System drawing from grid when enough power is available in battery

Hi There,

I am running some modelling scenarios for distributed systems for small island states in the South Pacific. I appear to have similar issues as this previous post but it appears that the power drawn is much more significant that a 'mop-up' of voltage differences between datapoints. I am using PVWatts instead of detailed as this is just an initial high-level economic assessment of PV-Battery systems but I would have expected the battery max self-consumption charging structure to off-set a greater proportion of the load outside of the of the daytime PV output timeframes. It appears to always seek to flatten load instead of minimising load.

sam.nrel.gov/node/72582

Any guidance on modelling so that the PV and battery system are minimizing consumption from the grid would be appreciated. Perhaps I am missing something in inputs for how the system should operate?

See screenshot of the profile attached as well as core SAM file I'm using.

Thanks in advance,

Charlie

Upon reading further, I can see that the model prioritizes minimizing peak grid exposure rather than energy consumption from the grid. However, I would have thought that by scaling up the capacity and power, the system would eventually reach a size that would allow the peak-shaving structure to result in a net-zero draw from the grid.

It appears that as the battery in these scenarios also has spare charge left in the battery and intuitively, I would have thought it could reduce the peak exposure still as it would spread this discharge equally across the timeframe where there is no PV power available.

Cheers,

Charlie

Hi Charlie,

You are correct that the model is essentially trying to reduce the grid peak. The algorithm seeks to reduce the grid power purchased over the course of 24 hours within the context of the monthly peak. So for instance, if the monthly peak power from the grid is 100 kW, and the peak power over the next 24 hours is 80 kW, the battery will not discharge, since it can make no effect in the demand charge for the month.

It sounds like what you are most interested in is a dispatch strategy to maximize self-consumption. To do this, I think I would recommend the manual dispatch controller, as it gives you more fine grained control on what kind of strategy you'd like to employ. Using a manual strategy, the approach of up-sizing the system should eventually result in an "off-grid" usage scenario, where no power is ever purchased from the grid.

If you'd like, I can try and put together a quick example in one of the files you attached, just let me know.

Best,
Nick DiOrio