Ras Laffan in Qatar remains the world's largest LNG export complex on paper, with 14 liquefaction trains. Missile strikes on 18 and 19 March 2026 took out Trains 4 and 6, worth 12.8 million tonnes a year, so the biggest fully available plant in 2026 is Cheniere's Sabine Pass at about 30 mtpa.
That gap between nameplate capacity and available capacity is the reason I rebuilt this ranking rather than copying a 2025 table. Two of the ten largest plants are wholly or partly out of service as of August 2026, and Freeport is one of them. Every entry below states what the number measures: design capacity, capacity currently able to load, or capacity still under construction.
The largest LNG export terminals in 2026
| # | Terminal | Country | Capacity | Basis |
|---|---|---|---|---|
| 1 | Ras Laffan | Qatar | 77.1 mtpa national capacity across 14 trains, about 64.3 mtpa available after the March 2026 strikes | Nameplate, reduced by 12.8 mtpa of damage |
| 2 | Sabine Pass | United States | About 30 mtpa from 6 trains | Operating nameplate |
| 3 | Bintulu (MLNG) | Malaysia | 29.3 mtpa from 9 trains, close to 30 on the complex total | Operating nameplate |
| 4 | Corpus Christi | United States | Above 25 mtpa now, heading to 28.9 mtpa once Stage 3 is complete | Nameplate, expansion 93 percent complete |
| 5 | Plaquemines | United States | 2.6 Bcf/d nominal, 3.2 Bcf/d peak at full build, roughly 20 mtpa | Phase 1 in service Q4 2026, Phase 2 mid-2027 |
| 6 | Golden Pass | United States | 18.1 mtpa peak from 3 trains | First export cargo 22 April 2026, still ramping |
| 7 | Gorgon | Australia | 15.6 mtpa from 3 trains on Barrow Island | Operating nameplate |
| 8 | Freeport | United States | About 15 mtpa, currently offline | Halted after an explosion and fire, restart expected late 2026 |
| 9 | North West Shelf, Karratha | Australia | 14.3 mtpa after the permanent retirement of a second train | Current nameplate, down from 16.7 |
| 10 | Cameron | United States | 13.5 mtpa | Operating nameplate |
Two structural facts sit behind that table. Six of the ten plants are on the US Gulf coast, and the single largest site is not American but Qatari. Country totals are where published figures diverge most sharply. The United States appears at roughly 102.3 mtpa on a nominal export-capacity basis and at 136.5 mtpa on an operational basis at the start of 2026, a spread of 34 mtpa on the same country in the same year, because the higher count credits peak rates at the new Gulf trains. In flow terms the Energy Information Administration puts US exports near 17 Bcf/d averaged across 2026.
Australia shows the same problem in miniature: 81.8 mtpa on an export-capacity estimate, 85.1 mtpa across 10 plants in February 2026, and above 87 mtpa on a fuller operational count. Qatar sits at 77.1 mtpa nameplate with 12.8 of it unavailable. None of these figures is wrong, and I would not compare a number from one basis against a number from another.
Ras Laffan lost two trains, and the repair clock runs in years
The missile attacks on Ras Laffan Industrial City damaged Trains 4 and 6, together rated at 12.8 mtpa, plus the Pearl GTL plant on the same site. Energy Minister Saad Sherida Al-Kaabi put the effect at 17 percent of Qatar's LNG export capacity and an estimated 20 billion dollars of lost annual revenue. QatarEnergy expects repairs to run 3 to 5 years and declared force majeure, with long-term buyers in Italy, Belgium, South Korea plus China exposed to reduced deliveries.
For anyone planning gas-linked freight, the practical consequence is that those volumes moved to the spot market. That is the mechanism which pulled US cargoes toward Europe and Asia through spring 2026, and it is why the ranking above is unstable in a way an LNG terminal list normally is not. Qatar's answer is North Field East, a 32 mtpa addition originally targeted for commissioning in 2026 and now reported as slipping into 2027, followed by North Field South at 16 mtpa around 2028 on the path to roughly 142 mtpa.
Sabine Pass turned 10 and still leads the US fleet
Sabine Pass shipped its first cargo in February 2016, and 10 years later its 6 trains, rated at about 30 mtpa, make it the anchor of American export volume. Two projects reshaped the US position during 2026. Golden Pass, the QatarEnergy and ExxonMobil venture at Sabine Pass in Texas, loaded its first export cargo on 22 April 2026 and is rated at 18.1 mtpa peak across 3 trains. Plaquemines, Venture Global's Louisiana plant and the eighth US export terminal, produced first LNG in December 2024 and reaches 2.6 Bcf/d nominal once both phases finish, with Phase 1 in service in the fourth quarter of 2026.
Corpus Christi is the third mover. Stage 3 was 93 percent complete during 2026 and takes the site above 25 mtpa immediately, with 28.9 mtpa the figure once every train is running. Those additions show up in monthly flow data rather than in annual capacity tables. US exports hit a record 11.7 million tonnes in March 2026, an estimated 17.9 Bcf/d and the second-highest month on record, with Europe taking most of it. Full-year 2026 exports are expected to rise about 13 percent.
