PV simulation with time step < 1 hour

Start
Prev
1
Next
End

djacobowitz
Topic Author

22 Nov 2013 00:09 #1982 by djacobowitz

PV simulation with time step < 1 hour was created by djacobowitz

I have 1-minute weather data and would like to obtain 1-minute estimates of output from a PV installation.

Regardless of which PV mode I choose, the time step option in Simulator Options is grayed out; I can't change it.

I also do not know what format I should put the weather data in. There are other support articles that say that if you want to do sub-hourly timestep you need to use SMW format rather than TMY. I won't have trouble reformatting the weather data as long as I understand the format.

If it helps, I'm not too particular about the model itself at this time. Accuracy is less important than having realistic 1 minute output. I'll be using these results in a storage dispatch simulation. :-)

Regards,
Dave J

Please Log in or Create an account to join the conversation.

Paul Gilman

22 Nov 2013 16:30 #1983 by Paul Gilman

Replied by Paul Gilman on topic PV simulation with time step < 1 hour

Hi Dave,
For now, SAM's PV model only runs with hourly time steps. That's because we have not tested all of the algorithms in the model for smaller time steps. Some of the approximations (for example, cell temperature) are reasonable for hourly time steps, but may not be for smaller time steps.
You could work around this by building your own model in C++ or one of the other languages supported by the SAM Software Development Kit (SDK) . The pvwattsfunc module calculates a PV system's output for a single time step (using the PVWatts DC-to-AC derate factor to characterize the system's performance).
Adapting the PV model so it will run with smaller times steps is on our To Do list, but it's not a high priority at the moment. We'd like to hear from anyone else who would use this capability (and has the sub-hourly weather data that would be required to do so).
Best regards,

Paul.

Please Log in or Create an account to join the conversation.

djacobowitz
Topic Author

22 Nov 2013 18:52 #1984 by djacobowitz

Replied by djacobowitz on topic PV simulation with time step < 1 hour

Thanks for the info, Paul.

Right now the integration of high penetrations of PV into power systems is getting a lot of policy attention, and I expect that to increase over time -- as penetrations get higher. In order to understand the implications of intermittency (and the value of reducing it or managing it with curtailment or battery storage), people are going to need good sub-hourly models of PV output. So I think you will see demand for this capability increase. :-)

I'll look into the SDK and the pvwattsfunc module. I'm comfortable programming, but I was hoping to avoid anything hard-core!

Regards,
Dave Jacobowitz

Please Log in or Create an account to join the conversation.

djacobowitz
Topic Author

22 Nov 2013 19:03 #1985 by djacobowitz

Replied by djacobowitz on topic PV simulation with time step < 1 hour

By the way, Paul, do you use SWIG to make the Python bindings for the SDK? I'd love to see Perl bindings and you can use the same config files to set it up.

Please Log in or Create an account to join the conversation.

Paul Gilman

25 Nov 2013 22:03 #1986 by Paul Gilman

Replied by Paul Gilman on topic PV simulation with time step < 1 hour

Hi Dave,
We did use SWIG at first to generate the original Java wrapper with some customization. We wrote the other wrappers by hand.
We would like to add a Perl wrapper to the SDK, but don't have the resources to do so at the moment. We'll try to work it into a future workplan. Thanks for the suggestion!
If you'd like to develop your own tools for calling SSC from Perl, in addition to SWIG, there's PerlXS and Inline::C .
Thanks also for the feedback on how you would use a subhourly PV model.
Best regards,

Paul.

Please Log in or Create an account to join the conversation.

djacobowitz
Topic Author

26 Nov 2013 17:42 #1987 by djacobowitz

Replied by djacobowitz on topic PV simulation with time step < 1 hour

Thanks, Paul.

The C API is working fine for me right now. If I feel motivated, I may indeed write a wrapper module in Inline::C (or if I have to, raw XS). If I do, I'll send it to you for potential incorporation.

Regards,
Dave J

Please Log in or Create an account to join the conversation.

Start
Prev
1
Next
End

Moderators: Paul Gilman

NREL is a national laboratory of the U.S. Department of Energy, Office of Energy Efficiency and Renewable Energy, operated by the Alliance for Sustainable Energy, LLC.

NREL Home

NREL Research

Follow NREL

NREL is a national laboratory of the U.S. Department of Energy, Office of Energy Efficiency and Renewable Energy, operated by the Alliance for Sustainable Energy, LLC.