On 10 September, Australia published three Bills. From 1 January 2028, no one will export LNG from Australia without a licence, and every licence will carry a quantity of gas the holder must sell at home — set each year by a regulator and adjustable by a minister.

The policy is branded a 20 per cent reservation. Applied to Australia’s own legislation and its own regulator’s forecasts, it reserves about seven cargoes in its first year.

The gap between the headline and the mechanism is worth knowing. The reason to read on is narrower: Australia had to legislate for a power Canada already holds, and has never used.

In 60 seconds

  • The 20% is a ceiling, not a quota. The binding number is a “demand calibrated quantity” the regulator sets each June. The explanatory memorandum is explicit: it “cannot be more than the 20 per cent”.
  • Our model says seven cargoes in 2028. Above what the Gladstone ventures already sell domestically, the scheme diverts about 26 PJ in its first year — 0.6% of Australian LNG exports. By 2030 it reaches 33 cargoes, or 2.9%.
  • It hardens because Bass Strait is emptying. The obligation rises 60% between 2028 and 2030 while the scheme itself does not change. Southern production falls 46%.
  • Western Australia is the control group. It reserved 15% in 2006, never intervened again, and its exporters already plan more domestic gas in 2028 than the new federal formula requires.
  • The government is describing its own Bill incorrectly. Its explainers say 110% of forecast demand. Section 24(5) says 110% of demand less non-exporter supply. In 2028 the difference is 27 cargoes.
  • An Australian exporter is buying Canadian LNG. Six days after its CEO said every cargo Canada locks in is one Australia foregoes, Santos signed for 1 mtpa from Ksi Lisims. Its stated reasons are purely commercial — but under the drafting, holding international supply may make relief from the domestic obligation harder to obtain.
  • Bottom line: the volume is noise. The instrument is not. Every North Asian long-term contract signed in 2026 went to Qatar, the United States or Malaysia — none to Australia.
Exposure draft Bills of 10 September 2026; consultation closed 24 September. Forecasts from the AEMO 2026 GSOO and ACCC Gas Inquiry, June 2026.
169 PJThe 2028 obligation, against a 272 PJ ceiling
−46%Southern Australian production, 2026 to 2030
$910mMaximum penalty for missing the obligation

What Was Actually Legislated

Three Bills went out on 10 September and consultation closed a fortnight later. They repeal the Australian Domestic Gas Security Mechanism and the domestic price cap, and replace both with something more durable: a licence. Verified

Licences run 20 to 50 years and cannot be transferred. The obligation attaches not to the project but to “the person who holds legal title to the liquefied natural gas immediately before” loading — the lifter, which in Australian LNG usually means a foreign equity partner. Missing the obligation carries a maximum penalty of $910 million. Verified

The 20 per cent everyone quotes is section 17. But section 24 overrides it whenever a smaller number applies. Each June the Australian Energy Regulator sets a “demand calibrated quantity” for every licence holder, and where that is lower it becomes the obligation. The explanatory memorandum leaves no doubt: the calibrated amount “cannot be more than the 20 per cent DSO”. Verified

So the real question is how the regulator will calibrate. Section 24(5) says the target is forecast domestic demand, less forecast supply from everyone who is not a licence holder, plus 10 per cent of that residual. The minister may reduce the 10 per cent to zero. Verified

The government’s own explainer, its roundtable slides and its ministers all describe this as “110% of forecast demand”. That is a different calculation, and a much larger one. The Bill governs. Interpretation

