You calculated your photovoltaic system’s production in PVGIS, but the study you received gives a different result? Or are you comparing two reports with the same number of panels and different annual kilowatt-hours? Before choosing the higher estimate, you need to check whether the calculations actually describe the same installation.
PVGIS is the European Commission’s Joint Research Centre tool for solar radiation data and production calculations. It is useful for planning, but the result depends on the data and settings used. A difference alone does not prove that the equipment is wrong or that one offer is better.
Keep the complete reports, not just the final figure
Ask for the results file and the input data used for each calculation. A screenshot showing only “annual production” does not let you investigate the cause of the difference.
Note which calculation relates to the final proposal and which was done only for an initial assessment. If one parameter is based on a site survey and another is a temporary assumption, ask for them to be clearly distinguished. This will help you discuss what needs to be confirmed before the study is finalized.
This process complements the guide to comparing offers, with a specific focus on documenting the production calculation.
Check the version, database, and panel model
The phrase “calculated with PVGIS” does not describe all the settings used. Ask for the version, solar radiation database, and panel technology selection to be shown. Do not compare an old saved report with a current calculation without recording these differences.
The JRC update notes describe, for example, the addition of the “crystalline silicon (2025)” option, with coefficients adjusted for newer panel technologies. This option can change the estimate compared with the original coefficients, even at the same nominal power.
This does not mean you should choose whichever model gives you more kWh. Ask the designer to explain which option they used and why it is suitable for the proposed equipment. Tool updates are a reason to keep calculation details, not a reason to treat every older study as useless.
Make sure the same roof is being described
Compare the map location, panel power, tilt, orientation, and mounting method. In the PVGIS user guide 5 these are distinct inputs. Do not assume they are the same just because the reports are for the same home.
In particular, ask whether the actual array tilt or the optimization option was used. A theoretically optimal scenario needs to be distinguished from a solution that can actually be installed on your roof.
If there are two different roof surfaces, ask for an explanation of how each was represented and how the total was calculated. For array layout choices, see the separate south-facing and east–west guide. The aim here is to identify what changed between the calculations.
Do not simply add up all the loss percentages
A report may include losses from different effects, along with the “system loss” parameter. The percentage entered for this parameter is not necessarily the final overall percentage shown in the results.
The official JRC FAQs explain that individual losses are calculated sequentially and multiplicatively. Therefore, they are not added together as independent deductions from the same initial amount of energy.
Ask for an explanation of what each percentage includes and what additional adjustments were made after the calculation. If someone reduces the result again “for the inverter” or adds a general allowance, ask which loss they believe has not already been taken into account. The answer needs to be supported by evidence, to avoid both omissions and double counting.
Distinguish the horizon from nearby shading
The fact that the map knows your location does not mean it has captured every obstruction on your rooftop. The JRC methods documentation clarifies that the built-in horizon model is based on terrain and does not represent shading from very nearby objects, such as houses and trees.
So ask for an explanation of where the chimney, parapet, or neighboring building was assessed. If the final study gives lower production because it includes documented local shading, the difference may have a specific cause. However, a vague reference to “conservative assumptions” without an explanation is not enough.
Read the average together with the variability
The grid-connected system results in PVGIS 5 show average annual production, monthly distribution, and year-to-year variability. The average is not a promise for every future year, and the monthly figure is not a weather forecast for the specific month ahead.
Ask for the variability indicator to be explained in plain language. Do not turn it into a guaranteed minimum yourself. If you are comparing it with actual measurements from a system, keep a record of the measurement period, data completeness, and any periods when the system was not operating.
Also, keep production from the savings. For an economic assessment, it is necessary to examine how you use energy and which financial assumptions apply in your case.
Ask for a comparison that explains the difference
We recommend rerunning the calculations using common inputs and changing only one option at a time. Ask for a short note documenting the change and its impact, along with anything that remains uncertain. This gives you an explanation, rather than two numbers competing with each other.
At Solar Home, we combine an assessment of your site with your consumption needs. You can request a study and quote, so we can discuss the estimated production using clear assumptions.
