WorldAtNet Flagship Report · Energy
Is the Global Energy System Entering Its Biggest Transformation in History?
Oil, LNG, nuclear, solar, wind, batteries, hydrogen, electric vehicles and now artificial intelligence are colliding inside a single system, and geopolitics is deciding who wins.
Every generation believes its own energy moment is unprecedented. This time the claim is harder to dismiss. The global energy system is not simply adding new sources of power alongside old ones, the way it did when gas joined coal or nuclear joined hydro. It is being pulled apart and reassembled at the same time, by five forces that rarely move together: a genuine technological revolution in renewables and storage, a resurgent nuclear industry, a war that shut down a fifth of the world's seaborne oil trade, an artificial intelligence boom that is rewriting electricity demand forecasts month by month, and a hardening contest between the United States and China over who controls the minerals and machinery that everything above depends on.
The numbers tell a story that would have seemed implausible a decade ago. In 2025, according to the International Energy Agency's Global Energy Review 2026, solar power became the single largest source of growth in global energy demand, the first time any modern renewable technology has held that position. Electricity demand grew at nearly triple the pace of overall energy demand, prompting the agency to declare that the world has entered what it calls the Age of Electricity. And yet oil, gas and coal all still grew that year too. The transition is not a straight line from fossil fuels to clean power. It is a widening system in which almost everything is expanding at once, layered on top of a geopolitical order that is more fractured than at any point since the Cold War.
Oil: War, Peaks and a Plateau
For most of 2026 the oil story has not been about the energy transition at all. It has been about the Strait of Hormuz. When war between the United States, Israel and Iran engulfed the Gulf earlier this year, tanker traffic through the strait, a route that normally carries close to a fifth of the world's seaborne crude and products, collapsed to a fraction of its usual volume. The IEA's Oil Market Report described cumulative supply losses from Gulf producers exceeding a billion barrels within ten weeks, with more than fourteen million barrels a day shut in at the peak, an unprecedented disruption by any historical measure. Prices swung wildly, briefly touching well above one hundred dollars a barrel before easing as shipping through the strait gradually resumed in June. Our earlier explainer on the Strait of Hormuz crisis walks through why that single chokepoint still holds so much power over the world economy.
Strip away the war and the underlying trend is quieter but no less significant. The IEA's longer range Oil 2025 outlook projected that global demand would keep rising only modestly toward a plateau near 105.5 million barrels a day by the end of the decade, with petrochemical feedstock, not cars or trucks, becoming the dominant source of growth from 2026 onward. Demand for oil burned as fuel, separate from plastics and chemicals, may already be peaking as early as 2027, even as jet fuel use keeps climbing. China, the engine of global oil demand growth for two decades, is expected to see its own consumption top out around the same year as electric vehicles and high speed rail displace diesel and petrol at scale. Oil is not disappearing. It is settling into a long plateau interrupted, as 2026 has shown, by sudden geopolitical shocks that can still move the entire global economy.
LNG: From Scarcity to Glut
Liquefied natural gas is heading in almost the opposite direction from oil. After four years of tight supply following Russia's invasion of Ukraine, the market is turning toward what analysts increasingly call a multi year glut. More than 174 million tonnes of new liquefaction capacity is under construction worldwide, according to BloombergNEF data cited by Bloomberg, which would lift global supply by more than forty percent by 2030. The United States accounts for roughly half of that build out, with Qatar and Canada following. Meanwhile China, once assumed to be the market's main source of future demand growth, is importing less LNG as domestic gas production rises and a new pipeline deal with Russia reduces its need for seaborne cargoes.
The IEA's own LNG capacity tracker shows more than 450 billion cubic metres a year of export capacity has reached final investment decision since 2019, with 2025 setting a record for new project sanctioning. Analysts at BNP Paribas and Morgan Stanley expect prices in Europe and Asia to fall below ten dollars per million British thermal units by late 2026, and possibly toward eight dollars in 2027. For gas exporting nations that bet heavily on LNG as a bridge fuel and a source of export revenue, the coming years look far less lucrative than the post war boom years suggested they would be.
