The honest version of this comparison is that it is rarely decided on cost alone, and when it is, the answer is usually “pipeline, if one is possible”.
The shape of the two cost curves
A pipeline’s cost is dominated by distance. Every extra kilometre is more steel, more right of way, more compression to push gas along it. The plant at either end is comparatively modest.
LNG is the opposite. The liquefaction plant and the import terminal are enormous fixed costs incurred whether the cargo travels five hundred miles or eight thousand, and the marginal cost of extra distance — a few more days of charter and boil-off — is small.
Draw both curves and they cross. Onshore the crossover falls somewhere between roughly two and four thousand kilometres, depending on terrain, volume, and how expensive money is at the time. Below that a pipeline wins comfortably. Above it, LNG.
Water moves the crossover a long way
The crossover distance assumes a pipeline can be built. Across deep water it often cannot, and where it can the cost per kilometre rises steeply: subsea pipelaying, deepwater compression, and repair operations that need specialist vessels rather than an excavator.
So the practical rule is less about distance than geography. Two land masses with a continuous route between them favour a pipeline out to a few thousand kilometres. Two land masses separated by an ocean favour LNG at almost any distance, which is why the trades that define the LNG market — Qatar and Australia to east Asia, the US Gulf to Europe and Asia — are all trades no pipeline could serve.
Volume matters as much as distance
A pipeline is efficient at high, steady throughput and terrible at low or variable throughput, because the capital is sunk regardless. LNG scales in cargo-sized increments and, more importantly, in ship-sized increments: a market that needs supply for three winters can charter an FSRU and send it away afterwards.
That is why small and seasonal markets frequently choose LNG even at distances where a pipeline would look cheaper on a spreadsheet. There is no such thing as a temporary pipeline.
Optionality is a real asset
A pipeline connects one seller to one buyer. Both are then locked in: the seller has no other outlet for that gas and the buyer has no other source down that pipe. That mutual dependency is sometimes stable for decades and sometimes becomes the whole story, as Europe discovered.
A cargo, by contrast, can go anywhere a ship can reach. The value of that flexibility does not show up in a delivered-cost comparison but it is priced in the market constantly: it is why destination clauses were fought over, why arbitrage between basins exists, and why buyers pay for portfolio supply rather than a single source.
Set against that, LNG’s flexibility has physical limits. Cargoes still have to pass maritime chokepoints, and a canal closure re-routes a fleet.
Speed of delivery
A pipeline of any length is a multi-year project with a right of way to acquire, often across borders. An import terminal is a multi-year project too — unless it floats, in which case a country can be importing within about a year of deciding to.
When supply is needed urgently, that difference decides the question on its own, regardless of the long-run cost comparison.
Emissions
Piped gas is generally the lower-emitting route for the same molecule, because it avoids liquefaction’s 8 to 12 per cent energy cost and the fuel burned moving a ship. That advantage narrows over very long pipeline distances, where compression fuel accumulates.
The comparison is also swamped by something neither chain controls well: upstream methane leakage. Fugitive emissions at production and along transmission are poorly measured and heavily weighted in any lifecycle calculation, and a leaky pipeline chain can easily be worse than a tight LNG one. Anyone quoting a confident number for either should be asked what leakage rate they assumed.
What this means for reading the data
This site carries both: 829 LNG terminals and 3,748 gas transmission pipelines from the same tracker. They are not alternatives in the data the way they are in a project decision. Pipelines feed liquefaction plants and carry regasified gas away from import terminals, and the proximity links between the two datasets are there to show where the chains meet.
Where you do see them competing — a proposed import terminal in a market a new pipeline could also serve — the tracker will usually carry both projects at proposed status for years, and only one will ever be built.