Solar output across the US hit 45.5 terawatt hours in May against 43.4 from coal, the first month on record where the panels finished ahead of the fuel that ran the country for a century

A generation table is a boring document.

Rows of fuels, columns of months, numbers that shift a little and mostly stay in the order they were in last year.

One row moved past another this spring.

Solar and storage still took 91 percent of all new grid capacity in that quarter, which says more about how little else is being built than about how much solar is.

The margin between them came to about 5 percent, which in a table like this is close to nothing and also the entire point.

Nothing unusual happened to make it.

No outage, no heat wave, no policy switch, just an ordinary month in which two long running trends finally crossed.

Why the crossing had to happen in spring

Solar output does not peak in May.

Total generation peaks in June or July, when the days are longest and the sun is highest, and any solar plant will make more in high summer than it does in late spring.

Its share of the mix is a different curve.

Share peaks earlier, because the sun is already strong in April and May while cooling demand has not started, so the denominator stays small while the numerator is already large.

Coal was moving the other way.

Retirements accelerated through the last five years as gas and then solar undercut it on running cost rather than build cost, which is what strands a plant that has already been paid off.

What 45.5 terawatt hours is

A terawatt hour is a billion kilowatt hours.

A typical American house gets through roughly ten thousand kilowatt hours in a year, so the month’s solar output is about 4.5 million houses supplied for twelve months, generated in thirty one days.

It came off a very scattered machine.

Rooftops, parking canopies, ranch land, retired farmland and desert, spread from Maine to Hawaii, with 6 million separate installations now counted nationwide.

None of it is one thing anybody can point at.

That is the difference between this and a coal plant, because a coal plant is a place, and solar output is an accumulation.

The numbers behind the crossing

The month came in at 45.5 terawatt hours of solar and 43.4 of coal.

Solar was up 17 percent on the same month a year earlier. Coal was down 11 percent.

The five year picture is starker.

Solar held 5.4 percent of the mix in May of 2021 and 12.8 percent in May of 2026, while coal fell from 19.7 percent to 12.2 percent across the same five Mays.

The analyst who published the figures put a rank on it.

Nicolas Fulghum of Ember called it a move from a niche contributor to the third largest and fastest growing source of power in the country, which is behind gas and nuclear and now ahead of everything else.

What the coverage left out

The fleet that did this was built earlier.

New installations in the first quarter of 2026 came to 7.8 gigawatts, which sounds healthy until you set it beside the same quarter a year before, because it is a 27 percent drop.

Solar and storage still took 91 percent of all new grid capacity in that quarter, which says more about how little else is being built than about how much solar is.

The forecast is flatter still.

Residential is expected to fall 21 percent this year, the five year national outlook has additions running roughly flat even as electricity demand rises, and 457 projects sit with pending permits exposed to delay.

Coal did not collapse either.

It hit an all time monthly low of 39.3 terawatt hours in April and then rose to 43.4 in May.

So the crossing happened because solar climbed faster than coal recovered, and the same crossing is forecast for Texas on its own grid this year.

What decides whether it holds

Summer is the test.

Cooling load lifts total demand sharply in July and August, which grows the denominator faster than the panels grow the numerator, and solar’s share can slip back under coal’s.

Storage is what changes that argument permanently.

On the ninth of July the batteries on the California grid discharged 12.99 gigawatts at once, covering 36 percent of regional demand at the peak.

The following afternoon solar itself supplied 72 percent of demand across the same system.

Texas has been setting the same kind of marks, with batteries covering a fifth of evening demand in March and renewables meeting 79 percent of the load the day after.

Coal’s remaining case is the dark hours.

A plant that exists to cover the evening loses that role the moment a battery fleet can hold the afternoon and release it after sunset, which is precisely what six separate records in the first half of this year describe.

California has been living in that future for months, where solar has outrun gas on 82 percent of days this year.

Four other regional markets set their own solar records inside three months.

May’s number will not stand as a record for long.

What will stand is the ordering, because one row has now finished above the other in a published table, and no revision puts it back.