Nuclear: The Unlikely Comeback
Few forecasters expected nuclear power to be one of the defining stories of the 2020s energy transition, yet it has become exactly that. Global nuclear generation is set to add close to fourteen gigawatts of new capacity in 2026 alone, according to Rystad Energy, the largest annual addition in almost thirty years, with most of it commissioned in China but meaningful contributions from India, Bangladesh, Turkiye and South Korea. In the United States, the Department of Energy is pursuing a target set out in 2025 to quadruple domestic nuclear capacity to four hundred gigawatts by 2050, and December 2025 brought the selection of the Tennessee Valley Authority and Holtec to pursue early small modular reactor deployments.
Small modular reactors, once a distant promise, are now attracting serious capital. The IEA's Path to a New Era for Nuclear Energy report found plans for up to twenty five gigawatts of SMR capacity already in motion, much of it driven by technology companies seeking dedicated, carbon free power for data centres. Under supportive policy, that figure could reach one hundred twenty gigawatts by 2050. Meta has entered a multi gigawatt power purchase framework with next generation nuclear developers, Rolls Royce SMR has been chosen as preferred bidder for new reactors in the United Kingdom and Czechia, and China's Linglong One, the first commercial onshore small modular reactor, is due to begin operating in the first half of 2026. Nuclear's revival is being driven less by climate policy and more by a blunt industrial need for firm, always on electricity that intermittent renewables cannot supply on their own.
Why the Nuclear Turnaround Happened
- Data centres and AI workloads need continuous, carbon free power that batteries alone cannot yet provide at scale
- Governments in the US, UK, China and across Central Europe now treat nuclear as an energy security asset, not only a climate one
- Small modular reactor designs promise factory style manufacturing that could finally break the cost overruns that plagued large plants
Solar: The New King of Power
If one technology defines this decade of the energy transition, it is solar photovoltaics. Global annual additions surpassed six hundred gigawatts for the first time in 2025, according to the IEA, bringing cumulative installed capacity to around 2,800 gigawatts, the largest of any power generation technology on the planet. Thirty countries installed more than a gigawatt in a single year, nearly double the number that did so in 2020. The IEA's Electricity Mid Year Update 2026 projects that solar output will overtake wind power in 2026 to become the world's second largest source of renewable electricity after hydropower, adding roughly six hundred terawatt hours of new generation this year alone.
That growth carries a concentration risk. China commissioned nearly 370 gigawatts of solar capacity in 2025 by itself, accounting for well over sixty percent of the global increase, and controls an estimated eighty to ninety five percent of global manufacturing across polysilicon, wafers, cells and modules. For every country trying to build a domestic solar industry, the economics of competing with Chinese scale remain punishing. Solar has become simultaneously the clearest success story of the energy transition and the clearest illustration of how dependent that transition now is on a single country's industrial base.
Wind: Steady but Strained
Wind power had its strongest year in over a decade in 2025, with annual capacity additions rising nearly forty percent to around one hundred sixty gigawatts, a new record according to the IEA. The growth came despite persistent supply chain bottlenecks, permitting delays and financing pressure that have dogged the sector, particularly offshore wind projects in Europe and the United States. China again dominated, commissioning 117 gigawatts of new wind capacity in 2025, forty eight percent more than the year before, and now supplies more than half of the world's turbine nacelles and blades.
Outside China, the picture is more uneven. Global Energy Monitor's analysis of planned wind and solar projects found that the combined pipeline in G7 countries has stayed roughly flat at around 520 gigawatts since 2023, even as calls from bodies like IRENA for G7 nations to more than double their annual additions have gone largely unanswered. Wind remains essential to the transition, but it is increasingly a story of two speeds, a Chinese buildout advancing at industrial scale and a Western pipeline struggling against grid queues, community opposition and higher financing costs.
