Large PV + battery, wrong calculated SOC and System power

Hello,
I'm modeling a large PV array of about 9GW with a large battery to simulate fully renewable power production of a geographical region.
The battery charges and discharges with a detailed "Custom time series", built according to region's power demand and PV array production (when PV production is higher than the demand, the excess power goes to charging the battery, and when PV production is lower than regions demand, the battery provides additional power to region's grid).
The battery is connected to hybrid inverter's DC side.
The simulation has multiple errors:
1. Battery SOC jumps above and below set limits of minimum 15% and maximum 95%
2. At night, PV production is 0 and battery SOC is 0%, but "System power generated" is 9GW.
3. Custom time series array is set to a negative value (for battery charging), but apparently the number is too high, as the results for same time slot are 0.
    If I try lower numbers (less than 1e+6), the system can charge the battery.

Hi Lev,

SAM is designed to model system sizes at the project level rather a regional level, and large numbers can cause problems with some internal calculations and the way results are displayed. Can you model a smaller system and scale up your results accordingly?

Best regards,
Paul.

Hi Paul,
Thank you for prompt reply.
I'll try to model smaller system and scale it up.

Hi Paul,
What are the battery limitations?
Charge/Discharge power, current, C-rate?
Any other limitations?
Thanks.

Hi Lev,

The battery model should work with values of those parameters that are realistic for a physical battery. SAM does not set limits to those input values. Because some of the calculations involve iterative algorithms, it is hard to predict when an unrealistic input value might cause problems with the calculation.

Also, SAM is written in C++, which has a maximum integer value of 2.147e+9 as described here: docs.microsoft.com/en-us/cpp/c-language/cpp-integer-limits?view=msvc-160 . Specifying very large system or component sizes can result in integers with values greater than that maximum. The total installed PV capacity in the world is less than 600 GW (600 million or 6e+8), and the total installed wind capacity is less than 800 GW. Again, if you are using SAM to model PV, wind, battery, or other installed capacity in a geographical region, it is best to model one or a few representative systems, and then scale up the output of those individual systems rather than modeling one very large system.

Best regards,
Paul.