The grid spent 13 years getting less green

Here is a shape nobody expects to find in the climate data: from 1990 to 2003, the renewable share of the world's electricity went down. It started the period at 19.3 percent and bottomed out around 17.8 — thirteen years of a supposedly inevitable transition running in reverse. For anyone who assumes the world's grid has been greening steadily for as long as anyone has measured it, the first stretch of the record is a correction.

Nothing was wrong with the dams. The problem was the denominator. The 1990s and early 2000s were the great age of coal-fired growth in Asia, and new fossil generation swamped a renewable fleet that was still mostly hydropower. China's own renewable share slid from about 20 percent in 1990 to 15 by 2003; India's fell from roughly 25 to under 12. The world was electrifying much faster than it was greening.

Then the line turned. Wind and solar — a rounding error in 1990 — began growing faster than electricity demand itself, and the world's renewable share climbed back through its old peak and kept going, reaching about 28 percent by 2020. The steepest conversions were engineered, not inherited: Britain went from under 2 percent renewable electricity in 1990 to about 40 percent two decades into this century; Denmark went from 3 percent to 79.

The green grid didn't grow out of the old one. It had to outrun it.

That is the lesson buried in the U-turn: a share is a race. Renewables lost the 1990s not by shrinking but by growing more slowly than everything else, and they have been winning since only because wind and solar now outbuild the competition. Ten points of share gained in seventeen years, against a grid that never stopped expanding, is what winning looks like — slow arithmetic, fast construction.