Fig. 1The ceiling, and the formula underneath it
East coast domestic supply obligation, PJ. Navy: the 20% ceiling, which is also what the government’s own wording produces. Gold: what section 24(5) produces.
0100200300PJ20% ceiling 272 PJ272As drafted 169 PJ1692028gap: 27 cargoes20% ceiling 283 PJ283As drafted 252 PJ2522029gap: 8 cargoes20% ceiling 269 PJ269As drafted 269 PJ2692030ceiling binds20% ceiling — also what the government’s own explainer producesWhat s24(5) produces
Source: Insight Stream calculation applying s24(5) of the exposure draft Domestic Gas Reservation Bill 2026 to AEMO 2026 Gas Statement of Opportunities forecasts (Step Change, 26 March 2026) and AEMO Queensland LNG feedgas. Cargo equivalents at 70,000 t.
How we modelled itForecast demand and non-exporter supply are AEMO’s own published Step Change figures; non-exporter supply is proxied by southern (Victoria, NSW, South Australia) production. The 20% ceiling is applied to Queensland LNG feedgas. Conversions: 54.4 PJ per Mt LNG, derived from Geoscience Australia’s paired 2023 figures; one 70,000 t cargo = 3.81 PJ. Varying the supply proxy by a quarter in either direction moves the 2028 obligation between 91 and 246 PJ — the conclusion holds across the range. The regulator’s actual method will be a legislative instrument, and it has not been drafted.

Seven Cargoes

Run the formula and the first regulated period produces an obligation of about 169 PJ. The three Gladstone ventures already sell roughly 143 PJ a year into the domestic market on contract. The incremental gas — the gas that would not otherwise have stayed home — is about 26 PJ. Estimate

That is seven cargoes, against Australian exports of roughly 1,127 cargo-equivalents a year. Six-tenths of one per cent. Estimate

Fig. 2What “20 per cent” delivers
Share of Australian LNG exports, with cargo equivalents. Australia exported an estimated 78.9 Mt in 2025–26.
The headline“20% of LNG exports reserved”20.0% of exports20.0%≈225 cargoesThe gross obligation, 2030What exporters must sell at home6.3% of exports6.3%≈71 cargoesThe incremental diversion, 2030Above what they already sell at home2.9% of exports2.9%≈33 cargoesThe incremental diversion, 2028First year of the obligation0.6% of exports0.6%≈7 cargoesAustralian LNG exports ≈ 78.9 Mt a year, about 1,127 cargo-equivalents. Shares are of that total.
Source: Insight Stream calculation. Export baseline: Department of Industry, Science and Resources, Resources and Energy Quarterly, June 2026. Domestic sales baseline: ACCC Gas Inquiry June 2026 interim report, Table 1.1.

Even the gross obligation — counting gas the ventures already sell at home — peaks at 6.3 per cent of exports, under every assumption we tested. Estimate

The government’s own claim is “up to 200 additional petajoules” a year, and it justifies the scheme against “AEMO forecast possible shortfalls of up to 140 petajoules”. Verified The 140 PJ appears nowhere in AEMO’s central case; it belongs to the Accelerated Transition scenario, in 2034 and 2036. In the central Step Change case the largest gap anywhere in the series is 106 PJ, in 2035 — and in 2028, the first year the obligation bites, the forecast gap is zero. Verified

The fair answer to that: AEMO puts the probability of an annual supply gap at 59 per cent by 2030 and 100 per cent from 2033, and the ACCC expects structural shortfalls from 2030. A scheme that takes three years to stand up has to be legislated before the problem arrives. The timing is defensible. The sizing rhetoric is not. Interpretation


It Hardens Because Bass Strait Is Emptying

The obligation rises about 60 per cent between 2028 and 2030. Nothing in the scheme changes. What changes is the denominator.

Fig. 3The residual widens on its own
East coast domestic demand against supply from producers other than the LNG exporters, PJ. The shaded residual is what section 24(5) multiplies by 1.10.
0100200300400500PJ20262026 demand 475.8 PJ2026 non-exporter supply 318 PJ20272027 demand 454.8 PJ2027 non-exporter supply 319 PJ20282028 demand 436.2 PJ2028 non-exporter supply 283 PJ20292029 demand 436.9 PJ2029 non-exporter supply 208 PJ20302030 demand 435.8 PJ2030 non-exporter supply 170 PJDomestic demand436 PJSupply fromnon-exporters318 → 170 PJthe residual×1.10 = the obligation
Source: AEMO 2026 Gas Statement of Opportunities, Step Change, 26 March 2026 — Figure 9 data workbook and Table 8. Non-exporter supply proxied by southern (VIC, NSW, SA) available production.

