Battery charged by PV, option for DC-connected battery in 2015.06.30

I am having trouble getting the set-up right so that the battery in a PV system gets charged by the PV output rather than by the grid. Could this be related to the option which seems to have disappeared since the beta 2015.05, to select AC or DC connected battery system?
I've tried various set-ups but as far as I can tell the battery only gets charged by the grid, even if the PV power is much higher than the load. If I disable charging by the grid, the battery discharges from initial condition and stays discharged.

What I want to model is a PV-battery system with constant load and no grid. This option isn't available directly, but I think that it should still be easy to do, since the algorithm prioritises the PV and battery to supply the load and the PV output to charge the battery. Then any grid usage indicates that the PV and battery has insufficient capacity, and I can subtract the grid usage to see how much of the load can be supplied.

I am also getting an error:
"ssc variable does not exist: 'battery_bank_replacement'
Simulation cashloan failed."
which has appeared mysteriously. I don't know what I changed that caused it to appear and I can't work out what to change to get rid of it. Any ideas of where to look?

Good morning,

The battery will charge from PV only if PV production exceeds the electric load and the "Charge from PV" option is checked in the "Manual Storage Dispatch Controller" for the time period of interest. If "Charge from grid" is selected within the dispatch controller, the battery will charge from the grid if needed. Please note, if both charging from PV and charging from the grid are enabled, then the battery will by default charge from PV if available and only charge from the grid in the event that PV production is not sufficient to both meet the load and charge the battery.

If you've setup your dispatch controller and believe it is not working correctly, please send your .SAM project file to SAM.Support@nrel.gov and Nicholas.DiOrio@nrel.gov and we will investigate to determine if there is an error in the controller.

The option to select AC or DC connected batteries has been removed while issues regarding the DC connection are investigated.

It may be somewhat difficult to model a PV-battery system no grid using SAM, as the models will seek to satisfy the electric load with PV+Battery and then purchase the remaining power required from the grid. There are potentially several ways to de-emphasize these grid purchases, including through the electricity rate structure. Your approach sounds like a good way to remove the grid purchases.

Unfortunately there is a bug in the battery replacement module which will be addressed via a patch later in the week. To get rid of the "ssc variable does not exist: 'battery_bank_replacement' issue, please check "No replacements" in the "Battery Bank Replacement" section.

Thank you for the detailed comments, we will continue working to improve the battery model.

Hi,

Did anything further happen with regards to this question?

I've been trying to model the same thing, essentially off-grid with any grid power draw meaning an on-site installation shortage. I have the same issue with the battery discharging once in the beginning of the simulation but never recharging even when the PV power exceeds the load.

I haven't had the "ssc variable does not exist: 'battery_bank_replacement" issue though.

Thanks

Hello,
SAM's battery model is designed to model grid-connected photovoltaic systems that meet a residential or commercial building's electric load. It assumes that the battery is connected to the system on the AC side of the photovoltaic inverter.
Are you using the "Manual Dispatch Model?" If so, be sure that "Charge from PV" is checked for the time periods that the PV output exceeds the load. Also, you can specify whether to prioritize meeting the load or charging the battery from the PV array.
If you would like me to review your file, please email me a copy of your file.
Best regards,

Paul.

I actually restarted the program and remade the model. It works this time. I did notice something that may be worth noting: if the "Charge from grid" is checked and the % capacity is set to 0, this problem still happens (it was something I tried as a workaround earlier).

Thanks for your response.