Revised 14 September 2026 · v1.0.0

Liquefying natural gas takes a lot of shaft power — historically supplied by burning some of the gas itself in a turbine to drive the refrigeration compressors. That single design choice is most of the reason one tonne of LNG can carry anywhere from a best-in-class 0.15 tonnes of CO2-equivalent to several times that, depending on the plant. Global LNG supply averages just under 20 grammes of CO2-equivalent per megajoule delivered, which the International Energy Agency estimates works out to roughly 350 million tonnes of CO2-equivalent a year worldwide, about 70% of it CO2 and the rest unburnt methane. A separate estimate puts the average across several international facilities at 0.35 tonnes of CO2-equivalent per tonne of LNG produced — more than double what the best plants already report.

The gap between that average and the best performers isn't one technology. It's a stack of them, applied in different combinations depending on site, gas composition and, increasingly, what power is available nearby.

Where the bar actually sits

LNG Canada's Kitimat, British Columbia plant is the reference point most public comparisons reach for. LNG Canada itself put Phase 1's greenhouse-gas intensity at about 0.15 tonnes of CO2-equivalent per tonne of LNG when the project was announced — below British Columbia's own regulatory benchmark of 0.16 tonnes CO2e/tonne for new LNG facilities. That's a plant running on gas-turbine-driven compressors, air-cooled, with no grid electricity involved: proof that careful design of a conventional plant already gets close to top-quartile, without electrification.

Five levers, one table

Waste-heat and turbo-expander figures: turbomachinery trade literature summarised in a GPPS (Global Power and Propulsion Society) viewpoint on decarbonizing gas-turbine-driven LNG plants. Electric-motor efficiency range and CCS status: see sourcing in the sections below.

The standout lever: replacing the turbine with a motor

An earlier piece on this site looked at a smaller-scale version of this same idea — the Starter-Helper Motor, an electric machine that shares a gas turbine's shaft and takes on part of its compression load without replacing it, documented at Qatar's RasGas Train 3 and in a Toshiba retrofit case study. That piece also cited a 2023 Siemens Energy feasibility paper comparing a straightforward driver swap against replacing the whole turbine-compressor string with an electric one: a single-shaft industrial gas turbine runs at roughly 25–30% efficiency, an aeroderivative turbine at 38–42%, against an electric motor exceeding 95%. Once transformer, drive and transmission losses are counted, a full electric-drive design still comes out to about 40% overall drivetrain efficiency against 34–36% for the mechanical-drive design it replaces — a real but much smaller gap than the headline motor number suggests, once you count everything upstream of the motor too.

That efficiency case is showing up in order books, not just feasibility studies. Electric-motor-driven configurations held about 13.4% of the LNG liquefaction train market in 2025 and are projected to grow at roughly 15.4% a year through 2035 — the fastest-expanding drive category tracked. Built examples now exist at real scale: Freeport LNG in Texas is reportedly the world's largest all-electric LNG plant, with 675 MW of installed electric power built in from the start rather than retrofitted; Woodfibre LNG in British Columbia, powered by BC Hydro's largely hydroelectric grid, is reported to cut its emissions profile roughly 14-fold versus a conventional gas-turbine design; and ADNOC's Ruwais LNG project in Abu Dhabi is positioned as the first LNG export site in the Middle East and Africa to run on grid power rather than on-site combustion.

Siemens Energy efficiency comparison and Starter-Helper Motor case studies (RasGas, Toshiba): as sourced in "The Helper Motor Trick That Already Adds LNG Throughput Without a New Turbine" on this site. Market-share and growth figures: LNG Liquefaction Train Market report, Global Market Insights. Freeport LNG, Woodfibre LNG and ADNOC Ruwais: reported in industry trade coverage (Oil & Gas Journal, LNG Industry, GE Vernova) — none of these three has been independently verified against the operator's own technical filings for this piece.

