All original content is created in Ukrainian. Not all content has been translated yet. Some posts may only be available in Ukrainian.Learn more

How to calculate the output of solar panels using the Global Solar Atlas

Post cover: How to calculate the output of solar panels using the Global Solar Atlas
This content has been automatically translated from Ukrainian.
If you are planning to install solar panels, one of the first questions is how much electricity will they actually produce over the year?
Honestly, I installed my panels without such research. I just directed them south and installed them vertically - at an angle of 90°. One of the reasons for this decision was to reduce snow accumulation in winter so they could work better at low sun positions.
After a few months of operation, I became curious: how much will generation drop in winter?
I have 3.84 kW of solar panels installed, and even a small generation in winter would be useful. For example, to partially power the boiler, pumps, and a couple of light bulbs.
Of course, each installation is unique. Generation is influenced by orientation, tilt angle, installation location, weather, and many other factors.
That’s why I wanted to not just guess, but try to estimate this more technically.
For this, I used Global Solar Atlas - a service that allows you to model the expected generation of a solar power plant considering its capacity, orientation, and tilt angle.
You can evaluate this using the free service Global Solar Atlas. This is not an advertisement. This service is quickly found on Google, seems to have no ads, and is supported by several companies.
It allows you to specify the installation location, capacity of the solar station, orientation, and tilt angle of the panels and get an approximate generation forecast.
Screenshot 2026-08-19 at 16.35.55.png
In this article, I will show this with my own example.

Opening Global Solar Atlas

First, we open Global Solar Atlas and find on the map the location where the solar panels are installed.
In my case, it is the Ivano-Frankivsk region.
Screenshot 2026-08-19 at 16.36.09.png
The service immediately shows basic information about the solar potential of this area.
For example, for my point, Global Solar Atlas shows:
  • GHI - 1154.5 kWh/m² per year - global horizontal solar radiation;
  • DNI - 1054.7 kWh/m² per year - direct normal radiation;
  • DIF - 579.5 kWh/m² per year - diffuse radiation;
  • GTI opta - 1366.2 kWh/m² per year - radiation on a surface with an optimal angle;
  • OPTA - 38° - optimal tilt angle of the panels.
The last two indicators are particularly interesting for comparison.
The service estimates that for this area, the optimal tilt of the panels is approximately 38°, while my panels are installed vertically - at 90°.
This means my installation is consciously different from the optimal one.
And that’s why simply taking the number from the map is not enough. It is necessary to set the real parameters of your own solar station.

Setting Up Your Own System

In the Choose PV system to calculate energy yield block, you can change the configuration of the PV system. First, we will select Small residential.
Screenshot 2026-08-19 at 16.36.22.png
The standard solar station will be displayed:
PV system configuration
Pv system: Small residential
Azimuth of PV panels: Default (180º)
Tilt of PV panels: Default (38º)
Installed capacity: 1 kWp
Screenshot 2026-08-19 at 16.36.34.png
But we need to set our parameters. Click Change PV System.
Screenshot 2026-08-19 at 16.36.42.png
I set my parameters:
Параметр Значення
Тип системи Small residential
System size / Потужність сонячного поля 3,84 kWp
Azimuth of PV panels / Азимут 180°
Tilt of PV panels / Кут нахилу 90°
What is kWp? kWp (kilowatt-peak) is the nominal or peak power of the solar power station.
Screenshot 2026-08-19 at 16.36.51.png

Why 180°?

An azimuth of 180° means that the panels are directed south.
For my installation, this corresponds to the actual positioning of the panels.

Why 90°?

90° is a vertical installation.
This means the panels are actually installed on a vertical wall of the house, not on a sloped roof.
This is important to specify in the calculator because the tilt angle directly affects the amount of solar energy the panel surface receives throughout the year.
Click on Open detail:

Screenshot 2026-08-19 at 16.54.59.png

What is the result
Screenshot 2026-08-19 at 16.37.07.png

After entering these parameters, Global Solar Atlas calculates the expected generation.
For my system, it shows:
2.971 MWh per year
or approximately:
2971 kWh per year.
My station has a capacity of 3.84 kWp, so if we convert the result to specific output:
2971 ÷ 3.84 ≈ 774 kWh/kWp per year.
This is already a quite convenient number for comparing different solar installations.
For example, one could say that in my conditions, each installed kilowatt of solar panels, according to Global Solar Atlas, should produce approximately 774 kWh per year.

What do 2.971 MWh mean?

This does not mean that my station will produce the same amount of electricity every day.
Screenshot 2026-08-19 at 16.37.10.png
On the contrary - generation throughout the year is very uneven.
And this is where it becomes interesting to look not only at the annual result but also at the monthly.
кут 90 градусів
кут 90 градусів
Here it is worth noting one point: the figures presented in the Average hourly profiles interface are not a ready table of monthly generation. This is an hourly profile that shows at what hours of the day and in which months generation is expected.
That is, Global Solar Atlas allows you to look at the solar station not only from the perspective of "how many kilowatt-hours will there be in a year," but also to understand when exactly this energy appears.