AEMO forecasts southern production falling from 318 PJ to 170 PJ by 2030, a 46 per cent decline driven by the depleting Gippsland fields. Demand barely moves. The residual the formula acts on widens by more than 100 PJ without a single policy decision. Verified

By 2030 the calibrated quantity finally exceeds the 20 per cent ceiling, and the ceiling starts to bind. The bite of Australia’s reservation scheme is set by the decline rate of Bass Strait. Interpretation


Western Australia Already Did This, Twenty Years Ago

There is a natural experiment inside Australia, and it is the part of this story most useful to anyone still building export capacity.

Fig. 4Western Australia meets the new obligation before it starts
PJ, 2028. The federal formula applied to WA’s own residual, against domestic gas WA’s LNG project participants have already told the state they expect to supply.
WA obligation, 2028The formula applied to WA’s own residual290 PJ290 PJAlready planned for 2028Under existing State Agreement commitments312 PJ312 PJWA wrote 15% into each State Agreement in 2006, before the plants were built.
Source: Insight Stream calculation on AEMO 2025 WA Gas Statement of Opportunities (19 December 2025) demand and production data; commitments from the WA Government’s WA Domestic Gas Statement, November 2025.

Western Australia reserved 15 per cent of LNG production for the domestic market in 2006, wrote it into each State Agreement before the plants were built, and priced it in. It has not had to intervene since. A 2024 parliamentary review left the offshore policy unchanged. Verified

The east coast did the opposite. It connected its domestic market to export parity at Gladstone with no reservation at all, and has spent the nine years since escalating.

Fig. 5Two tracks
Australian domestic gas interventions since 2006.
2006201720222028Western AustraliaReserved before theplants were built15% reservedno further interventionobligation already metEast coastDid not — and hasintervened four timesGladstone startsexport parity arrivesADGSM$12/GJ price cap+ Heads of Agreementlicence regimeEarly clarity cost less than late intervention.
Source: WA Department of Energy and Economic Diversification; Customs (Prohibited Exports) Amendment (Liquefied Natural Gas) Regulations 2017; Treasurer’s media release, 22 December 2022; ACCC Gas Inquiry; exposure draft Bills, 10 September 2026.

The lesson is not that reservation works or fails. It is that early clarity cost less than late intervention — and that a jurisdiction which sets the terms before the capital is committed does not end up writing licence regimes afterwards. British Columbia is at the point on that curve where Western Australia stood in 2006. Interpretation


The Cost Is Not the Volume

If the scheme diverts seven cargoes, why is the reaction so loud?

Because the volume is not what changed. What changed is that a sovereign now sets an annual number, with ministerial discretion over it, on an asset class financed on twenty-year certainty. Wood Mackenzie, modelling for the Australian industry, put it precisely: “Discretion of this breadth is a cost independent of how it is exercised, because parameters set by annual judgment rather than transparent rules cannot be modelled with confidence and are priced as a risk premium.” Verified

Three features sharpen it. Relief for pre-existing contracts must be requested in the original licence application and determined by 31 December 2027 — before a single regulated period has run and before the regulator’s calibration method exists. Buying third-party gas to comply is penalised rather than rewarded: acquiring gas that is not certified “additional” increases your own obligation by the same amount. And the method for splitting the aggregate between exporters is in neither Bill. Verified

“The proposed scheme may have a retrospective effect by effectively replacing an existing contractual right to export with a requirement to obtain regulatory approval to continue exporting.” — Institute of Energy Economics, Japan, submission to the Australian Government, 30 June 2026

