The war that closed the Strait of Hormuz at the end of February has done two things to American LNG at once. In the short run it turned every working US liquefaction train into a strategic asset, with Europe and Asia bidding hard for flexible cargoes. In the long run it pulled forward a fresh round of investment decisions on the Gulf Coast. Since March, three large US export projects have reached final investment decision (FID), and a fourth North American project followed in Canada in September. The question for US gas markets is not whether these projects are good business on paper. It is when their gas arrives, and what the world will look like by then.
Our reading is that the 2026 class of US LNG projects was sanctioned in a crisis but will be delivered into a very different market. The supply gap opened by the Gulf outage is concentrated in 2026 and 2027. The plants approved this year are scheduled to start up in 2030 and 2031, after the largest wave of liquefaction capacity in the history of the industry has already landed. That timing gap matters for buyers, for the traders who signed much of the offtake, and for the US gas producers who will have to feed the plants.
What was approved, and when it starts
The sequence is short and easy to follow. On 13 March Venture Global announced FID and an $8.6 billion project financing for Phase 2 of CP2 LNG in Louisiana, bringing total financing for the project to $20.7 billion. The company puts CP2's peak production capacity at 29 million tonnes per annum (Mtpa) and says it has sold nearly all of that on long-term contracts, mainly to customers in Europe and Asia. On 15 May Caturus announced FID on the 9.5 Mtpa Commonwealth LNG terminal in Cameron Parish, closing $9.75 billion in project financing within total commitments of $21.25 billion, with operations to begin in 2030. On 3 June Delfin Midstream took a $5 billion FID on its first floating liquefaction vessel, which it describes as the first FLNG facility in the United States, with expected capacity of 4.4 Mtpa and first LNG scheduled for 2030.
The International Energy Agency's Global LNG Capacity Tracker, updated on 1 October, counts four FIDs between January and October 2026 worth nearly 46 billion cubic metres per year (bcm/yr) combined: the three US projects plus LNG Canada Phase 2 in September. By the tracker's own nameplate figures (7.6 bcm/yr for CP2 Phase 2, 12.9 for Commonwealth and 6.0 for Delfin), the US share comes to roughly 26.5 bcm/yr. The IEA's Gas Market Report for the third quarter places the start of commercial operations for the enlarged CP2 project in the second half of 2030, with Delfin and Commonwealth also starting in 2030. LNG Canada Phase 2 is listed for 2031.
None of this gas helps this winter, or the next one. That is the central fact of the 2026 investment class.
The gap these projects were meant to fill
The scale of the disruption explains the appetite. According to the IEA, LNG loadings from Qatar and the United Arab Emirates fell by 35 bcm year on year between March and June, while non-Gulf production rose by around 27 bcm, offsetting about three quarters of the loss. For the full year, the agency expects combined Qatari and Emirati LNG supply to fall by about 45 per cent, or 54 bcm, with new projects in North America, Africa and Australia adding close to 50 bcm. Spot prices in March reached their highest monthly averages since January 2023, and in the second quarter Platts JKM averaged $17.5 per million British thermal units, 45 per cent above a year earlier.
The more important figure for investors is the IEA's estimate of cumulative LNG supply losses between 2026 and 2030: about 140 bcm, once the near-term disruption and the medium-term damage at Ras Laffan are combined. The agency says this is equivalent to 15 per cent of the new LNG supply set to be added globally over the period, and that the impact is largely concentrated in 2026 and 2027. In other words, the shortage is front-loaded. Markets may stay tighter than expected for two years, and then the original supply wave resumes its course.
Arriving into the peak of the wave
That wave is enormous. The IEA tracker says more than 330 bcm/yr of new export capacity is due between 2025 and 2030 from projects already under construction at the start of 2025, the largest build-out over any comparable period. Annual additions from post-FID projects are expected to rise from around 30 bcm/yr in 2025 to a peak of around 90 bcm/yr in 2028, before tapering in 2029 and 2030.
Much of that wave is American and is already in steel. The tracker lists Port Arthur Phase 1 and Rio Grande Trains 1 to 3 for 2027, CP2 Phase 1 for 2027, Corpus Christi Midscale Trains 8 and 9 for 2028, Louisiana LNG for 2029, and Port Arthur Phase 2 and Rio Grande Train 4 for 2030, with Rio Grande Train 5 in 2031. Golden Pass, the joint venture of QatarEnergy and ExxonMobil, sent its first cargo on 22 April and expects to export about 18 million tonnes a year once its three trains are running, with Trains 2 and 3 to follow once Train 1 is stable.
The Gulf itself is not standing still either. Qatar's North Field East expansion, rated at 43.5 bcm/yr, has been pushed back to at least the first half of 2027 after the attacks on Ras Laffan, and North Field South, at 21.8 bcm/yr, is listed for 2028. The UAE's Ruwais LNG is also listed for 2028. Delay is not cancellation. By 2030, when the 2026 US class begins to load cargoes, Qatar's expansion trains are scheduled to be operating alongside the full 2027 to 2029 American cohort.