Batteries: The Fastest Mover
No power technology is growing faster than battery storage. Global deployments reached around 108 gigawatts in 2025, up forty percent on the previous year and eleven times higher than in 2021, according to the Global Energy Review 2026. Lithium iron phosphate chemistry now accounts for around ninety percent of new deployments, up from well below half only five years ago, prized for its lower cost and longer cycle life over the more energy dense chemistries used in many electric vehicles. China alone accounted for roughly sixty percent of global additions, with its installed battery storage capacity reaching almost 150 gigawatts by early 2026, according to energy research group Ember, and a new national target of 300 gigawatts by 2030.
The rapid buildout is starting to expose its own bottleneck. Lithium prices at the start of 2026 were more than double their level a year earlier, even after remaining around seventy percent below their 2022 peak, according to the IEA's Global EV Outlook 2026. Battery costs fell eight percent over the same period thanks to manufacturing efficiency gains, but sustained mineral price increases could reverse that trend as cheaper stockpiles are drawn down. Storage has moved from a niche grid service to the technology increasingly credited with making variable solar and wind usable around the clock, but its economics remain tied to the same volatile mineral markets shaping every other part of this transition.
Hydrogen: A Reality Check
Few parts of the clean energy story have fallen further from their early promises than hydrogen. Between 2021 and 2024 the sector was defined by sweeping announcements, multi gigawatt electrolyser projects, national hydrogen strategies and corporate joint ventures worth tens of billions of dollars. By 2026, as our earlier report on the green hydrogen economy detailed, the industry has entered a far more sober phase. The IEA's committed project pipeline has shrunk by ten million tonnes to twenty seven million tonnes by 2030, and more than one hundred gigawatts of announced electrolysis capacity risk never reaching a final investment decision. Flagship projects including Australia's twelve gigawatt Hy Energy scheme and a major Air Products joint venture in Texas have been scaled back or cancelled.
Low emissions hydrogen production reached almost one million tonnes in 2025, the IEA's Global Hydrogen Review found, a milestone in relative terms but still a sliver of total global hydrogen output, most of which still comes from unabated fossil fuels. The core problem is what industry analysts call a buyer's dilemma, steelmakers, shipping firms and fertiliser producers say they want clean hydrogen but few are willing to sign the long duration supply contracts that project financing requires. Hydrogen still has a serious role to play in sectors that are genuinely hard to electrify, but 2026 has made clear that role will arrive later, and at a smaller scale, than the industry once promised.
Electric Vehicles: A Two Speed World
Global electric car sales are expected to reach twenty three million units in 2026, close to twenty eight percent of all new car sales, according to the IEA's Global EV Outlook 2026. That follows a 2025 in which sales topped twenty million for the first time, a fifth consecutive year of roughly 3.5 million unit annual growth. Around five percent of the world's car fleet is now electric, already displacing an estimated 1.2 million barrels of oil demand a day. Europe posted the strongest major market growth in 2025, up more than thirty percent, helped by tighter EU emissions standards, while emerging markets in Southeast Asia and Latin America are seeing even faster growth in electric two and three wheelers.
The United States tells a very different story. New EV sales fell forty five percent year on year in the final quarter of 2025 after federal tax credits for new and used electric vehicles were terminated, and Rest of World's analysis of the IEA data found that with virtually no government financial support remaining, American buyers, who overwhelmingly favour large SUVs, are increasingly out of step with a global market where Chinese manufacturers alone supplied around sixty percent of electric car sales in 2025. Under current policy settings, the IEA projects the global EV fleet could grow more than sixfold by 2035 to as many as 510 million vehicles, but the pace and geography of that growth will remain sharply uneven.
