Comparing Satellite-Derived Receding-Horizon and Day-Ahead Solar Forecasts: Cost Optimization of Residential PV and Battery Storage
Jamie M. Bright1, Chathurika Mediwaththe2, Paul Scott2, Nicholas A. Engerer1
1Fenner School of Environment and Society, College of Science, The Australian National University, Canberra, Australia
/2College of Engineering & Computer Science, The Australian National University, Canberra, Australia

We compare the performance of a receding-horizon (updated every 5-min throughout the day) and a day-ahead (once at start of day) solar power forecast with regards to cost optimization in application to a residential photovoltaic (PV) installation with battery storage.
Whilst benefits are offered from day-ahead forecasting, we demonstrate clear observable benefits for using a receding-horizon power forecast.
We find that receding-horizon forecasting improves upon the baseline scenario (no PV and storage) reducing the power-to-average ratio (PAR) to 46.5% (from 1.6669 to 0.7750), and offering a cost saving down to 3.9% (from 386.13 to 15.21 AU cts).

Area: Sub-Area 11.4: Solar Forecasting Applications for PV Grid Integration (Joint between Topic Areas 10 and 11)