Over the last fifteen years solar generation has grown faster than almost any energy source in history. Installed-capacity charts are impressive — an exponential curve climbing off the top of the frame. But those very charts create a false sense of completion. Capacity answers “how much can be produced at the perfect moment”, whereas the grid is broken by the opposite question: what to do when it cannot be produced at all.
Composition matters more than the sum
Start with an honest picture. Below is how the contribution of each source to total generation shifted in a model region. Watch not the height of the whole area but its stratification: coal is squeezed from below, gas and carbon-free sources are reshuffled, and solar and wind rise as a thin but fast-growing layer on top.
Stacked area: the height of each band is that source's contribution to annual generation.
The latest-year snapshot shows where the system has arrived. Renewables are already a visible share — but “visible” and “sufficient” are different words, and the gap between them is measured not on a yearly average but on one specific dark, windless evening.
Donut: the generation mix for the most recent year in the series.
The evening-peak problem
Electricity demand is not flat across the day. It has two humps: a morning one, as the city wakes, and a higher evening one, as people come home. The trouble is that the evening peak lands exactly when solar output falls to zero.
A day of demand, normalised to the peak. The evening maximum sits around 19:00.
Flexibility as a resource of its own
Engineers long ago stopped talking only about “megawatts” and started talking about “megawatt-hours” and “ramp rate”. A battery that covers the two hours of the evening peak has a different value in the system than a same-rated solar panel producing at noon. Controllable demand — when large consumers shift their load — often turns out cheaper than any new source.
What this changes for a reader
When a headline says “renewables reached N% of energy”, it is worth asking three questions. Is that a share of installed capacity or of actual output? Over what averaging window — a year, a day or an hour? And most of all: what covers the evening peak when the sun is down? The answer to the last question — coal, gas, imports or storage — says more about the maturity of the transition than any midday generation record.