That submission matters more than any analyst note. The IEEJ’s members are Japan’s major energy suppliers and industrial consumers, and Australian LNG is about 40 per cent of Japan’s LNG imports and 10 per cent of its primary energy supply. Its language is not diplomatic: the must-sell obligation “may be perceived as inconsistent with fundamental principles of a market economy”, and the scheme “is likely to discourage additional upstream investment in Australia”. Verified

The response is already visible in the contracting register. Every North Asian long-term SPA we could verify in 2026 was signed with Qatar, the United States or Malaysia. None with an Australian project. Verified

All of this is landing in the tightest Pacific market in years. Qatar shipped 18 cargoes in the six months after the Strait of Hormuz closed, against 509 in the same period a year earlier; two Ras Laffan trains, 17 per cent of Qatari capacity, are damaged with a three-year repair. The IEA now expects global LNG supply broadly flat in 2026, having forecast 7 per cent growth two quarters earlier. Australia, meanwhile, ran at 97.2 per cent of nameplate in August. Verified It is the most reliable large supplier left in the basin, and it has just written itself an option to keep gas at home. Interpretation


Six Days in September

On 9 September, Kevin Gallagher told the National Press Club that Canada was “stepping into the LNG supply gap Australia has allowed to emerge”, and that “every LNG cargo Canada locks in for the future, is a cargo Australia foregoes”. Verified

The exposure draft Bills landed the next day. Five days after that, on 15 September, Santos announced a non-binding heads of agreement with Western LNG to buy about 1 mtpa of Canadian LNG from the Ksi Lisims project, free on board, for up to twenty years from around 2031. Verified

Santos gives four reasons, all commercial: grow LNG sales “without a corresponding increase in upstream capital investment”; supply flexibility for Asian customers; geographic diversification into Canadian LNG with “a secure, relatively short shipping route to premium north Asian markets”; and portfolio optimisation. The release does not mention Australian policy, the reservation scheme, GLNG or the Australian domestic market. Verified Take that at face value — and note that the pattern is older than the policy. Woodside contracted 2.5 mtpa from Commonwealth LNG in Louisiana in September 2022, and took FID on the 16.5 mtpa first phase of its own Louisiana LNG in April 2025. Verified

Fig. 6The pattern predates the policy
Australian majors’ North American LNG positions, against Australian reservation policy milestones.
202220232024202520262027Australian majors,North American LNGWoodside — Commonwealth LNG2.5 Mtpa offtake, US GulfWoodside — Louisiana LNG FID16.5 Mtpa, US GulfSantos — Ksi Lisims1 Mtpa HOA, Canada PacificAustralianreservation policyreservation announcedexposure draft BillsThe pattern predates the policy. What is new is the Pacific coast.
Sources: Woodside/Commonwealth LNG media release, 5 September 2022; Woodside Half-Year Report 2026 (Louisiana LNG FID April 2025); Santos media release, 15 September 2026; Australian Government announcements of 22 December 2025 and 10 September 2026.

What is new is the coast. Santos’s is the first Pacific-facing North American position taken by an Australian exporter, and it lands in the same fortnight as the Bills. Interpretation

One piece of drafting is then worth reading closely. Relief from the domestic obligation on the grounds of a pre-existing export contract requires three cumulative tests, and the second is that the licence holder cannot reasonably produce or acquire additional gas. Four days before the Santos announcement, King & Wood Mallesons published its reading: “if the licence holder could ‘reasonably’ purchase gas (including from international sources) … the second limb would not be satisfied.” Verified

An international supply portfolio is, on that reading, evidence that relief is not needed.

