There’s something fundamentally inconvenient about solar power.
It stops working every evening.
You can build the world’s largest solar farm, cover entire deserts with photovoltaic panels, spend billions on transmission infrastructure, and generate enormous amounts of electricity during the day.
And then, at around six in the evening, the sun clocks out.
No overtime.
No night shifts.
No exceptions.
For decades, this has been one of solar power’s biggest limitations.
Electricity isn’t particularly interested in whether the sun is shining.
Your refrigerator needs to run at midnight. Factories operate through the night.
And millions of people switch on their lights just as solar generation begins disappearing.
Which makes what happened in 2025 rather extraordinary.
For the first time in history, solar generated more electricity worldwide than wind.
Not by a projection.
Not by installed capacity.
By actual electricity produced.
According to Ember’s Global Electricity Review 2026, published in April, solar generated 2,778 terawatt-hours of electricity in 2025, overtaking wind’s approximately 2,715 TWh.
That’s roughly as much electricity as the entire European Union consumes in a year.
And solar accomplished this despite one rather obvious handicap.
The sun still sets.
The numbers are getting ridiculous
In 2015, the world generated just 256 TWh of solar electricity.
Ten years later, that figure had multiplied more than tenfold.
The most astonishing part is how quickly the growth is accelerating.
Solar generation increased by 636 TWh in 2025 alone, a 30% jump in a single year. That one year’s additional solar output was more than twice the entire world’s solar generation in 2015.
Read that again.
We added more than two 2015s worth of solar electricity in just twelve months.
And much of that expansion happened in one country.
China.
China added 336 TWh of solar generation in 2025, accounting for approximately 53% of the global increase. Its solar electricity output grew 40% in a single year. In fact, China added more solar electricity in 2025 than the entire world added in 2023.
India wasn’t sitting still either.
Solar generation grew by 37%, while wind generation increased by 28%.
India also added more new solar capacity than the United States for the first time.
These aren’t small improvements to an experimental technology anymore. They’re the numbers of an industry being built at extraordinary speed.
But something else was happening while everyone was watching solar panels.
And I think it’s equally interesting.
The 45% nobody should ignore
Imagine running a factory where your raw material is free, but you can only use it when the supplier decides to show up.
That’s essentially the solar power business.
Sunlight is free.
Capturing it costs money.
And using its electricity hours after sunset introduces another cost entirely.
Storage.
Historically, batteries were expensive enough to make storing large quantities of electricity difficult to justify economically in many situations.
Solar farms could generate cheap electricity at noon, but getting that electricity to customers at eight in the evening was another matter.
Then battery prices began falling.
And falling.
In 2025, battery pack prices for stationary energy storage dropped 45% in just one year, reaching approximately $70 per kilowatt-hour, according to Ember.
That’s not a typo.
A 45% decline.
Now, there’s an important distinction here. That’s the price of battery packs, not the complete installed cost of an energy storage project. You still need power electronics, construction, grid connections, financing, and other equipment.
But the direction is unmistakable.
The technology that allows solar electricity to travel through time is becoming dramatically cheaper.
And suddenly, the economics begin to look different.
What happens when electricity becomes too cheap at noon?
Here’s where the story gets particularly interesting.
In some electricity markets, solar has become so abundant during the middle of the day that prices occasionally fall below zero.
Negative electricity prices.
Think about the absurdity of that for a second.
We spent more than a century building power stations to produce electricity.
Now, during certain hours, some markets have so much electricity available that generators may effectively pay to keep supplying it rather than shut down.
This doesn’t mean every household gets paid to use electricity.
These are wholesale market conditions, and they depend on local grid constraints, contracts, and operating requirements.
But they expose a problem.
The world is getting remarkably good at producing solar electricity during daylight hours.
It hasn’t become equally good at moving all that electricity to the hours when people actually need it.
That’s where batteries enter the picture.
A battery can charge when solar electricity is abundant and discharge when generation falls and demand remains high. It doesn’t manufacture electricity. It moves it from one point in time to another, with some energy lost along the way.
And that simple capability is becoming increasingly valuable.
According to Ember, the world added approximately 250 GWh of battery storage capacity in 2025, an increase of 46% from the previous year.
The report estimates that the new battery capacity installed that year was equivalent to enough storage to shift around 14% of the average daily increase in solar generation into other hours.
Not all solar power. Not all night.
But enough to illustrate why the next chapter of solar power may look very different from the last.
The bigger story isn’t solar versus wind
There’s another number buried in Ember’s report that deserves considerably more attention.
In 2025, global electricity demand increased by 849 TWh.
Solar generation alone increased by 636 TWh.
That means solar supplied the equivalent of approximately 75% of the entire increase in global electricity demand.
Add wind, and together they accounted for 99%.
And when other clean sources, including nuclear and hydropower, were included, clean electricity generation increased by more than global demand.
The result?
Worldwide electricity generation from fossil fuels actually fell slightly in 2025.
Just 0.2%.
Hardly the end of coal or natural gas.
Fossil fuels still accounted for a large share of global electricity production.
But consider what that means.
For most of modern history, growing electricity demand meant burning more fuel. More factories, more homes, more air conditioning, more data centres, and more power stations consuming coal and gas.
In 2025, the world consumed more electricity without increasing the total amount generated from fossil fuels.
That is a meaningful structural development, even if one year doesn’t establish a permanent trend.
And it happened while solar was still fundamentally a daytime electricity source.
Imagine what becomes possible as storage gets cheaper, grids become more flexible, and electricity markets learn to use abundant daytime generation more effectively.
The sun doesn’t need to work night shifts anymore
There’s an important caveat to all this enthusiasm.
Batteries don’t eliminate solar’s intermittency. They have finite storage duration, require significant investment, and cannot independently solve prolonged periods of low renewable generation. The future electricity system will still need complementary generation, transmission networks, flexible consumption, and other forms of storage.
Nor did batteries cause solar to overtake wind in 2025. Solar crossed that milestone because of the enormous expansion in generating capacity.
But the relationship between the two technologies is changing.
For years, the question was how cheaply we could generate electricity from sunlight.
Increasingly, the question is how cheaply we can make that electricity available when the sun isn’t shining.
That’s a different engineering challenge. And potentially an even bigger commercial opportunity.
Because a solar panel makes electricity when nature allows it.
A battery gives us greater control over when we use it.
For the first time in history, the world generated more electricity from sunlight than from wind.
But twenty years from now, we might look back at 2025 and remember something else entirely.
It was the year storing sunshine started getting a lot cheaper.
And that might change everything about how we power our nights.
Sources
Ember, Global Electricity Review 2026, published April 21, 2026. Global generation, solar and wind growth, China and India trends.
Ember, 2025 in Review. Historical solar growth, new capacity additions and fossil generation changes.
Ember, Global Electricity Insights. Battery pack prices, energy storage deployment, electricity market dynamics and system flexibility.