Downloading and Importing from the new NSRDB Viewer

[CivilNJ]

I tried using the new NSRDB Data Viewer and located my site, went to Data Download (Point) and selected MTS1 and MTS2 tabs to retrieve TMY2 or TMY3 files. I receive a download but I receive format errors on import into PVsyst. I also noted that I do not see this option (from the Help File) - "Import Meteo Data"/”NREL National Solar Resource Database” . Was this removed in version 6.34? I am using US TMY2 and TMY3 and it throws errors.

Error:

Conversion error of string "Source" => LongInt and then it continues with additional ones for "State", etc.

I assume that something is incorrect in the header but I want to be sure that I am using the proper file and import option.

[kcradford]

Is NSRDB support going to be added now that Solar Prospector is no longer going to be supported, and is scheduled to be decomissioned?

 

"The Solar Prospector tool is no longer actively supported and is scheduled to be decommissioned on September 30, 2016. Please use the National Solar Radiation Database (NSRDB) to access NREL solar geospatial data going forward. If you have a need for the Solar Prospector tool to remain active, contact us at solar.tools@nrel.gov."

[PVFan]

I too am having this problem. Solar Prospector is no longer working and I am getting the same error messages when trying to import data from the NSRDB Viewer. I am using PVsyst 6.47.

[SolarJ]

Jean-Marie,

When downloading the data from the viewer, which data do we select on the PSM tab beyond the GHI described in the instructions? Particularly in the "Years" section. Do we select from PSM tab or MTS2? Thank you for the clarification.

[dtarin]

Jean-Marie,

When downloading the data from the viewer, which data do we select on the PSM tab beyond the GHI described in the instructions? Particularly in the "Years" section. Do we select from PSM tab or MTS2? Thank you for the clarification.

 

You will select TMY from the PSM tab. I select the items in the picture attached.

[SolarJ]

Solarguru,

Thank you for the response. Do you have to change the data field numbers in the *.MEF file? They are not lining up properly we looking at the fields in the file

[dtarin]

Here is my MEF file you can try.

NREL PSM 5.zip

[ghbanta]

I am trying to get data downloads from NSRDB to replace Solar Prospector. I now have found the settings for the NSRDB download wizard to get the correct data and have the "Steps to Import a File Downloaded from NSRDB into PVsyst"

I have also downloaded NREL_PSM_5.zip and placed the file in the meteo directory. I also see a Solar Prospector MEF in the same directory but I don't see the NREL_PSM as a choice for hourly data.

How do I '"install" the MEF file so it becomes an import choice?

Thanks,

Gary

[dtarin]

Inside the ZIP file is the MEF file. You will need to extract that into the meteo directory. You will need to have a site created first for your location. Then you will go to Import ASCII meteo file, under ASCII source, select your PSM weather file, change your names if necessary under "Hourly file to be created", and under conversion, select the MEF file (I have named it NREL PSM 5, but you can rename it to anything), click start conversion.

You will want to see 1990 grayed out in reference year. After you convert, view the meteo file and check the quality. You will most often need to implement a time shift. Re-run the conversion by adding the time shift by opening the conversion template and entering the shift under the date tab. Sorry if this was very brief or repetitive.

[SolarJ]

Thank you!

[headfast]

Dear all,

I am trying to obtain TMY data from the NSRDB to import it into PVsyst. I already got the TMY file and I am following the steps from this link to add it as my meteo data into PVsyst

https://nsrdb.nrel.gov/sites/default/files/NREL%2520NSRDB%2520into%2520PVsyst.pdf+&cd=1&hl=es-419&ct=clnk&gl=ve

But when I finish the 7th step I cannot get the Ok button to turn on and cannot proceed with the conversion, I followed the exact steps by using a TMY.csv on the ASCII source file option and then click "open" on the conversion section.

I get a red message that says "please define a significant comment".

I do not know why this is. Can you please help me out?

Thanks,

Charles

[cliebner]

Did you try changing the "Format description" to something more descriptive than the default text "New conversion format for meteo ASCII file"?

[gonecrawfishin]

Inside the ZIP file is the MEF file. You will need to extract that into the meteo directory. You will need to have a site created first for your location. Then you will go to Import ASCII meteo file, under ASCII source, select your PSM weather file, change your names if necessary under "Hourly file to be created", and under conversion, select the MEF file (I have named it NREL PSM 5, but you can rename it to anything), click start conversion.

You will want to see 1990 grayed out in reference year. After you convert, view the meteo file and check the quality. You will most often need to implement a time shift. Re-run the conversion by adding the time shift by opening the conversion template and entering the shift under the date tab. Sorry if this was very brief or repetitive.

 

It's not clear to me what time shift is appropriate. I downloaded NSRDB data and did not select "convert UTC to local time". I then used the .MEF file which was posted by solarguru on 10/17/16 which included a shift of 5 hours. My site is located in UTC-5. I'm getting errors that "GlobHor" data is undefined, and also that my time seems mismatched. Does anyone know what the most appropriate way is to treat NSRDB data, specifically PSM v3?

