheless, applying the same annual coefficient uniformly nationwide will result in overestimation in some regions and underestimation in others. If annual forecasts are to be used in practice, it is necessary at minimum to establish reference values that incorporate regional differences.

In practice, it's common to consider an approximate range of 1,000–1,200 kWh per 1 kW per year, but where within that range you choose is important. If solar radiation is good and installation conditions are favorable, generation tends to be on the high side, while in regions with harsher weather it tends to be on the low side. The important point here is not to use the reference generation figure as a mere fixed number, but to use it as a starting point — "this is about typical for this region."

Also, you need to be clear for yourself about which assumptions are included in this reference generation amount. For example, whether it is a higher value based purely on solar irradiation conditions or a practical value that already includes a certain amount of typical losses will change how you apply subsequent loss adjustments. If you perform calculations without sorting this out, you may end up subtracting losses twice or, conversely, not subtracting them at all.

To make annual forecasts less prone to fluctuation, first establish a reference generation value appropriate to the region and clearly state what assumptions that figure is based on. By multiplying this reference generation by the installed capacity, you can get an initial outline of the site's annual generation potential. This is the second item.

Item 3 Reflecting Orientation and Installation Angle

The third item is orientation and tilt angle. In annual forecasts of solar power generation, ignoring this factor can lead to large discrepancies with reality even when the installed capacity is the same. This is because the solar panels receive different solar irradiation depending on which direction they face and the angle at which they are installed.

Theoretically, a combination of orientation and tilt that is more exposed to solar radiation is advantageous. However, in practice, roof pitch, site conditions, racking plans, and building constraints mean you rarely achieve those ideal conditions exactly. When installing on an existing roof, you often follow the roof’s orientation and angle, and even for ground-mounted systems the optimal solution can change depending on layout and land conditions. That is why, for annual forecasts, you need to consider not “what if under ideal conditions” but “what if using the orientation and angle that will actually be adopted.”

A practical approach for use in the field is to apply azimuth and tilt correction factors. For example, multiply the annual generation potential calculated from the regional reference generation by a coefficient corresponding to the orientation and tilt conditions. If conditions are close to ideal, use a value near 1.0; if slightly unfavorable, use a slightly lower value; and use an even lower value for more unfavorable conditions. The important point is not to treat all systems as a single group. If installation surfaces face multiple directions, it is better to assess each surface’s conditions and aggregate them, which will more closely reflect reality.

One thing to be careful about here is not to jump to the conclusion that "it has no value unless it is south-facing." In practice, even if installations are distributed east–west, if you can increase the total installed capacity the annual total can still be sufficient, and there are cases where the timing of generation better matches demand periods. Conversely, even if a site is south-facing, large shading, a small number of panels, or strict spacing constraints mean you cannot simply call it advantageous. For annual projections, it is important to consider the actual overall layout, not just the theory of orientation and angle.

Orientation and tilt angle are the individual factors most easily overlooked after system capacity. However, simply applying this correction brings annual forecasts much closer to reality than a simple calculation based only on system capacity. As the third item, this is a condition that should always be taken into account.

Item 4 Estimate the impact of shadows and surrounding obstacles

The fourth item is the impact of shadows and surrounding obstacles. In annual forecasts of solar power generation, shading is one of the factors that easily causes discrepancies with on-site results. Even if installed capacity, regional reference generation, orientation and tilt have been accounted for, it is not uncommon for actual performance to fall well below expectations because shading was not sufficiently considered.

Factors that cause shading include various items such as surrounding buildings, trees, fences, rooftop equipment, handrails, antennas, utility poles, and adjacent structures. Moreover, shadows are not fixed; they move with the seasons and the time of day. Something that is fine in summer can be affected in winter when the sun’s altitude drops and long shadows are cast. In real sites there is considerable variety: cases affected only in the morning, cases affected only in the afternoon, and cases where only some rows are repeatedly affected.

