Interpolation of climate data

[Ricardo]

Dear PV SOL users & team,

I have been using PV SOL Premium 2018 for some months now and a question arose on climate data sensitivity. I am designing a project located in north-east Spain, more specifically in the area marked in yellow in the screenshot below (in a town called Santa Eugenia de Berga, Spanish ZIP code 08507), but the closest climate data as is would be from either Barcelona or Gerona, marked in red below:

 

I understand PV SOL is able to interpolate data via the “Create climate data for new location” in this screen, but could you please clarify which data points are used for that calculation? When I create the new location “Santa Eugenia de Berga” I notice that irradiation is higher than that of either Barcelona or Gerona, which leads me to think there are more points involved in this calculation. Could you please let me know which points would be taken into account for this simulation?

 

And la st but not least, I understand PV SOL interpolates data but doesn’t take into account local conditions, right? In this specific area I’m looking at there is a lot of fog, but I imagine this is not considered. Could you please confirm this as well?

 

Thank you very much for your support.

[developer_mh]

Dear Ricardo,

thanks for your question. In PV*SOL we use the climate data provider MeteoNorm that also takes care of the climate data interpolation. For some countries, we additionally provide extra climate files. This is the case in Spain as well, where the climate files that you can select correspond to the UNE EN 94003:2006 "Datos climáticos para el dimensionado de instalaciones solares térmicas."

These climate data files might differ from what you get out of MeteoNorm. In your investigation, the locations of Barcelona and Girona show climate data from the UNE standard, while the interpolated location of Santa Eugenia de Berga shows MeteoNorm climate data. You will also see that when you create your own climate data location near Girona or Barcelona, you will have values of around1610 kWh/m² as well.

When I use MeteoNorm directly, i.e. the stand-alone version, and create the location of Santa Eugenia de Berga, I get the information that it is 100% satellite data that is used for the interpolation.

Micro-climate features like fog are not considered, no. But there are some interesting specialities of the MeteoNorm interpolation algorithm that is worth a read:

http://www.meteonorm.com/images/uploads/downloads/mn72_theory7.2.pdf

Hope that helps, kind regards,

Martin

[Ricardo]

Dear Martin,

Thanks for your very quick and very helpful answer, as well as for the info on the MeteoNorm interpolation algorithm, which is indeed interesting.

However, I am still confused that such high differences exist between MeteoNorm and the UNE standard. I tried creating a location in Barcelona and it shows 1583 kWh/m², which is almost 4% higher than the 1526 kWh/m² shown as the standard in Barcelona. I did the same thing for Gerona and the values are 1610 and 1416 kWh/m² for my new point and the standard, respectively (both are very close, right in the city center); in this case the difference is almost 14%.

As you know, this has a direct impact on the production estimates for a solar plant and therefore on its profitability/feasibility. Do you have any suggestions on which of the two values is more trustworthy? The degree of uncertainty is very high for us...

Thanks in advance, kind regards,

Ricardo

[developer_mh]

Hi Ricardo,

I understand the uncertainties that these difference may cause. The fact is that there is no single "truth" for climate data for a given location, but different use cases and data sources. MeteoNorm tries to compile long term typical meteorological years for any location around the world, usually comprising 20 to 30 years of ground based or satellite measurement. That means the climate data you get from MeteoNorm is very typical for a location when looking at the last 20 or 30 years. This kind of data is generally accepted as valid input data for photovoltaic system analysis.

In some countries however there are published standards with predefined climate dataset of varying provenience. Some projects or investors in these countries require PV simulations to use these data as input in order to introduce some kind of comparability between the simulations. This might be reasonable when there are a lot of climate data providers that offer climate data of very incosistent quality.

To be honest, I don't know how the authors of UNE EN 94003 2006 had their data from. Given that the process of standardisation usually takes several years, these data were likely recorded well before 2006, I guess around 1990. MeteoNorm uses data from 1991 - 2010, so perhaps this is the reason for the difference in the annual irradiation sum. But we would have to know the source of the 94003 2006 data to really understand the differences.

If you plan PV systems in Spain, and no one is asking you or your company to use the UNE data, I would personally go with the more up-to-date data of MeteoNorm.

Kind regards,

Martin

[Ricardo]

OK, thanks a lot!