PV simulation with 1-min data

Hello,

I'm attempting to run an ILR sensitivity analysis with 1-min weather data from ORNL (MIDC) in SAM 2015.06.30. The model runs with my input file in SAM CSV format (attached) The average GHI, LCOE and other figures seem to be off by a factor of about three in comparison to the Knoxville (ID 723260) TMY3 file available directly in SAM. (I realize that there won't be complete agreement between a single year of data and a TMY file, though I was expecting them to be closer.)

Any ideas on where the issue lies, or what to try next? I've attached the report PDFs for 1-min and TMY runs.

Thanks,
Jeremy

Hi Jeremy,

I ran SAM's Solar Resource File Checker macro on the one-minute file, and it reports problems with the solar resource data. For example, solar resource data in a given time step might be inconsistent with the sun angles in that time step. If you look at the first few rows of data, the global horizontal irradiance (GHI) and diffuse horizontal irradiance (DHI) values are small but greater than zero for night-time hours. There are also hours in the summer when the irradiance values in the weather file seem unrealistically high. These kinds of problems are causing SAM's incident irradiance calculations to be invalid -- in the time series results, you can see many hours when the calculated plane-of-array (POA) irradiance is negative.

Best regards,
Paul.

Paul,

Thanks for the response. Is there a way to run the file checker macro with no error limit?

Regards,
Jeremy

Hi Jeremy,

It is not possible to change the limit on the number of errors the macro reports. I'll see if that is something we can change for a future version.

Best regards,
Paul.

Hi Paul,

Thanks.

I ended up solving my original problem by correcting the longitude (needed a negative sign) and time zone (had been EST, changed to -5). I had not looked at the example in the help file carefully enough. Another suggestion would be to specify the format for the header fields (Lat, Lon, TZ, elev) explicitly in the help file (weather_format.htm) since SAM doesn't throw an error for an unexpected value.

Also, it appears that SAM puts the location data for the selected TMY file into the PDF report when specifying a weather file on disk. The attached PDFs show Ann Arbor, for example, when the weather file was from ORNL.

Regards,
Jeremy

Hi Jeremy,

Thanks for the feedback.

I will try to make the weather file format description for SAM CSV a little clearer in Help. The fact that you were getting simulation errors for every time step was a clue that the lat/lon/tz data was wrong that I missed -- sorry about that.

As for the location data in the report, the report generator reads the data from the SAM input page rather than from the weather file itself. When you use the "Use a specific weather file on disk" option, SAM does not load that header data from the file onto the input page, so the report generator uses the wrong data. You can work around that issue by using the Folder Settings options to add the folder that contains your weather file to SAM's internal list of weather file folders. That will allow you to choose your weather file from the library list instead of using the "Use a specific weather file" option so that the correct location data appears on the PDF report.

I will bring both of these issues to the attention of the development team so we can improve that behavior in a future version of SAM.

Best regards,
Paul.