Dear Vineet,
Thank you for attaching the file. In order to completely replicate your analysis, I also need a copy of your weather file. (Note that you can embed weather data in a SAM file by clicking "Copy to project" on the Location and Resource page.)
However, I think I can explain what you observed without your weather file. I ran a simulation in your file using the TMY2 file for Phoenix, Arizona to generate the following scatter plot of hourly power output (kWh) vs direct normal irradiance (W/m2). As you can see, there are a few hours between about 110 W/m2 and 430 W/m2 where the system generates power, but most of the power is generated at higher DNI levels.
/sites/default/files/content/images/scatter-power-vs-dni.bmp
The following hourly graph shows once instance where the DNI (red) is 120 W/m2, and the power output (green) is about 11 MWh/h. You can see that it is an hour that follows several hours of sunshine, and that the field HTF temperature drop (blue) lags the drop in solar resource. In other words, the HTF temperature fluctuates at a slower rate than the fluctuation in DNI because the fluid stores heat. That could be one explanation for why there are some hours that the system generates electricity during periods of low DNI.
/sites/default/files/content/images/hourly-dni-temp-output.bmp
Best regards,
Paul.