China Can Buy U.S. LNG Without Shipping It to China
Chinese buyers can purchase U.S. LNG for delivery elsewhere, so China-bound export data alone cannot establish whether a purchase commitment has been met. After the Trump-Xi summit, any LNG pledge needs its own counting rules.
Primary lensTrade policy
Sub-topicPolicy monitoring
Evidence base7 records used
Use casePolicy monitoring
A Chinese company can buy American liquefied natural gas without sending it to China. For exporters assessing the Trump-Xi summit, that is more than a shipping detail. A sale to a Chinese customer can be commercially valuable while adding nothing to China's physical imports.
China's official readout of the September 24 talks says the economic teams reached a new joint arrangement. It gives no LNG purchase volume, delivery timetable or tariff schedule. The readout therefore provides no basis for adding a summit-related LNG commitment to a sales forecast. If a purchase announcement follows, its value will depend partly on what it counts. An existing contract filed with the U.S. Department of Energy shows why that question deserves attention before the headline quantity does.
USTR has already raised related measurement questions. Its June 5 Board of Trade consultation sought comments on monitoring trade flows by dollar value and timing, and on sharing data between the two governments. LNG illustrates the problem of choosing a measure. The notice did not establish LNG eligibility or a rule for counting purchases. Counting purchases is distinct from calculating the tariff savings on a negotiated basket.
China Gas can choose another destination
The agreement is between Venture Global Plaquemines LNG and China Gas Hongda Energy Trading. A public summary filed with DOE describes a February 2023 contract for approximately 1 million tonnes a year. Purchase and delivery obligations begin at Phase 2 commercial operation, subject to the contract's conditions. Contractual delivery takes place at the connection between the terminal's loading arms and the LNG tanker, before the cargo reaches a foreign port.
The destination provision is equally consequential. It lets the buyer choose among countries permitted by the applicable DOE export authorizations and U.S. law. A separate provision permits resale or transfer to authorized countries, or to purchasers that accept the relevant destination restrictions in writing. The buyer's Chinese identity does not require the cargo to enter China.
For a portfolio buyer, those rights allow a U.S. purchase to serve customers elsewhere. Suppose China Gas directs an authorized cargo to a European terminal. The U.S. seller has made a sale to its Chinese customer, but the cargo is unloaded in Europe. There is no contradiction between those accounts.
The filing establishes that possibility under this agreement. It does not establish the terms of every Chinese purchase, or rule out later amendments. Still, it gives an exporter a concrete reason to resist treating a buyer's nationality as a destination requirement. The contract must supply that requirement.
DOE's destination data cannot measure all Chinese buying
DOE's monthly reporting instructions keep purchaser and destination information separate. Reports follow the month of departure. Destination means the country of physical unloading, but exporters may report an anticipated destination if it is unknown when the report is due, then correct it. A China destination total consequently cannot show all the U.S. LNG that Chinese companies purchased for their global portfolios.
The destination also stops short of identifying ultimate consumption. DOE's destination-reporting policy replaced its former end-use-country tracking requirement with actual delivery reporting. Gas delivered in one country can have a subsequent journey that this field does not describe.
When reconciling a sales forecast with DOE data, the exporter needs to match cargoes by reporting period and destination. A difference between the customer account and the country total calls for examining the transaction. It does not establish that either record is wrong.
China already appears in the pre-summit export record
DOE's July report, released September 22, lists China as the destination for 4.6 billion cubic feet of U.S.-produced LNG exports by vessel in July and the same rounded amount in June. The January-July total was 9.1 Bcf. These figures describe exports reported before the talks. They do not establish July arrival dates or a summit commitment.
For scale, DOE recorded 518.7 Bcf of total U.S.-produced LNG exports in July and 3,686.4 Bcf over January-July. China's share of the latter was roughly 0.25%, calculated from the published figures. The recorded volume was small, but it supplies a baseline for assessing later claims of growth.
These totals cover U.S.-produced LNG, including exports through Mexican terminals. They should not be presented as a count of direct shipments from U.S. ports to China. Establishing that route requires terminal-level records. The aggregates also do not reveal the relevant buyers, tariff treatment or reasons for the deliveries.
A purchase pledge needs its own counting rules
Any subsequent LNG commitment should first be read for its definition of an eligible purchase. A commitment measured by Chinese purchasers might count cargoes delivered elsewhere. For one measured by imports into China, the buyer's identity alone would be insufficient. The agreement would need to supply the accounting rules. The current summit readout specifies neither an LNG obligation nor a method for counting one.
Existing contracts matter here. An announcement could refer to newly signed volume, previously contracted supply or some combination, depending on its terms. An exporter assessing the benefit needs to establish which applies and when delivery obligations begin. Adding an announced quantity without resolving those points risks counting sales already in the forecast.
A gap between a company's purchases and DOE's total reported for China would not, on its own, establish a contract breach. The company may have exercised destination rights like those in the Plaquemines agreement. Nor would permitted resale necessarily earn credit toward a government pledge. That answer would have to come from the pledge's own rules.
A national export forecast needs a further check. EIA's September 1 review of the first half of 2026 described exports near maximum output and attributed growth to added terminal capacity. A Chinese sale could change a cargo's destination without adding production. Forecasting higher total U.S. exports therefore requires evidence of available capacity or higher utilization, separate from the purchase amount.
An exporter can value an enforceable sale on its contractual terms without waiting for tariff relief. Crediting the sale toward a government promise requires the promise's own rules. Subsequent exports reported for China can be compared with the pre-summit series, with shipment records used to verify arrivals. That reconciliation should precede any claim that a sales announcement has produced the same quantity of Chinese imports.
From reading to review
Run the numbers on your lane.
The duty calculator runs the current stack for any HTS code and origin. A free account opens full tool output, AD/CVD detail, Chapter 98 processing, and available exports.