2018.11.11 version VS 2020.2.29 version

Hi Paul,

I compared the default molten salt tower cases using the 2020 and 2018.11.11 versions. I used des_moines as the location.
The output is different, 2018 version gave the value that was around 5% higher. I wanted to investigate further, so I started with comparing the incident thermal power on the field, that was slightly different too. In fact, for some DNI values mainly at the end of day version 2018 had thermal power while 2020 version didn't. The  optical efficiency was also higher for 2018 version, and therefore all other parameters are higher after the solar field.

Is this something you are aware of and part of the modifications you made to the 2018.11.11 version?

Many thanks

Hi Mehdi,

You can use the release notes to see what changes we made to the models between versions: nrel.github.io/SAM/doc/releasenotes.html .

I am not aware of any changes to the Molten Salt Power Tower model that would explain a 5% difference in total annual output for the default system. The default system design is optimized for Daggett, California, so the difference could be caused by changing the weather file to Des Moines, Iowa.

Best regards,
Paul.

Hi Paul,

I ran the default cases using Daggett  data, the difference is lower, around 2%, but still exists. When I checked different parameters such as optical efficiency, input to the power block, and power block efficiency there are small differences. I checked the release note and couldn't find any item that can cause this. 

Many thanks,
Mehdi

Hi Mehdi,

The CSP team has identified the cause of this difference as a new cooling model implemented for the Rankine cycle in SAM 2020.2.29. Please see the revised release notes for a description:

nrel.github.io/SAM/doc/releasenotes.html#2020-2-29-mspt

Thank you for bringing this to our attention.

Best regards,
Paul.