How to model battery in SAM from a system wide approach i.e. taking into account other generation resources in isolated grid ?

Hi, (attention to Nick DiOrio)

I am doing a study on an isolated grid system (Alice Springs, QLD), modelling scenarios of high PV uptake (50-70% of energy mix) with battery storage. I used SAM to model PV and trying to brainstorm how to approach the battery. I realize from the webinars and documentation that SAM battery model works behind the meter. Is there a way to adequately work around this limitation and consider the battery within the energy system along with other generation resources other than additional PV (e.g. existing PV + gas plant)?

My initial thought was to use the net load profile supplied by the local generator for PV+battery modelling (e.g. gas generation), but I am not quite sure how to reconcile this with existing solar if the battery is to achieve the desired load matching system wide.

Dear Amr,

As of SAM 2017.1.17, SAM's battery model is available with the Detailed Photovoltaic model when combined with the PPA Single Owner financial model, which means that you can use it to model a power project with no building load, i.e., one that is not behind the meter.

In that configuration, only the manual dispatch mode is available, so you have to manually define the charge and discharge schedule of the battery using the six dispatch periods and weekday/weekend matrices on the Battery Storage input page. The system's electrical power output is the sum of the PV power and battery power. (The battery power is negative when the battery is charging.)

You could model the PV/battery system in this way, and then outside of SAM, determine how the gas plant would need operate given the PV/battery system's output.

You might also consider using HOMER for this kind of analysis. It is designed to model power systems with a mix of different resources, and is useful for analyzing different dispatch strategies.

Best regards,
Paul.

Hi Amir,

Sorry for the delayed response, and thanks to Paul for answering the question. Paul's recommendations are what I would also recommend. It sounds like it may be difficult to precisely model your scenario in SAM, but it's possible you could model the PV and battery component with your modification to the net load profile to reflect the influence of a local generator. As you note, it not be easy to optimize the mix of these sub-systems in this way, though perhaps through an iterative approach you could arrive at a reasonable analysis solution. This may be more in the HOMER territory, though we'll be curious to hear if you are able to accomplish your modeling with SAM.

Best regards,
Nick