Monthly P90 Values

[Shashank Sharma]

Hello PV,

I want to know that, can we find out P90 monthly values in our PV syst report. If no, there is any calculation or general formula for it.

[Linda Thoren]

Dear Shashank Sharma,

P50-P90 estimations rely on yearly data. Calculating P90 for monthly values is not advised. Yearly weather data has variation (around 3-4% RMS), while monthly data can vary significantly year-to-year, leading to unreliable probability profiles for each month. You can read more about the P50-P90 evaluations in the Help in the link below:
https://www.pvsyst.com/help/index.html?p50_p90evaluations.htm

Regards

[sagar b]

On 9/29/2023 at 11:42 PM, Linda Thoren said:

Dear Shashank Sharma,

P50-P90 estimations rely on yearly data. Calculating P90 for monthly values is not advised. Yearly weather data has variation (around 3-4% RMS), while monthly data can vary significantly year-to-year, leading to unreliable probability profiles for each month. You can read more about the P50-P90 evaluations in the Help in the link below:
https://www.pvsyst.com/help/index.html?p50_p90evaluations.htm

Regards

Hi Linda,

In my one project, the default variability value of weather data is 8.1%. I've used the Meteonorm data source.

Due to this, P(90) estimation is calculated lower than expected. 

For most of the projects, I've noticed that this default variability value lies between 3-5%. Do you have any suggestion on how to customize it?

[Linda Thoren]

Hello,

High annual variability indicates that weather conditions can differ significantly from year to year at this site. The purpose of running a P50-P90 analysis is to account for these uncertainties. Therefore, it is not recommended to customize the weather data variability, though you do have the option to input custom values if needed.

[sagar b]

1 hour ago, Linda Thoren said:

Hello,

High annual variability indicates that weather conditions can differ significantly from year to year at this site. The purpose of running a P50-P90 analysis is to account for these uncertainties. Therefore, it is not recommended to customize the weather data variability, though you do have the option to input custom values if needed.

Thanks Linda for you prompt response.