West Africa signs off a $25bn gas pipeline that will outlive us all

Climate & Energy 4 min 80 sources

Fourteen nations backed a 6,000km Atlantic gas line to Europe - one of a run of decades-long fossil bets being locked in this week, even as solar quietly overtook coal.

Key takeaways

West Africa signed off a $25bn, 6,000km gas pipeline to Europe, committing the region to gas for decades before a single metre is built.
The same week brought a record gas-turbine backlog and new gas plants, even as US solar out-generated coal for the first time.
The choice that shapes emissions is rarely this year's fuel; it is the long-lived asset built once and locked in for 30 to 40 years.

A 6,000km bet on gas, signed in Freetown

West African leaders have formally endorsed the Nigeria-Morocco Atlantic Gas Pipeline, one of the continent’s most ambitious energy projects. [41] The 15-nation regional bloc ECOWAS signed the framework on Sunday evening in Freetown, Sierra Leone, clearing what analysts call the last major political hurdle before financing. [37][41]

The numbers are large. The pipeline would run about 6,000km along the Atlantic coast of 14 African nations, carrying Nigerian gas to Morocco and then into Europe’s grid via Spain. [41] Full cost: roughly $25bn. [41] At capacity it would move 30 billion cubic metres of gas a year and serve 400 million people. [41] Construction is expected to begin in 2028, built in phases over several years. [41]

This has been a decade in the making - first proposed in 2016, negotiated across more than a dozen countries. [41] Backers argue it could keep more value in Africa, where gas is often extracted, processed abroad, then sold back at three or four times the price. [41] The counter-fact sits inside the plan itself. A pipeline finished in the 2030s and paid off over decades commits the region to burning gas for decades more. That runs well past the point most climate targets assume it stops.

The week the world ordered decades of gas

The pipeline is the biggest of several long-lived fossil bets placed this week. GE Vernova, a major turbine maker, said its backlog of gas-turbine orders and reservations hit 116 gigawatts, and expects 125 GW by year-end. [35] It is now taking reservations for 2031 deliveries. [35] The buyers span about 100 companies in 26 countries; a fifth are data-centre operators. [35]

A gas turbine ordered today runs for 30 or 40 years. So does a pipeline, or a new power plant. In Wisconsin, advocates are pushing regulators to look at the long-term picture before approving new gas plants, warning that once built, the investment locks in for decades. [9] The choice being made is not which fuel to burn next year. It is a commitment that no single later year can undo.

Meanwhile, solar quietly passed coal

The same week carried the opposite signal. In May, US solar generated more electricity than either coal or wind for the first time - 47,147 gigawatt-hours, against coal’s 45,119. [65] Over the first five months of 2026, renewable generation rose 10.1% from a year earlier, while coal output fell nearly 11%. [65] Solar and wind together made 57% more power than US coal plants. [65]

But look closely and the lock-in shows here too. Even as renewables surged, US natural gas capacity still rose by about 6,300 megawatts over the year. [65] Clean energy is winning on the margin - the next unit of power built is increasingly solar or wind. [65] The gas being added now, though, is not this year’s decision. It is the 2050s being decided now, in steel and turbines that will still be running long after the ribbon is cut.

Much of the new gas demand traces to data centres. Off-grid gas plants built to power them “won’t fix the problem,” one clean-energy analysis argues, because they hard-wire fossil supply into computing for the life of the asset. [33]

The heat these bets are set against

The stakes are not abstract. France recorded about 5,700 more deaths than usual during its June heatwave, its health agency said. [62] Across Europe, a scientific analysis found record heat is “sucking soils dry,” with the drought made 80 times more likely by human-caused warming. [3] The thirsty atmosphere is now pulling water from rivers, crops and the very reservoirs that generate power. [3]

That is the tension in one week’s news. The physical world is registering the emissions already released, while the infrastructure being signed this week fixes decades more into place.

The other long bet: nuclear

Not every decades-long commitment points the same way on carbon. Governments are also locking in nuclear, which runs for 60 years but emits almost nothing while it does. India said it is advancing small modular reactors, targeting five nuclear units by 2033. [22] The US and Saudi Arabia reached a nuclear cooperation deal, and the Trump administration is pushing to speed new reactors, including for AI data centres. [18][10]

The lesson is not that gas is bad and nuclear good. It is that the biggest climate decisions are rarely about this year’s fuel. They are 30- and 40-year bets, made once, in concrete and steel - and very hard to unmake once poured.