So the projects sanctioned in response to a shortage will, on current schedules, start up after the shortage has passed and as the market digests the tail of the biggest supply surge it has seen. That does not make them bad investments. Long-lived infrastructure is built for decades, not for one winter. But it does mean the price environment that helped close the financing is unlikely to be the one these plants are born into.
Who carries the risk
Here the contract books matter. US liquefaction projects are financed against long-term tolling or sale and purchase agreements, so the developer is largely insulated from spot price swings. The risk passes to the offtaker. CP2 is described by its developer as nearly fully contracted. Caturus said in September that about 8.5 Mtpa of Commonwealth's 9.5 Mtpa is subscribed under long-term agreements, and its named counterparties include EQT, Glencore, Mercuria, PETRONAS and Aramco Trading. Delfin's first vessel is backed by agreements with Vitol, Expand Energy, Centrica and Gunvor.
That list is telling. A good share of the 2026 offtake sits with commodity traders, portfolio players and US producers rather than with state utilities buying for captive demand. These buyers will place cargoes wherever the netback is best. In a well supplied market around 2030, the margin between Henry Hub plus liquefaction and shipping costs on one side, and European or Asian hub prices on the other, will decide how hard these plants run. Utilisation of US capacity, which has looked guaranteed throughout 2026, may become a commercial variable again at the turn of the decade.
The domestic side of the ledger
For the US gas market, the story is about feedgas demand that is locked in years ahead. The Energy Information Administration's September Short-Term Energy Outlook forecasts US LNG exports rising from 15.1 billion cubic feet per day (Bcf/d) in 2025 to 17.4 Bcf/d in 2026 and 18.6 Bcf/d in 2027. Dry gas production is forecast at 111.7 Bcf/d in 2026 and 115.9 Bcf/d in 2027, with marketed production rising by 4.5 Bcf/d and 4.6 Bcf/d in those years. The Permian and Haynesville together account for more than 70 per cent of the forecast growth. The EIA expects Henry Hub to average $3.43 per MMBtu in 2026 and $3.28 in 2027.
Those numbers describe a supply base that is so far keeping pace with export growth. The 2026 FIDs extend the demand curve into 2030 and 2031, beyond the STEO horizon, and they come on top of the 2027 to 2029 cohort. The integrated model that Caturus is pursuing, pairing its own South Texas production of more than 1 Bcfe/d with its terminal, is one answer to the feedgas question. Most of the industry still relies on the broader market, and on new pipelines such as the Hugh Brinson line out of the Permian, which the EIA notes began interstate shipments in June and will ramp up through early 2027.
The next tranche is already queuing
The pipeline of candidates has not thinned. On 15 September Caturus announced a five train, 7.75 Mtpa expansion that would take Commonwealth to about 17.25 Mtpa, targeted for the early 2030s. Cheniere Partners signed a lump-sum EPC contract with Bechtel in May for the first phase of the Sabine Pass expansion, a single train with expected capacity of over 6 Mtpa including debottlenecking, and expects FID by early 2027, subject to FERC and DOE non-FTA approvals that were pending at the time. On 30 September Glenfarne welcomed a South Korean investment of more than $50 billion in Alaska LNG, a 20 Mtpa terminal at Nikiski fed by an 807 mile pipeline, saying commercial agreements now cover 13 Mtpa of the 16 Mtpa it is targeting for FID.
Meanwhile, the only levers that added US supply quickly this year were regulatory. In March the US government authorised Plaquemines LNG to raise its exports by 13 per cent, or 4.6 bcm/yr, and in early April Elba Island was authorised to increase exports to non-FTA countries by 22 per cent, or 0.8 bcm. Uprates and debottlenecking at existing plants respond within months. Greenfield projects take four years or more.
What to watch
Three markers will show whether the 2026 class is a well-timed bet or an expensive hedge. First, the pace of Gulf recovery: the IEA's forecast assumed a full reopening of Hormuz in the third quarter and restored operations at undamaged facilities early in the fourth, and any slippage lengthens the tight window. Second, the next round of FIDs, led by Sabine Pass Phase 1, the Commonwealth expansion and Alaska LNG: each additional decision taken on the strength of 2026 prices adds to the 2030s supply stack. Third, the forward curves for European and Asian gas in 2030 and beyond, which will tell the traders holding this offtake how much optionality they have actually bought.
The 2026 investment class is a reasonable response to a real shock. Buyers want diversified, contract-backed supply after watching a fifth of the world's LNG trade disappear behind a closed strait. But the shock is short, and the plants are long. US policymakers and producers should plan for a 2030 market in which American LNG competes on cost in a crowded field, not one in which every cargo sells itself.