Getting dates from simulation outputs

In using the SDK, I've noticed that pvwattsv5 doesn't output any dates to correspond with its time series values. The dates would presumably come from the weather data, which can be configured and input in many different ways. Is there some part of the ssc_data that will give me access to the dates from the weather data that was used (I'm aware that the TMY data has no specific year, etc.) to run the module?

Is there another module that can parse the weather and return dates as a part of its outputs?

Typically, I am running with an 8760 hr annual weather file that starts at Jan 1, but I don't want to assume that to be the case as the time step, start time, and end time could all be different for weather data from different sources.

Thanks for advice/pointers.

Sam

Dear Sam,

In general, SAM assumes that the first simulation time step is the first hour of Monday, January 1. (For subhourly data, the time step would be the first 15 min, 10 min, etc. of Monday January 1.) That means that SAM requires a weather file that follows the same convention.

The and modules available via the SAM API run a simulation for a single time step, so you could develop your own model that uses a different convention.

The SAM performance models use the row position of the record in the weather file to determine the simulation time step rather than the data in the time step fields. The exception is that the solar performance models use the time stamp data for sun angle calculations, because those calculations require a value for the Julian day. For example, for a weather file with typical year data, the sun angles at 10 am on January 10 for a file that uses 1988 data for January would use the Julian day number for January 10 1988.

As you correctly note, the module does not include the time stamp data as output. However, the module does, along with other useful information about the weather file.

Best regards,
Paul.

The wfreader compute module

This is interesting. If I have weather data that starts in June, for example, are you saying that feeding it into SAM as-is would cause subtle or not-so-subtle problems? Can you give me a sense of what those problems might be? I imagine specifically that I would need to provide all other hourly data (like customer loads) explicitly, but would it then be possible to do a run starting on an arbitrary day of the year?

Hi Sam,

We finally had time to look at this a little more carefully, and it looks like moving the June data in the weather file to beginning of the file should work for what you are trying to do, assuming that the data in the Month, Day, Hour, and optional Minute columns are consistent with the irradiance and meteorological data in each row.

That approach will not work if you use any inputs that are specified as month-by-hour matrices, such as shading losses, or the electricity rate schedules in the financial models. As you suggest, any other hourly data, including shading losses, load data, etc. would need to be consistent with the weather data.

Best regards,
Paul.