What the grid actually costs the electrification case

LNG Canada's own proposed Phase 2 is the clearest live example of electrification meeting a grid constraint rather than an engineering one. The plan under discussion would run refrigeration compressors on electric motors fed by BC Hydro's hydroelectric grid instead of gas turbines — and public reporting describes the project's economics and emissions performance as depending on how quickly BC Hydro can build the transmission capacity to reach the remote Kitimat site, not on any unresolved question about the compressor technology itself. Absent that power, the fallback is more gas turbines, which would narrow or erase the emissions advantage electrification was meant to deliver.

The same theme shows up in the other direction at Hammerfest LNG in Norway — already electrified, and already covered on this site in the helper-motor piece linked above — where a 2024 technical paper from Innomotics and Equinor engineers found that grid voltage disturbances can trigger compressor surge in a way a self-contained gas-turbine train never has to deal with, and that the drive's under-voltage ride-through characteristics needed specific strengthening to avoid a meaningful hit to plant availability. Trading a fuel-supply risk for a grid-reliability risk is a real engineering trade, not a free upgrade.

Combined cycle: the step that doesn't need a grid connection

For a plant that is going to keep running gas turbines regardless — because grid power isn't available, or the economics don't favour it — combined-cycle configuration is the lever that doesn't depend on anyone else's infrastructure. Turbine exhaust carries roughly 60% of the fuel's lower heating value at 425–550°C; routing that heat through a heat-recovery steam generator to make additional power or process steam, instead of venting it, is standard practice on new large-scale trains and is characterised in industry literature as lowering fuel consumption compared with simple open-cycle operation. It's the closest thing on this list to a decarbonization lever with no site-specific dependency.

Cooling: the least settled question on this list

Refrigerant condensing temperature sets how much compression work a train needs per tonne of LNG, and the cooling medium sets the condensing temperature. Water or seawater cooling generally allows a lower approach temperature than air cooling, which is why it shows up as an efficiency lever in the technical literature on direct seawater cooling for LNG plants. But it isn't a free win: it requires reliable water access, adds intake, treatment and discharge infrastructure, and brings its own permitting and environmental-impact questions, especially at sites facing water-scarcity pressure. Industry commentary on cooling-technology trends increasingly points toward hybrid or adiabatic systems — air as the primary medium, water used only at peak ambient conditions — as a way to capture some of water cooling's efficiency benefit without carrying its full water demand. This is the one lever on this list where the right answer is genuinely site-specific rather than generally decided.

Carbon capture: the backstop, not the first move

Even a fully electrified, combined-cycle, optimally cooled plant still has CO2 to deal with — from the feed gas itself, where CO2 is stripped out before liquefaction, and from whatever combustion remains. That's the gap carbon capture and storage is meant to close, and 2025–2026 has been the period where CCS tied to LNG and adjacent gas processing has moved from announcement to active construction rather than staying a slide in a strategy deck. Venture Global has described CCS as a core, actively-scaling part of its LNG facilities' environmental strategy; ExxonMobil has CCS projects coming online in 2026 designed to remove up to 1.2 million tonnes of CO2 a year from gas streams feeding Gulf Coast markets, LNG included; and Woodside's 2024 acquisition of Tellurian brought a Louisiana LNG project with an associated carbon-capture-linked ammonia development into its portfolio. None of this is cheap the way efficiency upgrades are — CCS adds capital and operating cost rather than saving it — which is exactly why it belongs at the end of the lever list, not the start of it.

What doesn't follow from any of this

None of the sources used here support a claim that any single lever gets a conventional LNG plant to zero, or that electrification is available to every site regardless of grid access. LNG Canada Phase 1's 0.15 t CO2e/t LNG figure is the project's own announced target, not an independently audited outcome, and the global 0.35 t CO2e/t LNG average cited above comes from a different methodology and system boundary than a single project's own reported number — the two aren't a like-for-like comparison without knowing exactly what each includes (flaring, fugitive methane, feed-gas CO2 handling all vary by how a figure is scoped). What the public record does support is the ordering: efficiency and heat-recovery measures are close to universally adopted already, electrification is the largest remaining lever where power allows it, cooling method is a genuine but site-specific trade-off, and carbon capture is the tool for whatever is left after the others have been applied.

Sources and caveats