That bites hardest on the ventures which are already net buyers. In 2025 the Queensland exporters sold 181 PJ into the domestic market and bought 228 PJ out of it, from producers other than themselves. Verified Two of the three carry a net domestic purchase requirement on the ACCC’s forward numbers. Verified

The Canadian gas itself cannot help. An obligation is discharged only by physically supplying gas into the Australian market, and only from the holder’s own gas or gas certified as “additional”. A Canadian cargo would have to be shipped to Australia and regasified — and Australia has exactly one import terminal, at Port Kembla, whose owner argued in the Financial Review on 23 September that the availability of regasification should stop pipeline constraints being “an automatic reason to reduce domestic supply obligations”. Verified

So the world’s third-largest LNG exporter has written an obligation that is easier to discharge if you can import LNG, and harder to escape if you can buy it. No one involved has said that is the intention. But it is the machinery they have built. Interpretation


The Same Lever, Closer to Home

Australia is not inventing a power. It is the only one of the three major Western exporters that had to legislate for one.

Fig. 7Three levers, one of them new
Statutory powers over gas exports. ✓ express power in statute · ~ general or implied only · — none.
Domestic-needstest on exportVary licence onregulator’s initiativeRevoke in thepublic interestSet the exportprice by regulationEver used ondomestic groundsAustraliaNew regime. Obligations from 1 Jan 2028.express statutory power✓express statutory power✓express statutory power✓general or implied power only~none—United StatesPresumption favours export. Never rescinded over objection.general or implied power only~general or implied power only~general or implied power only~none—none—Canada0 refusals in 58 applications. Never used.express statutory power✓express statutory power✓express statutory power✓express statutory power✓none—express powergeneral or implied onlynone
Sources: exposure draft Domestic Gas Reservation Bill 2026; Natural Gas Act §3 (15 U.S.C. §717b) and DOE Policy Statement, 83 Fed. Reg. 28,841 (2018); Canadian Energy Regulator Act ss.343–353. Record of use: DOE Order 5404, 2 April 2026; CER export licence application schedule.

The United States has the weakest hand of the three. Section 3 of the Natural Gas Act creates a rebuttable presumption in favour of export; exports to free-trade partners are deemed consistent with the public interest and granted “without modification or delay”. The Department of Energy has vacated nine long-term authorisations and never once over the holder’s objection, and its 2018 policy statement says it “does not foresee a scenario” in which it would. The January 2024 pause touched only pending applications, never a flowing molecule, and was enjoined within six months. Congress repealed the general domestic-use authority in 2015 — which is why the bill introduced in the Senate last December has to re-enact it. Verified

Canada is the other end of the scale. Under the Canadian Energy Regulator Act the Commission “must not issue a licence unless it is satisfied that the quantity of oil or gas to be exported does not exceed the surplus remaining after allowance has been made for the reasonably foreseeable requirements for use in Canada”. Unlike the American test, the applicant carries the burden. Verified

It goes further than that. The Commission may vary a licence on its own initiative and attach any conditions it considers appropriate (s348). It may suspend or revoke a licence if it considers that to be in the public interest (s351). And the Governor in Council may make regulations prescribing “the price at which or the range of prices within which” exported gas must be sold — and may set different prices for different countries (s353). Verified

Ottawa can do by regulation what Canberra spent eighteen months legislating, and more. It never has. Of 58 export and import licence applications to January 2023, 47 were approved and none was refused. Every revocation on record came at the licence holder’s own request. Verified

The reason is straightforward: Canada has never had Australia’s problem. It produced 8,249 PJ of marketable gas in 2025 against 5,302 PJ of domestic demand, and LNG was 1.5 per cent of production. AECO sits structurally below Henry Hub. There is no domestic price grievance to answer, and we found no federal, provincial or party proposal for reservation since LNG Canada started up — the debate runs entirely the other way. Verified

One qualification, and it is the one a Canadian reader should hold onto: Ontario and Quebec import just under half their gas from the United States, and most western Canadian gas reaching them transits US territory. “Canada is long gas” is a national truth with a regional exception, and regional exceptions are how these fights start. Interpretation

The honest objection

The strongest argument against the whole scheme sits in the ACCC’s own data. In the second half of 2025, under the Heads of Agreement, the LNG producers offered 85.9 PJ to domestic buyers and sold 9.2 PJ — while exporting 41.5 PJ as eleven spot cargoes. Verified The gas was offered. Buyers did not take it at those terms. That is a price problem, and a must-sell obligation into a deliberately oversupplied market treats the symptom. Australian Energy Producers, whose members would carry the obligation, size the calibration at “around 50 PJ” and argue producers “should not be forced to sell gas below cost”. Against all that: the three Gladstone ventures control about 84 per cent of east coast proved and probable reserves, and southern supply is collapsing. A market that concentrated does not obviously self-correct. Both halves are true.

