Analysis
LNG Canada’s partners approved Phase 2 on 28 September 2026, and Fluor reported a notice to proceed for its JGC joint venture the following day. The project would add two trains at Kitimat and raise stated capacity from approximately 14 to 28 million tonnes per year. These notices establish an investment and construction commitment, not another 14 million tonnes of LNG already available to buyers.
The economic question as of 6 October is how much of that incremental nameplate can be converted into reliable cargoes at a profitable netback. The press releases identify process units, storage, berth capacity and feedgas transport work, but give neither a complete project capital budget nor a commissioning date. Any unit-cost estimate built by dividing one contractor’s award by 14 million tonnes would therefore omit material scopes.
Evidence and measurement
LNG Canada says Phase 2 includes two trains, an LNG tank, a condensate tank, another berth and wider utility and processing systems. Fluor says its venture will perform engineering, procurement, fabrication, construction and commissioning, and will recognise a US$7.5 billion Fluor share. That number is one contractor’s share of a contract: it is neither the full joint-venture contract nor the owners’ all-in capital expenditure.
The distinction matters because an export terminal combines liquefaction with marine loading, storage, power, water, commissioning, contingency and upstream interfaces. A hypothetical division of US$7.5 billion by the 14 Mt/year increment yields about US$536 per annual tonne of capacity. This is only a scale check on Fluor’s share, not a published project cost per tonne or an investment return.
Engineering chain
Nameplate 28 Mt/year assumes the physical chain can deliver gas at specification, run refrigeration equipment and load ships. A train can be mechanically complete while production remains limited by gas treatment, rotating-equipment reliability, storage availability or shipping windows. Two trains double installed capacity only if connected systems and operating performance support them.
At a purely illustrative 85% utilisation, incremental annual output from 14 Mt/year would be 11.9 Mt; at 70%, it would be 9.8 Mt. The 2.1 Mt gap is 15% of nameplate. These figures use capacity times assumed utilisation; no source says either utilisation rate will occur. They also exclude LNG used or lost inside the system and commercial allocation among partners.
Economic mechanism
Netback should be tested per delivered tonne: sales receipt less feedgas purchase, pipeline charge, liquefaction fuel and operation, shipping, destination costs and any contract adjustments. Because neither realised sale price nor the full cost stack is disclosed in the two releases, a positive margin cannot be inferred from the FID alone. A long-term sale contract may smooth revenue but its terms remain essential.
For a sensitivity illustration, each US$10 per tonne change in netback across 11.9 Mt/year changes annual pre-tax contribution by US$119 million. At 9.8 Mt/year it changes by US$98 million. This multiplication is not a valuation: it excludes financing, taxes, working capital, depreciation and the timing of startup. It shows why utilisation and small unit-margin changes interact.
Illustrative sensitivity
The 670-kilometre Coastal GasLink system is another gate. LNG Canada says it will manage expansion, including five new compressor stations, but the statement gives no independently verified incremental pipeline throughput. Compression can increase flow only within pipe pressure, hydraulic, gas-quality and operating limits. A terminal schedule and a gas-delivery schedule must converge.
An investor or buyer should track separate milestones: compressor commissioning, contracted feedgas volumes, train mechanical completion, first LNG, sustained output and cargo departures. First LNG is not equivalent to a full commercial run rate. Delay in any one element can create idle downstream capital even if another package meets its own schedule.
Uncertainty and decision gates
As of the evidence cutoff, no completion date, project-wide capex or realised operating margin is established by these primary notices. The most useful future disclosures would state full funding, contracted feedgas and transport rights, train startup sequence and output after commissioning. Until then, 28 Mt/year is an installed-capacity target with execution and market exposure.
A disciplined appraisal keeps Fluor’s US$7.5 billion share in its proper scope and tests several utilisation and netback combinations. It should revisit the case when owners publish audited cost and operating data. The scenario figures here are transparent arithmetic, not a forecast of completion, LNG prices or investment returns.