Smaller time steps for linear Fresnel CSP

Dear Paul,

Currently I am working on yield analysis of CSP. I want to run the simulation for 1 mins interval for linear Fresnel technology. In support forum, in reply to the enquiry of one SAM you have mentioned that in current SAM version lower time step yield analysis for linear Fresnel technology is possible with an appropriate weather data. Could you please let me know how I should model the weather data file. I know that TMY3 or TMY 2 data file does not support sub-hourly data. Should I model my data file as SMW format.

With Best Regards

Shah

Dear Shah,

To run one of the CSP models using subhourly weather data, you should use the SMW file format. The format is described in SAM's help system:

Weather File Formats

We have included one sample .smw file in the collection of weather files that comes with SAM. You can find it in the default SAM weather file folder, which for the Windows version of SAM 2013.1.15 is C:\SAM\2013.1.15\exelib\climate_files. The file is NV El Dorado Hourly.smw. That file stores hourly data, but you can use it to see what the format looks like.

As I've mentioned in other posts, we have done some testing of the subhourly simulation capability of the physical trough model, and consider it to be an experimental capability. We have not tested the other CSP models with subhourly timesteps, so please be aware that they may not accurately model the system's performance at time steps shorter than one hour.

Best regards,
Paul.

Dear Paul,

Thank you for your reply. I had a several try for simulation with NV El Dorado.smw weather file. However, at each attempt SAM shows a simulation error with following massage:

Please check the input file. SVAL lookup error trnsys.user_trdsrc.

As I am new in SAM, it is almost impossible for me to figure out and solve the problem.

Could you please shade some light on this issue.

Thank you.

Best Regards

Shah

Dear Shah,

As I mentioned before, and as is mentioned in the Help system, The SMW file format is only fully supported by the physical trough model. It looks like the linear Fresnel model will not work with that format.

Best regards,
Paul.