For annual forecasts, it becomes easier to organize if you fold this shading effect into a shading correction factor. If there is almost no shading, use a value close to 1.0; if there is a slight effect, use a value lower than that; if shading is clearly significant, apply an even more conservative value. What is important here as well is that you do not simply "reduce by a little" uniformly. You need to look at from which direction, at what times of day, and to what extent shading appears, and consider the degree of its impact.

Also, checking shadows has its limits when done only at a desk. Even if drawings or aerial photos look fine, it is common to find unexpected obstacles once you visit the site. This is especially true for ground-mounted systems, large sites, or sites with multiple buildings, where small positional shifts can affect the assessment of shading conditions. Even for rooftop installations, the effects of rooftop equipment and the vertical projections of surrounding buildings can be difficult to understand without being on site.

Shadows should be considered not as "present or absent" but as "how much they affect performance over a year." To reduce variability in annual solar power generation forecasts, you must not ignore shadows — even if it's a hassle, including a shading assessment at least once is indispensable. As the fourth item, applying shading correction based on on-site conditions greatly increases the realism of the forecast.

Item 5 Subtract system losses

The fifth item is system losses. In annual forecasts of solar power generation, whether you include this or not makes a clear difference between theoretical and practical values. Even after accounting for installed capacity, regional conditions, orientation, and shading, those figures are still close to the generation potential. In reality, because electricity is converted, routed through wiring and equipment, and affected by temperature increases and soiling, the actual energy produced is reduced accordingly.

Representative examples of system losses include losses in conversion equipment, wiring losses, reduced efficiency at high temperatures, variability between panels, soiling, and performance differences over time. Although each of these may seem small when viewed individually, they add up to a non-negligible difference over the course of a year. What makes it difficult for practitioners to explain forecasted values is presenting figures that have not sufficiently accounted for these losses.

For annual forecasts, the approach of applying a system loss correction factor is practical for field use. For example, multiply the theoretical energy generation that has been adjusted for orientation and shading by an overall system correction factor. Even under favorable conditions it will not reach 1.0, and under harsher conditions it will be lower. What is important here is to clarify what is included in that factor and to what extent. If shading has been corrected separately, there is no need to again assume a large shading loss within the system losses. Conversely, if orientation or tilt disadvantages have already been included on the baseline generation side, you must avoid double-counting those effects.

For example, if the system capacity is 10 kW, the regional standard generation is 1,100 kWh/kW·year, the azimuth correction is 0.95, and the shading correction is 0.97, the annual estimate up to this point is 10 × 1,100 × 0.95 × 0.97, which is about 10,136.5 kWh. Multiplying this by the system loss correction of 0.85 yields a practical forecast for operations of about 8,616 kWh. The figures differ considerably, but this latter value is a more usable forecast for field use.

Subtracting system losses is not about being conservative with the numbers. It is the process of converting theoretical values into values that can actually be explained. If the annual forecast will be used internally, for comparisons, or for investment decisions, this fifth item is indispensable.

Item 6 Incorporate degradation and operational conditions

The sixth item is degradation and operating conditions. When the term "annual forecast" is used, it often refers to the forecast for the first year after installation, but in practice there are many cases where figures are handled with an eye toward continued operation rather than just the first year. What you should consider at that time are the differences caused by equipment degradation and operating conditions.

Solar power systems do not maintain exactly the same performance over long periods. In general, performance gradually changes with aging, and power generation can vary due to the buildup of dirt and differences in operational conditions. Whether regular inspections and cleaning are carried out, whether shading changes occur, and whether the surrounding environment remains unchanged can also affect annual energy production over the long term. Differences that are small in a first-year forecast can become non-negligible when viewed over multiple years.

Therefore, if you want to handle annual forecasts more practically, it is easier to organize them by placing the degradation/operational correction coefficient at the end. For a rough estimate for the first year, treating it as close to 1.0 is acceptable, but if maintenance conditions or assumed operation are clearly severe, you may take a slightly more conservative view. Conversely, when scaling out to existing facilities with performance data, you can derive site-specific corrections from that actual performance.

