solar forecasting

New technologies create new jargons, some of which become mainstream.  Now that more of us have solar batteries on the roof, we are learning more about the flow of power to and from the grid and the balancing act of using independent solar power as much as possible before dipping into power from the grid.

So on a clear sunny day the supply of power from the grid drops considerably as solar power kicks in.  But if a cloud happens to cover the solar panels, then suddenly the solar power disappears and the power from the grid kicks in.  This creates a usage spike which causes a surge in the price.  This surge possibly lasts only a very short period but in that period the sellers of power make a motza.

Those managing the grid need to be able to predict and prepare for surges in use, and those selling power need to know when the price is spiking.  Normal weather forecasting is too broad-brush to be helpful in doing this, so systems have been developed using cameras on the ground, satellites in space, modelling of persistence patterns and AI to predict the availability of sunshine (or lack of it) in ten minute intervals.