Inconsistent results with Generic CSP Solar Field Optical Efficiency Table

Hello SAM Wizards,

I'm trying to use Generic CSP to model a collector that is most efficient when the sun zenith angle is not zero, e.g. 30 deg, using a Solar Field Optical Efficiency Table, and getting inconsistent results.

One of the first steps in the SAM hourly calculations is to obtain the Field Thermal Power Incident, which I expected would be equal to (solar field area) * (BNI from weather file), except possibly for low values of BNI during startup/shutdown or transient cloud. [This result seems to be true for the default Central Receiver full simulation model.] Indeed, results from the default Generic CSP setup are as expected, provided that the "Unsorted data table mode" is in use.

However, Thermal Power Incident values in the hourly Data Tables were high by a factor of 1.0596 in my simulations, which use my tab-delimited table. Results for the default generic tab-delimited table are much closer, but still differing by a factor of 1.0028 ("Unsorted data table mode" is unchecked). After much experimentation, I found that results would be exactly correct (for both setups) if the first 3 values in the column for azimuth=0 are identical and highest value in the table. This is surprising, given that the table efficiencies should only be used in the subsequent step in the hourly calculations (I presume).

I'm using 2020.11.29 64bit rev 2 on OSX 10.11.6. I'm new to SAM.

Please let me know If I've missed something in the setup -- the .sam file is attached in original form. The changes that I made in the columns for zero zenith angle were: default tab-delimited table, 3rd entry changed to 1. For my table, first 2 entries changed to 0.96.

Best regards
David

Hi David,

The discrepancy you found is caused by the difference between the solar field area estimated value you see on the Solar Field input page and the value SAM uses as the design value for internal calculations during the simulation.

The value on the input page is an estimate based on the "Total optical efficiency" as described in the page's Help topic: samrepo.nrelcloud.org/help/gss_solar_field.htm .

The value used to calculate the solar field area for the simulation, on the other hand, is the optical efficiency at noon on the summer solstice.

Best regards,
Paul.

Hi Paul,

Thanks for the clarification. I tried a different weather file, and the factor by which the thermal power incident appears higher is different, which makes sense. However, I can't find the actual collector area used in calculations reported anywhere in the results, though I can back it out of the relevant columns in the data tables. This value is very useful when testing variations in collector design. I had read that Help topic previously, but it doesn't say anything about the actual collector area or how it is determined.

Thanks and best regards,
David

Hi David,

Ideally, the solar field area estimate shown on the Solar Field page for the Generic CSP model would be the same as the design value used in the simulations. Unfortunately, for now, we do not have resources to work on the Generic CSP model. If we do get funding to work on the model, we will address this and some other issues.

I did revise the Help topic to explain the difference between the solar field area estimate in the user interface and the value used for the simulation. That revision will be in the next version of SAM (Fall 2022).

Best regards,
Paul.