The important point here is not to view degradation and operating conditions overly pessimistically from the outset. What is required at the annual forecasting stage is not to incorporate low-probability worst cases, but to form a reasonable expectation based on standard operating conditions. However, it is also dangerous to treat the high theoretical values of the first year as fixed. In reality, power generation fluctuates with operating conditions, so you should understand that forecasts have a certain range.

Degradation and operating conditions are often left until later compared with equipment capacity and regional conditions, but they are important as a final refinement to turn the annual forecast into an explainable figure. Including this as the sixth item makes the annual forecast figures more resilient over practical operational timeframes.

Example calculation of an annual forecast using 6 items

Here is a concrete example of an annual forecast using six items. Suppose that at a given site the system capacity is 12 kW, the regional reference generation is 1,100 kWh/kW·year, the azimuth angle correction is 0.96, the shading correction is 0.95, the system loss correction is 0.85, and the degradation/operational correction is 0.99.

In this case, the annual projected power generation is calculated as 12 × 1,100 × 0.96 × 0.95 × 0.85 × 0.99. First, the installed capacity and the regional reference generation give 13,200 kWh. Multiplying this by the azimuth correction 0.96 yields 12,672 kWh, and applying the shading correction 0.95 results in 12,038.4 kWh. Up to this point, this is a value close to the theoretical one reflecting the installation conditions. Multiplying by the system loss correction 0.85 gives 10,232.64 kWh, and finally multiplying by the degradation and operation correction 0.99 yields approximately 10,130 kWh.

The point of this calculation example is that it isn’t doing anything complicated all at once. Taken one by one, you’re simply multiplying the factors that should normally be checked on site—system capacity, local conditions, orientation, shading, losses, and operation—in sequence. Even so, compared with the simple calculation using only system capacity, which gives 13,200 kWh, the forecast used in practice drops to about 10,130 kWh. That gap is precisely why it matters to assemble the annual forecast carefully.

Also, with this method you can see which factors are affecting power generation. For example, you can explain how much output would increase if shading conditions are improved, how to account for an unfavorable orientation, and how predicted values change depending on what level of system losses you assume. In other words, the annual forecast is not just a numeric result but also a tool for visualizing opportunities to improve conditions and for identifying where risks lie.

If a practitioner uses this for internal explanations or comparative evaluations, breaking it down into six items like this is extremely effective. Rather than simply saying “it’s X kWh per year,” you can show which conditions were considered and how they produced that result, which increases its persuasiveness.

Mindsets that are prone to failure in annual forecasting

Failures in annual forecasts of solar power generation are more often caused by oversimplifying the approach than by not knowing the formula. The most common mistake is multiplying the system capacity by a relatively high annual coefficient and treating the result as a fixed value. This makes regional variations, orientation differences, shading, and losses largely invisible. While that number can be useful for an initial assessment, using it as-is for internal decisions or proposal figures often leads to revisions later.

Another common mistake is overestimating the installed capacity itself. If you base capacity on the apparent area and assume the theoretical maximum, then carry that straight into the annual forecast, you’ll already be biased high from the outset. This is especially true for rooftops or tight sites: setting capacity without accounting for required clearances and inspection/maintenance access will inflate all subsequent figures. If you want to improve the accuracy of annual forecasts, you need to be realistic from the capacity‑setting stage.

Also, underestimating the impact of shadows can cause failure. It's easy to decide that a little shading can be ignored, but if shadows occur at the same time every day, they can't be ignored over the course of a year. Moreover, shadows may appear only in winter, only in the morning, or only on certain rows, so you can't simply categorize conditions as "with shadow" or "without shadow".

