CSP Physical Trough model - Field thermal power incident

Hi,

I'm doing some parametric runs with the CSP physical trough model for a plant similar to Andasol 1. I've noticed that when changing the SCAs' reflectance value (all else being equal) some of the parametric runs returned slightly different values for the yearly sum of the field thermal power incident variable.

For example:



Looking to the hourly values I've traced back this difference to one / two hours in the year where although the DNI value is positive, the field thermal power incident value is 0, as well as other variables such as incidence angle or field collector DNI-cosine product, to name a few. For example, in the 0.935 case, Jan 25, 02:00pm has the error.



I tried to find  an explanation for this issue but I don't have one. Do you have any idea of what might be happening here and what can be done to solve the issue?

Thank you very much.

Best regards,
João Cardoso

Hi João,

Would you mind attaching a .sam file that demonstrates these results?

Best regards,
Paul.

Hi, Paul.
Here it goes. Thank you.
Best regards,
João

Hi João,

Thank you for sharing the file.

January 25th is a low resource day, and that hour coincides with the hour that the HTF temperature reaches the design loop inlet temperature of 293 degrees. For the 0.935 case, I changed the design loop inlet temperature input on the System Design page to 290 degrees, and that field incident power value for 2 pm January 25 was not zero. That suggests to me that the controller is setting the input power to zero under some conditions that are marginal for power generation.

I will see if I can get a better explanation from the CSP programmers.

Best regards,
Paul.

Hello, Paul.

Thank you for your reply and ensuing analysis.

I still have trouble understanding such behaviour since that variable should reflect the total irradiation impinging the solar field. Looking at line 3304 of the physical model code ( github.com/NREL/ssc/blob/develop/tcs/sam_mw_trough_type250.cpp if I'm not mistaken) we can see that  q_inc_sf_tot = Ap_tot*I_b/1.e6, thus the power plant controller should not consider this value to be zero. Moreover, other values are also set to zero which should not be such as the incident angle.

A better explanation from the CSP programmers would be great, thank you.

Best regards,
João Cardoso