[dtarin]

It's not clear to me what time shift is appropriate. I downloaded NSRDB data and did not select "convert UTC to local time". I then used the .MEF file which was posted by solarguru on 10/17/16 which included a shift of 5 hours. My site is located in UTC-5. I'm getting errors that "GlobHor" data is undefined, and also that my time seems mismatched. Does anyone know what the most appropriate way is to treat NSRDB data, specifically PSM v3?

 

The result of the time shift is revealed after the conversion has happened. After complete, a box will come up asking for you to inspect the data, click yes, and select "check data quality" tab to see how far off you are, and if adjustment is needed.

If your glob Hor is undefined, it might mean that depending on the variables you imported from NREL, it may not be matching up with the MEF file I provided. You will need to update the MEF to point to the correct columns. When I download PSMv3, I only download the variables I need for PVsyst (GlobHor, DiffHor, Tamb, Wspd) + DNI. When I download a UTC-5 site, and I check "convert to local time" when download NREL PSM, a time shift of 5 hours generally works. If you did not check convert UTC to local time, you may need to do something different. Since I haven't done it with that unchecked, I cant provide any help there.

https://imgur.com/a/1mOXdl6

[gonecrawfishin]

I think I'm getting closer. I had been using the instructions of two different posters on this thread together.

Ok, so I checked "convert UTC to local time" when downloading a site in New York. In my MEF file, I have my time base is universal time and a +5 hour shift entered. I'm still getting an error of "The imported meteo data are shifted by -59 minutes with respect to the Clear Day model." Why might that be?

On the side, I tried having NSRDB leave the data in UTC then I changed the MEF file to have no time shift and I got "The hourly data are not defined. This meteo file is unuseable for simulations!". So, there I go moving backwards again.

Additionally, it was my understanding that the NSRDB PSM v3 data represents instantaneous values (or for some values, averages centered at the time label)? The PVsyst help menu seems to indicate that it is calculating solar position based on the half hour, so a value with a 10am time label is paired with a solar angle at 1030am? Thoughts on what would be most appropriate to account for that? If I brute force trial and error this until I get a green message from PVsyst, am I in the best place, or should I be adding in 30min offsets because I know better than PVsyst?

[dtarin]

If you are at -59 minutes, add 1 hour to the time shift in the MEF file and re-run. Are you downloading in half hour intervals on the NREL side? This should be unchecked. I just downloaded a NY site (albany) and with a timeshift of 5 hours, the result is 0. See images.

 

[dtarin]

MEF File attached

NREL PSMv3.zip

[gonecrawfishin]

Well that's just baffling, I used your MEFv3 file and got completely different results, see attached (left).

 

 

I am not downloading from NSRDB in 30 minute intervals & do have that unchecked. I'm questioning the difference in definitions of what, for example, 10am *means* for NSRDB vs PVsyst. But perhaps, as long as I can brute force PVsyst's time shift comparison to the clear sky model, then that's the best I can hope for. I added an hour and "the Meteo file seems to be OK". Unnerving, but ok.

I'm really appreciating the help so far.

[gonecrawfishin]

Holy moly, side note, when I compare the PSM v3 data to meteonorm or the Albany TMY3, it is HUGE, like, 1.5% more global horizontal.

I tried the PSMv2, which required a 30 min time shift to match the clear sky model. In the end, it actually required a 5hr + 30min time shift, which starts to make sense. But when I compare that one to Meteonorm or TMY3, PSMv2 drastically underestimates production, like 1.5% less than MN/TMY3, 3% less than PSMv3. I'm new to using the NSRDB directly, but boy is it all over the place! Something is wrong, and I don't believe I am using this data correctly yet.

[dtarin]

PSM can overestimate by 5% under clearsky conditions, and has a GHI MBE of +/-5%. See link: https://www.nrel.gov/docs/fy17osti/67722.pdf

[CStelzerKED]

The NSRDB TMY data at 33.26, -95.21 imports with PVsyst warnings that the "clearness index of the best clear days seems high with respect to the clear day model." I've checked another satellite grid nearby, and imported several different TMY, TGY years. They all say the same thing. Shows maybe 1.15 index at the highest day, and not crossing 1.0 until about 90 days. Also checked a flat dry spot in Arizona for comparison and didn't have that warning. For the Arizona data the max index was ~ 1.05, with a long stretch of days very close to 1.0. So something about the NSRDB data in NE Texas seems skewed. It appears to be overpredicting in the winter/spring/autumn, and almost spot on in the summer months.

Can anyone explain the story behind the data?

Best clear days Ktcs chart from meteo tables and graphs analysis V7.0.7

[Chusemi]

I'm not sure that it is a good idea to consolidate data from 10 data sources. If one data source pauses, all other data streams will be paused too.

I think you need one configuration, but write data to a new row in Excel for each data packet.