Charted: China Alone Is 27% of World Energy Demand — Top-3 Clears 49%

Our energy-systems research ledger answered the country question: how do large systems source, mix, and trade energy — primary stacks, electricity fossil shares, and import dependence from Japan’s ~88% to Australia’s deep export surplus. The EI Statistical Review update and later Ember / IEA mid-year refreshes then answered the vintage question: what moved in world TES, renewables-vs-coal in power, and capex. This post answers the distribution question desks trade next: how concentrated is the system at the top?

The interactive dashboard above is built as a concentration lens. Toggle Top-k ladder, Demand curve, Fuel exports, and Import risk. Filter by lens (demand / export / production / manufacturing) and metric (Top-1 / Top-3 / HHI). The punchline is deliberately multi-sided. On primary demand, China alone is about 27% of world TPES and the top three (China + US + India) clear ~49%. On fuel exports, concentration is higher still: seaborne coal’s top-3 bloc is ~72%, LNG ~61%, crude oil a more plural ~38%. On clean-tech manufacturing, the tip goes extreme — solar modules near 80% China, battery cells near 75%. Same theme, three different tops.

The headline ladder: Top-1 and Top-3 across energy lenses

Read the table as a family of market shares, not one slogan. Demand Top-1 tells you who burns and builds the energy stock. Export Top-3 tells you who can move molecules across oceans. Manufacturing Top-1 tells you who can make the hardware that greening depends on. Averaging “China’s energy share” across these rows is a category error — 27% of demand is not 80% of modules.

Demand: a thick top without a monopoly

Toggle Demand curve. The Lorenz-style panel for world TPES rises to ~27% at rank 1, ~43% at top-2 (China + US), and ~49% at top-3 once India enters. An equal-share line across six buckets sits far below until the rest-of-world residual. That geometry is oligopoly, not monopoly: no single country owns half of primary energy, but three countries own nearly half.

Coal consumption is the exception inside the demand family. China’s ~56% top-1 and ~74% top-3 make thermal coal a different concentration class from oil demand, where the United States leads at only ~16% and the top-3 (US + China + India) lands near 40%. Electricity generation sits between them — China ~32%, top-3 ~55% — which is why power-system stories often feel more China-centric than crude markets even when primary oil still looks plural.

Pair this with the global electricity generation mix when you need country-level stacks rather than world-share ladders. The concentration cut answers who owns the denominator; the mix cut answers what fills each country’s numerator.

Fuel exports: oligopolies with different tips

Open Fuel exports. The stacked bars put Top-1 / Top-2 / Top-3 / rest on one axis per commodity. Coal is the hardest seaborne market in the panel: Indonesia alone ~35%, top-3 with Australia and Russia ~72%. LNG is a flexible triopoly — United States ~22%, Australia and Qatar close behind, top-3 ~61% — the same corridor our LNG capacity post tracks from a capacity angle. Pipeline gas still shows Russia as top-1 (~18%) with Norway and Canada completing a ~48% bloc, even after European route rewiring. Crude oil exports remain the most plural major fuel: Saudi Arabia ~15%, top-3 only ~38%.

That ordering matters for security analysis. A desk that treats “energy trade concentration” as one number will mis-rank risk: coal and LNG concentrate at the top; crude does not to the same degree. HHI bands in the ladder view make the same point continuously — manufacturing rows sit in the extreme band (≥5,000), coal export and coal use in the high band, crude export in the plural band.

Manufacturing vs molecules: the clean-tech tip is sharper

Stay on Fuel exports → Export vs manufacturing contrast, or filter the ladder to Manufacturing. Solar module capacity near 80% China and battery cells near 75% China are a different political economy from Saudi’s 15% of crude exports. Greening the power mix can raise manufacturing concentration even while it lowers the fossil share of electricity in Ember-style censuses.

This is the bridge to our chokepoint commodities concentration work: midstream graphite, rare-earth separation, and refined gallium print even higher Top-1 scores than solar modules. Energy systems concentration and critical-minerals concentration are cousins — one tracks fuels and demand, the other tracks the materials inside the hardware stack.

