Smart grids are supposed to flatten the demand curve by reducing peak power (i.e. with demand response) and filling valleys (i.e. with EV charging). Price arbitrage using local energy storage is another driver for a flat demand curve.
So do you think in a few years the demand will be completely flat?
I bet it shouldn’t, let me explain why…
Which is a smarter grid, (1) one with so much distributed storage that the real demand can behave any way and it will be compensated locally so the resulting demand is seen as flat from the generation side and prices vary very little or nothing? Or (2) a grid where the -real- demand increases if there is cheap renewable available (windy and sunny days), and decreases when there is not so much renewable power available? I think it is easy to understand that the latter can be more efficient and cheaper, because the less storage needed, the lower the system price will be. A kWh of wind power that is consumed after being stored in a battery can be cheaper that a real-time diesel generated kWh but not cheaper than the same wind kWh without being stored.
An advanced and competitive future grid should be more able to shift load than to store generation. The second option is actually not so flexible, it means investing in additional assets in order not to need flexibility on the demand. (Yes, of course there will be a balance.)
Today generation mainly follows demand, and capacity has to be planned for the peak power consumption. If demand is more able to follow renewable generation availability, an efficient power system can be achieved without flattening the load curve.
In short and in an illustrative sense, by “surfing a wavy supply curve”…