02 · Lesson · why it matters

The decision that gets made once and lasts forty years

Some choices you make every year. The ones that shape the climate you make once, in steel and concrete, and then live inside for decades.

The choice that isn’t a choice anymore

This week West Africa signed off a $25bn gas pipeline. Fourteen countries, 6,000km, gas flowing from Nigeria to Europe. Construction starts in 2028. It runs for decades after that.

Here is the strange thing. Once the framework is signed and the steel goes in the seabed, nobody in 2045 gets to vote on whether West Africa burns gas. That was decided this week, in Freetown, by people who will mostly be retired before the first molecule flows.

This is the shape of almost every big climate decision. It looks like a choice about fuel. It is really a choice about time.

Yearly choices and one-time choices

Think about the difference between renting and buying a house.

Renting is a yearly choice. Each year you can leave, move, change your mind. Buying is a one-time choice that binds you for years. You take on a thirty-year mortgage in a single afternoon, and that afternoon shapes three decades.

Energy works the same way. Whether you turn the heating up tonight is a yearly choice - reversible, small. Whether your country builds a gas plant, a pipeline, a coal terminal, a highway - that is the afternoon that binds the decades.

Economists have a plain name for it: lock-in. A long-lived asset commits the emissions and the behaviour that come with it, for its whole life. You do not get to un-decide it each year. The decision was already made, once, at the start.

Why the asset, not the fuel, is the real bet

A gas turbine ordered today runs for thirty or forty years. So does a pipeline. So does a plant. The company ordering it is not deciding to burn gas in 2027. It is deciding to burn gas until the 2060s.

That is why the numbers this week matter more than they look. One turbine maker’s order backlog hit 116 gigawatts, with slots booked out to 2031. Each of those orders is a forty-year commitment placed now, by a hundred different buyers, in twenty-six countries.

And here is the part that catches people out. The same week, US solar out-generated coal for the first time. Clean energy is clearly winning the yearly race - the next unit of power built is more and more likely to be solar or wind.

Yet new gas capacity still went up. Both things are true because they answer different questions. The yearly question - what do we build next - is going green. The forty-year question - what did we commit to, in assets that will still be running in 2055 - keeps getting answered “gas.”

The trap of the sunk asset

Why can’t we just switch it off later, when the world has changed its mind?

Because a built thing fights to keep running. It cost billions. Someone borrowed to build it and must be paid back. People work there. A town depends on it. The money is already spent - it is sunk - and a plant that is cheap to keep running looks like a waste to close early.

So the asset earns its keep by running, and running means emitting, long after we wish it wouldn’t. The concrete does not care about next year’s climate summit. It was poured to last, and it lasts.

This is what makes lock-in different from an ordinary bad habit. A habit you can break next Monday. A $25bn pipeline you cannot. The whole point of building it was to make it permanent.

Who’s inside this, and how little any seat can see

None of this makes the pipeline villainous, or the people signing it foolish. West Africa has real reasons to want its own gas: to keep value at home, to power factories, to have a seat at the table. Each choice, seen alone, is reasonable. The lock-in is not any one person’s doing.

That is the humbling part. No single signer in Freetown decides the 2050s. But together, across a hundred rooms, the decades get quietly filled in. A pipeline here, a turbine order there, a plant approved in Wisconsin, a highway widened somewhere else. And you are downstream of all of it. The emissions warming the summer that killed 5,700 people in France last month were committed, mostly, by afternoons like this one, years ago.

The reader can’t see the whole grid of commitments from any one seat. Neither can the people making them. That is worth holding onto when the next “energy decision” scrolls past. Ask not what it burns this year. Ask how many years it was built to last.

03 · Lab · your turn

The Lock-In Ledger

Rehearse how a one-time build decision commits decades of emissions, and feel why the cheapest path is often the trap.

04 · Hope · carry this

The same week we locked in decades of gas, solar quietly out-generated coal for the first time. The thing we build most cheaply now is finally the clean thing - and that is the bet that gets easier every single year.