Freeport is offline, which the capacity tables do not show
Freeport LNG, rated near 15 mtpa, halted after an explosion and fire, with a restart expected toward the end of 2026. That is the second large outage in this ranking within a single year, and it lands on a plant with no war-risk explanation behind it. Freeport has a history of unplanned stoppages, so I treat its nameplate as a ceiling rather than a supply assumption when advising on cargo timing.
Put the two outages together and roughly 28 mtpa of nameplate capacity, Qatari plus American, was unavailable at some point in 2026. Set that against the roughly 300 mtpa held by the 3 largest exporting countries combined, and it explains why spot pricing stayed firm through a year of record commissioning.
Traders read such interruptions through separate lenses. Underwriters price business-interruption exposure, buyers examine diversion clauses in their sale agreements, and owners watch whether substitute cargoes lengthen ballast legs. One stoppage on the Texas coast therefore reshuffles chartering arithmetic far beyond the Gulf, and the ripple reaches berths that handle no gas at all.
Bintulu is the plant most rankings forget
Petronas has been liquefying gas at Bintulu in Sarawak since the 1980s, and the complex now runs 9 trains at 29.3 mtpa after Train 9 added 3.6 mtpa to the previous 25.7 mtpa. That places it within 0.2 mtpa of Sabine Pass, which makes the second and third positions a rounding question rather than a real hierarchy. I treat both as joint runners-up to Ras Laffan when advising on where sustained loading capacity exists outside the Atlantic basin.
Australia competes on plant count, not plant size
Australia reached second place by country without owning a single site of Qatari or Louisianan scale. North West Shelf at Karratha, running since 1989, now stands at 14.3 mtpa after a second train was permanently retired rather than merely idled, which is a real reduction from the 16.7 mtpa that older tables still carry. Gorgon on Barrow Island adds 15.6 mtpa from 3 trains drawing on the Gorgon and Jansz-Io fields. The balance of national capacity is spread across 8 further plants, which is a different risk profile: no single strike or outage removes a seventh of the country's output, but no single site offers the loading flexibility of Ras Laffan either.
What terminal capacity does and does not tell a charterer
- Nameplate is a design figure. Trains run above or below it depending on ambient temperature, feed gas composition plus maintenance cycles, so a 15.6 mtpa plant is not a promise of 15.6 mtpa.
- Berth count sets loading rhythm. Two jetties at a 30 mtpa plant produce a very different queue from four jetties at the same capacity.
- Force majeure moves cargo, not capacity. When Qatar declared it, the tonnage did not vanish from the market; it was replaced from further away, which lengthened voyages and tightened vessel supply.
- Under-construction capacity is not capacity. Plaquemines Phase 2 sits in mid-2027, and North Field East has already moved from 2026 into 2027.
- Unplanned outages are not rare events. Freeport at about 15 mtpa is offline pending a restart late in 2026, and no capacity table published in January would have shown it.
- Political geography is part of the number. Yamal LNG keeps exporting under sanctions pressure with most cargoes going to Asia in the summer months, while the related Arctic LNG 2 project has been effectively blocked.
If you move gas-adjacent freight rather than gas itself, the useful read is voyage length. Replacing Qatari volume from the US Gulf adds sea miles, and added sea miles absorb ships. We saw the same mechanism in reverse when Red Sea routing changed, which I covered in our Suez Canal and Red Sea routing guide. The fleet side of this market is set out in our ranking of the largest LNG shipping companies, and the Gulf transit question is handled in the Strait of Hormuz bypass guide.
Common questions
Which is the largest LNG export terminal in the world in 2026? Ras Laffan in Qatar, by design capacity, at 77.1 mtpa across 14 trains. After the March 2026 strikes removed 12.8 mtpa, about 64.3 mtpa is available, and the largest plant that is both undamaged and running is Sabine Pass at about 30 mtpa.
Which major plants are out of service in August 2026? Two. Ras Laffan Trains 4 and 6, worth 12.8 mtpa, face 3 to 5 years of repair, and Freeport at about 15 mtpa is halted after an explosion and fire with a restart expected late in 2026. Together that is roughly 28 mtpa of nameplate capacity.
How long will Qatar's damaged trains stay offline? QatarEnergy has said 3 to 5 years. Trains 4 and 6 are the two affected units, and the operator declared force majeure on affected long-term contracts rather than trying to cover them from elsewhere in the complex.
Is the United States now the largest LNG exporter? By capacity, yes, on either basis: 102.3 mtpa nominal or 136.5 mtpa operational at the start of 2026, against above 87 mtpa for Australia and 77.1 mtpa for Qatar. March 2026 exports of 11.7 million tonnes were a monthly record, and volumes are forecast to grow about 13 percent across the year.
Why do terminal capacity figures differ between sources? Because one source counts design capacity, another counts what trains can currently load, and a third adds projects due within 18 months. Australia appears at 81.8, 85.1 as well as above 87 mtpa for exactly this reason, and the United States at 102.3 or 136.5. I state the basis for each number above so the comparison stays honest.
Sources: QatarEnergy statements on the March 2026 Ras Laffan damage, US Energy Information Administration export and capacity data, operator disclosures from Cheniere, Venture Global, Golden Pass, Petronas, Woodside plus Chevron. Capacity figures are nameplate unless stated otherwise and change as trains commission or come offline.