AI and the Demand Shock
The variable nobody fully modelled a few years ago is artificial intelligence, and it is now reshaping electricity forecasts across the industry. Global data centre electricity consumption reached roughly 565 terawatt hours in 2026, a twenty six percent increase in a single year, with AI optimised servers alone accounting for around thirty one percent of that load, according to Gartner data reported by industry trackers. The IEA's own Key Questions on Energy and AI report found that electricity demand from AI focused data centres surged fifty percent in 2025 alone, even as overall data centre demand grew a more modest seventeen percent, broadly in line with the agency's projections. Total data centre consumption is expected to roughly double to around 945 terawatt hours by 2030, a figure comparable to Japan's entire annual electricity use today.
In the United States, data centres accounted for half of all electricity demand growth in 2025. The scale of capital now flowing into this build out is extraordinary, the IEA notes that the combined capital expenditure of just five technology companies now exceeds global investment in oil and natural gas production combined, with hyperscalers reportedly planning around 725 billion dollars of AI infrastructure spending in 2026 alone. That spending is a major reason nuclear, gas turbines and grid scale batteries are all being deployed simultaneously behind the meter at data centre campuses, a distinctly new source of demand competing directly with households and industry for the same constrained grid capacity.
Minerals as the New Oil
If oil defined twentieth century geopolitics, critical minerals are shaping this one. China refines an average of seventy percent of the world's supply across nineteen of twenty strategic minerals tracked by the IEA's Global Critical Minerals Outlook, and roughly eighty five to ninety percent of processing capacity for rare earth elements essential to wind turbines, electric motors and defence systems. Export controls introduced through 2025 and tightened further in 2026, covering gallium, germanium and a widening list of rare earth compounds, have already triggered price spikes of up to sixfold for materials like neodymium praseodymium oxide and slashed European import licensing approval rates below twenty five percent, according to independent analysts tracking the disputes.
Governments are responding with the kind of urgency once reserved for oil security. The United States hosted a Critical Minerals Ministerial in Washington in February 2026 with representatives from fifty four countries, launching a new forum intended to succeed the Minerals Security Partnership, alongside a proposed Strategic Critical Minerals Reserve modelled explicitly on the Strategic Petroleum Reserve. The Pentagon has struck a floor price agreement with MP Materials guaranteeing a minimum price for domestically produced neodymium praseodymium oxide. Yet analysts at the Center for Strategic and International Studies describe the West as facing a genuine trilemma, accepting managed dependence on China, paying an estimated thirty to fifty billion dollars over five to seven years to build independent capacity, or pursuing a hybrid strategy of stockpiling and diversification, with only a narrow window remaining to choose.
Climate Adaptation Cannot Wait
All of this is unfolding against a climate backdrop that is forcing adaptation up every government's agenda, independent of how quickly the energy transition itself proceeds. 2026 is tracking toward an average global surface temperature around 1.58 degrees Celsius above the pre industrial baseline, according to monitoring cited by the Climate Adaptation Center, which would make it one of the warmest years on record. June 2026 was the second warmest June ever measured globally, with more than two dozen countries setting national temperature records, and a heatwave in France broke century old records in Bordeaux. In the American Southwest, Lake Powell has fallen toward the critical elevation below which the Glen Canyon Dam can no longer safely generate hydroelectric power, a direct example of how extreme weather is now threatening energy infrastructure itself, not only the households and industries that depend on it.
The World Meteorological Organization's review of the year's opening months also recorded deadly flooding across southern Africa, where a World Weather Attribution study found that heavy downpour intensity has increased forty percent since pre industrial times. Our companion report, The Planet on Fire, examines these events in more detail. The link back to energy policy is direct, every heatwave that strains a power grid, every drought that threatens hydropower, and every storm that damages transmission lines is now a factor that utilities and governments must plan around, whether or not the broader transition to clean energy accelerates as hoped.
