BESS modeling -> time resolution / weather data

Hello everyone,I am currently working on my master’s thesis on second-life EV batteries as decentralized energy storage systems, focusing on self-consumption optimization, peak shaving, and grid interaction with PV. For this, I am using SAM to model different PV + battery configurations.My main question is about the time resolution of weather data:

• As far as I can see, SAM mainly provides hourly TMY files. From what I understand, these are typically used for annual energy balance studies.
• But if the goal is to analyze battery dynamics (C-rate limits, short-term charge/discharge, peak shaving, cycle counting), wouldn’t higher resolution (e.g. 15-min) be necessary?

So I would like to ask the community:

1. From your experience, are hourly TMY files generally sufficient for battery studies in SAM, or are there clear limitations?
2. If sub-hourly data is recommended:
   • What are the main hurdles in practice (file formats, matching load profiles, simulation run times)?
   • Are there any best practices or data sources you can recommend?
3. Additionally, if anyone has worked on similar use cases (PV + storage for self-consumption, peak shaving, or grid support), I would be very grateful for any examples, models, or lessons learned that could help me frame my thesis more effectively.

Thanks a lot in advance for your advice!Best regards,
Peter

Hi Peter,

SAM requires a weather file with one year of data, but that data can be hourly, 30-minute, 15-minute, 10-minute, 5-minute or 1-minute data. You can use data from a typical meteorological year (TMY) file that contains hourly data representing a long term historical period, or hourly or sub-hourly data for a file that contains data for a specific year.

You can download these different kinds of files (down to 5-minute resolution for some locations and years) from NREL's National Solar Radiation Database (NSRDB) either directly from the NSRDB Viewer website or by using the "advanced download" option on SAM's Location and Resource page.

You can find a list of other resources for weather data on the Weather Data page of the SAM website: sam.nrel.gov/weather-data.html .

Best regards,
Paul.

Hi Paul,

thank you for your quick response!

Do you know which temporal resolution I should use to properly simulate battery behavior? Would a 15-minute resolution be sufficient?

Best regards,
Peter

Hi Peter,

That depends on your analysis goals, how the battery will be used, and the availability of data. If the load is highly variable and you have access to 15-minute load data, and your analysis is focused on details of how the battery operates, then a 15-minute resolution would be appropriate. If your analysis is more focused on long term costs and cash flow analysis, an hourly resolution might be sufficient. If you are exploring how well a battery in a PV-battery system responds to rapid changes in the solar resource on windy, partly cloudy days, then a five- or even one-minute resolution might be best.

Best regards,
Paul.