Time step - Accuracy of the simulation

Dear Paul

I have a weather file with data every 10 minutes and I want to simulate a CSP Parabolic Trough Physical Model.

In the Simulator Options I have read the following warning referring to a 1 hour time step: “Note: Shorter timesteps will cause simulation times to increase and may not be appropriate for the accuracy of the models”.

After reading this warning, some doubts have arisen in me about the reliability of the simulation, so I do not know how to proceed:

1. Creating a SMW weather file with data every 10 minutes and run the simulation with a time step of 10 minutes.
or
2. Creating a TMY3 weather file with data every hour and run the simulation with a time step of 1 hour.

What is your advice? Which is the accuracy of any simulation different from the one with 1 hour time step?

Best regards

Igor

Dear Igor,

If you want to run the physical trough model with 10-minute time steps, you must use a weather file in the SMW format. The TMY3 format only supports 60-minute data.

Please note that the physical trough model is the only model in SAM that supports subhourly simulations. This capability is still in under development and testing, so please let me know if you have any issues.

Best regards,
Paul.

Hello Paul

Further to Igor's comments about sub-hourly time intervals and the requirement to use SMW rather than TMY3 input files.

I am modeling the CSP Linear Fresnel Novatec solar boiler and only require the electricity output without any financial calculations.

Has SAM been extended to incorporate sub-hourly intervals available for the Fresnel technology?

Currently, the SAM team is internalising the functions of the external TRNSYS module, which supports the CSP analysis.

Could you let me know what effect the internalisation of the TRNSYS will have on SAM ability to handle sub-hourly analysis?

Regards
Paul

Dear Paul,

In general, we are moving toward making all of SAM's performance models flexible enough to handle sub-hourly simulations, and our work to convert the CSP models from TRNSYS to SSC will help with that effort. Part of the effort involves making SAM capable of reading subhourly weather data, and the other part involves developing algorithms that account for transient effects that come into play at shorter time steps but can be ignored at the hourly time step.

As I mentioned in my earlier reply, we have only tested the physical trough model at time steps less than 60 minutes, and we still consider that capability to be in the testing phase. It is possible in the current version of SAM to run the linear Fresnel model at smaller time steps if you have the appropriate weather data.

Best regards,
Paul.

Dear Paul,

I made SMW file with 10 min time step weather data to design a model of physical parabolic trough, when I Choose weather data file in both versions of SAM 2013.1.15 and 2013.9.20 SAM reads only the Header i.e. it shows only the Location information but it does not show any Weather data information and when I click view hourly data it says "Could not understand weather file data type: only TMY2, TMY3, EPW formats are understood", I thought there is an error in my SMW file so I tried to run SAM with its default SMW file (NV El Dorado Hourly.smw), but SAM gives the same answer also for that file. Can you please tell me where is the mistake?

Best Regards
Vahan

Dear Vahan,

As you noted, SAM seems to not display the summary weather data for the .smw files on the Location and Resource page, and the weather data viewer does not recognize the .smw weather file type. Those are bugs that we need to fix.

However, SAM's physical trough model simulation engine does recognize the .smw weather file type, so you can run simulations, and view the weather data and results data in the hourly data viewer after you run simulations.

Best regards,
Paul.