Six Transitions, One System
- Oil is entering a long plateau, punctuated by war driven shocks like the Strait of Hormuz crisis
- Gas is swinging from scarcity to a multi year LNG glut as new export capacity comes online
- Nuclear and solar are both expanding rapidly, for almost entirely different reasons
- Hydrogen is retreating from its early promises while batteries scale faster than any prior power technology
- Electric vehicles are surging in China and Europe while stalling in the United States
- AI has become a new, fast growing source of demand that is reshaping investment in every other technology
What Comes After Oil
The honest answer is that no single technology comes after oil, because no single technology is replacing it. What is emerging instead is a more complicated, more electrified, and more contested system, one in which solar and batteries are scaling at a pace few analysts predicted even five years ago, nuclear has found a second life driven by data centres rather than climate targets, hydrogen has been humbled by hard economics, and oil itself remains capable of upending global markets overnight whenever geopolitics intervenes. Artificial intelligence has added a genuinely new variable, a source of electricity demand growing so quickly that it is now pulling investment into nuclear, gas turbines and storage all at once, sometimes in tension with the same decarbonisation goals that AI's own boosters claim the technology will help achieve.
Layered over every one of these shifts is a geopolitical contest over minerals, manufacturing and market access that increasingly resembles the oil politics of the twentieth century, only concentrated in a smaller number of hands. Whether this decade is remembered as the moment the world successfully rebuilt its energy system, or as the moment a fragile transition collided with war, industrial rivalry and a warming climate it could not keep pace with, will depend less on any single breakthrough technology and more on whether governments can manage all six of these transitions at once, under conditions none of them fully control.
Frequently Asked Questions
Straight answers to the questions readers ask most about the energy transition
Is the world past peak oil demand?
Not yet, but the growth rate has slowed sharply. The IEA projects global oil demand plateauing around 105.5 million barrels a day by 2030, with petrochemical feedstock overtaking road transport as the main driver of what growth remains. Demand for oil burned as fuel, separate from plastics and chemicals, could peak as early as 2027. War related disruptions, like the 2026 Strait of Hormuz crisis, can still cause sharp short term swings in both supply and price without changing that longer term trajectory.
Why is nuclear power suddenly popular again?
Two forces converged. Governments increasingly view nuclear as an energy security asset following gas supply shocks, and technology companies need firm, round the clock, low carbon electricity for AI data centres that intermittent solar and wind cannot reliably provide alone. That combination has driven a wave of small modular reactor investment, new build commitments across the US, UK, China and Central Europe, and direct power purchase deals between reactor developers and hyperscalers like Meta.
Will AI data centres cause an energy crisis?
They are already straining specific grids, particularly in Virginia, Ireland and parts of the US Southeast, where data centres can account for a fifth or more of local electricity demand. Globally, data centre electricity use is projected to roughly double by 2030, and the IEA expects data centres to account for a large share of total US electricity demand growth through the decade. Whether that becomes a broader crisis depends on how quickly new generation, transmission and storage capacity can be built to match it.
What happened to the green hydrogen economy?
It ran into hard economics. High production costs, weak long term offtake demand from industrial buyers, and reduced policy support, particularly changes to US tax credits, have led to a wave of project delays and cancellations since 2024. The IEA's committed hydrogen project pipeline for 2030 has shrunk by ten million tonnes, and installed electrolysis capacity, while still growing, remains a small fraction of early industry targets. Hydrogen retains a genuine long term role in hard to electrify sectors like steel and shipping, but that role is arriving later and smaller than originally promised.
Why does China dominate clean energy supply chains?
Decades of sustained industrial policy have given China commanding shares of global manufacturing across solar panels, wind turbine components, battery cells and, critically, the refining of rare earth elements and other critical minerals, where its average share across strategic materials tracked by the IEA is around seventy percent. That dominance has become a geopolitical lever, with export controls introduced in 2025 and 2026 causing sharp price spikes and supply disruptions for manufacturers in the US, Europe and beyond.
Are electric vehicles still growing despite the US slowdown?
Yes, strongly, just not evenly. Global EV sales are expected to reach twenty three million units in 2026, close to twenty eight percent of the total car market, driven by continued growth in China and Europe and rapid adoption of electric two and three wheelers across parts of Asia and Africa. The United States is the clear exception, where the end of federal purchase incentives led new EV sales to fall around forty five percent year on year in late 2025.
0 Comments