Import risk: high dependence is not the same as high Top-1

Toggle Import risk. The scatter puts import dependence on the x-axis and fossil primary share on the y-axis, with bubble size proportional to √primary EJ. Japan (~88% import dependence, ~86% fossil primary) and Korea (~82% / ~84%) sit in the stress quadrant. The EU-27 prints ~58% import dependence on Eurostat’s disclosed meter with still-high fossil primary (~69%). China is a large importer in absolute barrels and molecules but only ~22% dependent on the net-energy meter because domestic coal still dominates TPES. The United States prints mildly negative (net exporter); Australia and Saudi print deep export surpluses.

Concentration at the export tip and concentration of import exposure are related but not identical. A world where Indonesia, Australia, and Russia clear 72% of seaborne coal exports is a concentrated supply market; a world where Japan and Korea sit above 80% import dependence is a concentrated vulnerability map. Both belong in the same risk packet. Our research ledger already ranked import dependence; this lens asks how that exposure sits next to Top-1 / Top-3 supply shares.

Who is exposed under a concentration reading

More exposed: Northeast Asian importers whose power and industry still clear peaks through imported LNG and oil while solar/battery hardware also routes through China-dominated manufacturing; European systems that cut Russian pipe gas but remain ~58% energy-import dependent; thermal coal buyers who face a three-country seaborne tip; and analysts who treat Ember’s renewables-over-coal crossover as if it dissolved trade concentration.

Relatively better positioned: net energy exporters (US, Canada, Australia, Middle East producers) who sit on the surplus side of the import scatter; buyers who can diversify across plural crude markets even when LNG stays a triopoly; and grids that pair domestic renewables with storage fast enough to shrink import dependence without waiting for manufacturing to decentralise.

What would rewrite the next concentration cut: a Statistical Review year where India’s TPES share jumps enough to push top-3 demand past 55%; US LNG share climbing past 30% of exports; a material diversification of solar module capacity outside China that drops Top-1 below 60%; or a European import-dependence print that breaks clearly under 50%.

Caveats and methodology

• Different denominators. TPES demand, seaborne export tons, LNG volumes, and manufacturing capacity are not interchangeable. Top-1 percentages across rows are comparable as concentration meters, not as additive shares.
• Primary ≠ electricity. China’s ~27% of TPES is not the same object as ~32% of world electricity generation or Ember’s renewables-vs-coal mix shares.
• Rounded country shares. EI Review and secondary IEA manufacturing figures are rounded; derived HHI values are approximate and sensitive to how the residual “rest of world” is bucketed.
• Eurostat vs national balances. EU-27 import dependence (~58%, 2023) is disclosed; non-EU import figures are estimated from national balances / IEA and marked accordingly.
• Manufacturing shares are secondary. Solar module and battery cell Top-1/Top-3 use IEA supply-chain vintages labeled secondaryuseful for order-of-magnitude concentration, not for a single-year EIA-style census.
• Trade stance is net. Indonesia can be a coal exporter and oil importer; the stance label nets those flows.
• This post is a concentration companion. For country mix ledgers use the research post; for vintage deltas use the EI update and later Ember/IEA refreshes.

The shareable takeaway

Energy systems are concentrated at the top — but which top depends on the lens. China alone is ~27% of world primary energy demand; China + US + India clear ~49%. Seaborne coal and LNG export markets concentrate harder (72% and 61% top-3). Crude oil exports stay comparatively plural (38% top-3). Clean-tech manufacturing is the extreme tip: solar modules ~80% China, battery cells ~75%. Import dependence still peaks in Japan and Korea even when supply-side Top-1 labels sit elsewhere. The theme answer is sharper: countries source and mix energy in many ways, but trade and manufacturing power sit in thick tops that mix-share headlines alone will miss.

Related reading: Energy systems research, Energy systems EI update, Global electricity generation mix, and Chokepoint commodities concentration.