Furthermore, it is risky to perform calculations while it remains unclear where loss adjustments were applied. If the regional baseline generation already includes some losses and you then apply an additional large system loss, you will underestimate. Conversely, if you use a regional coefficient close to the theoretical value and do not subtract any losses, you will overestimate. Understanding what assumptions each coefficient is based on is very important for annual forecasts.

In other words, what tends to cause failures in annual forecasts is being too hasty with the numbers. Speed is necessary in practice, but if you skip clarifying the assumptions for the sake of speed, you will end up losing time later. Simply going through the six items in order can significantly reduce such mistakes.

How Practitioners Can Improve Forecast Accuracy

If a practitioner wants to improve the annual forecast accuracy of solar power generation, a stepwise approach to increasing accuracy is more effective than diving into detailed analysis from the outset. First, grasp the outline of the annual generation potential using installed capacity and regional reference generation. Next, incorporate azimuth and tilt, check shading conditions, and add system losses and operational conditions to bring the estimate closer to practical values. Keeping this order makes it easier to see how much the numbers changed at each stage.

It is important to record not only the annual forecast itself but also the set of underlying assumptions. What is the installed capacity, what is the regional reference generation, how was the azimuth/tilt correction applied, has the shading correction been confirmed on site, and what do the losses include? If you document these, you won’t be confused when you review the numbers later. Conversely, if only the numbers remain, making revisions or providing explanations becomes very difficult.

Additionally, if possible, incorporating month-by-month trends and comparisons with actual results will further improve accuracy. Annual forecasts are convenient, but because there are seasonal differences, some things may not be visible from annual totals alone. Especially if you want to see overlaps with self-consumption and demand, you should not stop at annual values alone, but also take monthly and time-of-day perspectives. That said, it’s not necessary to require that level of detail from the outset; first, prioritize constructing the annual forecast using the six items.

Also, never forget that the accuracy of on-site conditions is directly linked to the accuracy of annual forecasts. If the layout, obstacle positions, site elevation differences, or relationships with surrounding buildings are unclear, estimates for shading and orientation corrections will fluctuate. No matter how much you refine your desktop calculations, if the input conditions are coarse the results will remain coarse. In practice, it is not uncommon for the way on-site conditions are gathered to affect prediction quality more than the calculation formulas.

The annual forecast for solar power generation may look complex, but it is actually just an ordered list of six items to check. That is why carefully reviewing each item is the most efficient way to improve accuracy.

Summary

To make an annual forecast of solar power generation a figure usable in practice, assembling it around six items is effective. The first is installed capacity, the second is reference generation by region, the third is orientation and tilt angle, the fourth is shading and surrounding obstructions, the fifth is system losses, and the sixth is degradation and operational conditions. By organizing these six items in sequence, you can arrive at an annual kWh estimate that is far closer to reality than a simple calculation based only on installed capacity.

In practice, you don’t need perfect numbers from the start. However, if you base an annual forecast solely on installed capacity, the figures tend to shift during subsequent detailed reviews. That’s why it’s important to have a framework of at least six items and be able to explain which assumptions produced which figures. The reliability of an annual forecast depends not on the complexity of the calculations but on the thoroughness with which the assumptions are organized.

Especially in situations where you want to increase the accuracy of orientation, shading, and siting conditions, how accurately you can determine the on-site positional relationships makes a big difference. If your understanding of obstacle locations on the site or of candidate installation positions is insufficient, both shading correction and placement assessments tend to become less reliable. Whether on a roof or at ground level, accurately capturing site conditions directly ties to improvements in the quality of annual forecasts.

If you want to obtain such on-site conditions in a more practical way, LRTK, an iPhone-mounted GNSS high-precision positioning device, is effective. Because it makes it easier to record potential equipment locations and obstacle positions with high accuracy, it makes it easier to connect to annual predictions that take shading and layout conditions into account. If you want annual solar power generation estimates to be truly usable figures, in practice it is a major advantage to prepare not only the calculation formulas but also a system that can accurately acquire on-site conditions.