Recognizing Weather Data in P50/P90 Analysis

8 posts / 0 new
Last post
amartinez
Recognizing Weather Data in P50/P90 Analysis

Hello:

I am having issues with SAM recognizing weather files for P50/P90 analysis. I have 15 years of weather data for my particular project site in TMY format. (I have also converted it to SAM format using the macro functionality, but that did not seem to help.) I can run each of the files individually in a normal simulation run, but when I try to point to the file folder for P50/P90 analysis it does not recognize the files. SAM indicates that "it is not possible to empirically determine the P90 value with less than 10 years of weather data".

It gives me this message after it scans the folder and indicates the 15 files are present.

Any help would be appreciated.

Thanks,
Andrew

Paul Gilman

Dear Andrew,

Are you using the "_yyyy" convention for the file name (for example: weather_file_1990.csv, weather_file_1991.csv)?

You also might try running the weather file checker macro on each file to be sure there is not a formatting issue that is preventing SAM from recognizing the file as a weather file.

If you have a way to share the files with me, I would be happy to investigate further. You can either compress them into a .zip file and email them to me at sam dot support at nrel dot gov, or email me a link to the files on a file sharing service (Dropbox or similar).

Best regards,
Paul.

Paul Gilman

I just downloaded a set of weather files for "Denver CO" from solar prospector and used them for a P50/P90 analysis. Note that the Solar Prospector weather files are suitable for SAM's solar performance models, but will not work with the Wind model.

Here are the steps I used:


  1. Download the 12 files from Solar Prospector. I did it using SAM's download feature, but you could instead download them directly from the Solar Prospector website. You have to download each file individually -- a little tedious, but it just takes a couple of minutes. Note that the file names indicate the year using the "_yyyy" that SAM requires for P50/P90 simulations, so you do not need to change the file names.

  2. Put the files in a folder. To avoid potential problems, I put the files in their own folder:
    Screenshot of folder containing single year data files from Solar Prospector

  3. In SAM, click P50/P90, and select the folder above as the weather file folder.

  4. Click Run P50/P90 analysis.

    Screenshot of P50/P90 simulation results

Best regards,

Paul.

kalimazoop

Thanks for posting this Paul. I went through this process recently and discovered some problems with certain historical files downloaded from the Solar Prospector. I went to Solar Prospector directly instead of downloading them through SAM. I adjusted the data format to align with the SAM Expected format outlined in the "build a weather file" guidelines under help. The SAM "build a weather file" macro did not work on these files.

I discovered the following in the P50/P90 program function:
- the hourly timestamp hase to be 0-23. The files I got had 1-24
- It does no like leap years. It processes the data but does not display the result. The files with leap years get noted in the error log.
- Some Prospector files do not have RH.
- Some Prospector file headers do not line up with typical CSV weather file format that SAM likes. May require moving non-solar values into columns to line up with headers.
- Weather file checker will note a -RH value at a certain address that does not really exist.
- SAM requires to be restarted several times to remove old iterations of macro checks or analysis with faulty files.

It would be nice to have a P50/P90 results report generator.

kalimazoop

Paul, could you clarify how the historical files can be downloaded from the prospector through SAM?

kalimazoop

nevermind, I just noticed you can select a specific year from the drop down in the download section.

Paul Gilman

Thank you for this information.

We will investigate this further and make some changes to the weather file checker macro to better handle the special case of files from the Solar Prospector database for the Summer 2015 version of SAM. I will also see if our colleagues in the Solar Prospector project can resolve some of the issues with the files in that database.

Also, please keep in mind that the number of files available in the Solar Prospector database (12) is really not enough for a statistically meaningful P50/P90 analysis. The NSRDB locations with 30 or 45 years worth of data are better suited for this kind of analysis.

Best regards,
Paul.

Paul Gilman

Hello again,

Would you mind telling us what point on the map you are investigating? If you would prefer not to tell me on the forum, please email me the information at sam dot support at nrel dot gov.

Also, please try downloading the files from the Solar Prospector website, but using the "TM2" format instead of "CSV" and let me know if you see the same issues.

Best regards,
Paul.

Theme by Danetsoft and Danang Probo Sayekti inspired by Maksimer