Why I am interested in winter

My main reason for making this calculation is not annual generation.
In summer, 3.84 kW of panels already produce enough energy.
I am more interested in winter and possible blackouts.
In the house, there is a boiler, pumps, and basic electricity consumption. Therefore, even relatively small generation on a sunny winter day can be useful.
And here, vertical installation of the panels becomes an interesting compromise.
I consciously opted for vertical installation 90° over the optimal angle 38°.
кут 90 градусів
кут 90 градусів
The reasons are practical:
  • the panels are less covered with snow;
  • snow slides off a vertical surface more easily;
  • in winter, the sun is low on the horizon;
  • the panels take up less space;
  • there is no need to make a separate sloped structure on the wall.
But for this, one has to pay with lower annual generation.
Here is, for example, the Average hourly profiles graph for 38-degree installation:
кут 38 градусів
кут 38 градусів

Optimal angle vs. real

For my area, the service shows an optimal angle of about 38°.
This means that if I were designing the system solely for maximizing solar generation, I would aim for approximately this angle.
But in real life, maximum annual generation is not always the only goal.
It is important for me to receive energy exactly when it is needed, as well as to have the simplest possible structure.
Therefore, 90° is not a "installation mistake," but a conscious compromise.
And now it can at least be approximately evaluated with numbers. Visually, one can see that the graph with 90 degrees in the cold months is more "active." But in terms of values, not significantly more. So I took the values from both tables and fed them into AI for comparison. And the result is:
Місяць 38° 90° 90° від 38°
Листопад ≈150 кВт·год ≈123 кВт·год 82%
Грудень ≈161 кВт·год ≈177 кВт·год 110%
Січень ≈208 кВт·год ≈227 кВт·год 109%
Лютий ≈268 кВт·год ≈258 кВт·год 96%
Листопад–лютий ≈787 кВт·год ≈785 кВт·год ≈100%
Over the year, vertical installation loses about 30% of generation compared to the optimal angle of 38°. But if we look at the cold months - November, December, January, and February - the difference practically disappears (in these months). And in December and January, vertical panels even provide about 10% more generation.
The ratio is percentage-based. But I also had AI calculate a 10 kW station. For some reason, November was calculated significantly better for the 10 kW station. Of course, AI might have miscalculated something. So please - verify all information yourself.
Місяць 38° 90° 90° від 38°
Листопад ≈480 кВт·год ≈480 кВт·год 100%
Грудень ≈434 кВт·год ≈465 кВт·год 107%
Січень ≈527 кВт·год ≈589 кВт·год 112%
Лютий ≈644 кВт·год ≈644 кВт·год 100%
Листопад–лютий ≈2085 кВт·год ≈2178 кВт·год 104%
And the data is even a bit more interesting. Not exactly fire, of course. But the minimal need for snow cleaning is a big plus.
Of course, if selling electricity or actively using it - the angle of 38 degrees in my case is +30% annual yield.

An important nuance regarding the forecast

Global Solar Atlas explicitly warns that such estimates are primarily intended for preliminary analysis.
The service uses standard assumptions regarding many parameters of the PV system. For detailed design, tools are needed that allow for more parameters of the power station itself and use more detailed meteorological data.
Therefore, the figure 2971 kWh per year should not be taken as a guarantee.
This is a forecast.
And the most interesting thing for me is to compare this forecast with the actual generation of my 3.84 kW panels in a year.
Then it will be possible to check how close Global Solar Atlas turned out to be to reality specifically for vertical panels installed on a south-facing wall.
And I will draw conclusions regarding/increasing the home solar power system.
Like it?React
🧵

This post doesn't have any additions from the author yet.

Results of the SES (3.84 kW) work for July 2026: less generation, but more benefit
Aug 6, '26 17:15

Results of the SES (3.84 kW) work for July 2026: less generation, but more benefit

How much does a home solar power station (4kW) actually cost to build yourself: a detailed estimate
Jul 27, '26 20:22

How much does a home solar power station (4kW) actually cost to build yourself: a detailed estimate

Підсумки роботи СЕС (3.84кВт) за Червень 2026: обігрівач у підвалі, парадокси моніторингу та 35% автономності
Jul 3, '26 16:19

Підсумки роботи СЕС (3.84кВт) за Червень 2026: обігрівач у підвалі, парадокси моніторингу та 35% автономності

I bought a solcloudy logger for my KS5200PS.
May 31, '26 18:38

I bought a solcloudy logger for my KS5200PS.

How to securely attach solar panels to hollow brick: real experience
Jul 28, '26 16:52

How to securely attach solar panels to hollow brick: real experience

Error 08 on the Must inverter: how I caught this error in SBU mode
Aug 14, '26 16:04

Error 08 on the Must inverter: how I caught this error in SBU mode

I took an additional Must HBP19-5548 Vpm for my KS5200PS. An extra 5kW for storage. But also additional tasks that need to be solved. I'll answer the question right away - why not an inverter + battery? Two "all-in-one" stations cannot be connected in a system and it's difficult.
Aug 11, '26 11:05

I took an additional Must HBP19-5548 Vpm for my KS5200PS. An extra 5kW for storage. But also additional tasks that need to be solved. I'll answer the question right away - why not an inverter + battery? Two "all-in-one" stations cannot be connected in a system and it's difficult.

Aug 9, '26 11:00

8 серпня 2026 Сонячний звіт

Aug 8, '26 11:00

7 серпня 2026 Сонячний звіт

Aug 7, '26 11:00

6 серпня 2026 Сонячний звіт