Fig. 8Offered, and not bought
Gas offered to Australian domestic buyers under the Heads of Agreement, July–December 2025, against gas sold and gas exported as spot cargoes. PJ.
Offered to domestic buyersUnder the Heads of Agreement85.9 PJ85.9 PJActually sold domesticallyAt the offered terms9.2 PJ9.2 PJExported instead, as spot cargoes11 cargoes41.5 PJ41.5 PJThe gas was offered. Buyers did not take it at those terms. July–December 2025.
Source: ACCC Gas Inquiry 2017–2030, June 2026 interim report, §1.4, published 10 July 2026. The Heads of Agreement ceased on 31 December 2025.

What would prove this wrong

  • The regulator’s calibration method, when it appears in 2027, allocates on a basis that produces materially more than ~170 PJ in 2028 — our reading of s24(5) would be wrong.
  • Parliament amends s24(5) to match the government’s own explainers, lifting the first-year obligation by about 100 PJ.
  • A North Asian buyer signs a new long-term Australian SPA on normal terms before the Bills pass — the sovereign-risk reading weakens considerably.
  • Any Canadian or US federal reservation proposal emerges, which would move the lever from dormant to live and change this analysis entirely.

What This Means for You

If you areWhat the evidence saysThe question to ask
A Canadian LNG developer or producerYour export licence already contains a domestic-needs test and a ministerial approval. It has never bound, because Canada is long gas.Have you read the conditions on your own licence, and do your offtakers know what s348 and s351 say?
A North American gas producerThe US lever is the weakest of the three and the politics have no legislative traction. Exports ran 17.4 Bcf/d in the first half, up 23%.What would have to happen to Henry Hub before the public-interest test stopped being a formality?
A trader or portfolio playerThe diverted volume is noise. The repricing of Australian supply risk is not, and it is already in the 2026 contracting register.What discount are you applying to Australian term supply that you were not applying in 2025?
An Asian buyerRelief for existing contracts must be applied for in the original licence application and determined by 31 December 2027.Has your Australian counterparty told you what it intends to file, and when?
A BC or Alberta policymakerWestern Australia reserved before the plants were built and never intervened again. The east coast did not, and has intervened four times.Is it cheaper to set the domestic terms now, or to legislate them in 2040?

Signposts

IfWhat it tells you
The Bills are amended so the 10% applies to demand, not the residualThe first-year obligation roughly doubles. Watch the drafting, not the press release.
The AER’s calibration instrument lands in 2027 with a per-exporter allocation ruleThat is the number that matters. Nothing before it is more than an estimate, including ours.
A Japanese or Korean utility signs new Australian term supplyThe sovereign-risk premium is smaller than the submissions suggest.
LNG Canada Phase 2 reaches FIDReported as possible in early October. Fourteen more mtpa from a jurisdiction that has never pulled its lever.
Ontario or Quebec faces a winter supply squeezeThe one route by which a Canadian domestic-supply argument becomes real.

Australia’s reservation scheme will not reshape the Pacific LNG balance. Seven cargoes will not be noticed in a market missing most of Qatar. What it does is establish, in a Western democracy with a mature export industry, that the right to export is a licence the state grants annually rather than a contract the state honours. Canada wrote itself that power seven years ago and has never used it. The interesting question is not whether Ottawa would. It is whether anyone buying Canadian LNG has read the statute.

IS

Insight Stream — LNG/LPG & Waterborne Trade Analysis

LNG contract economics, LPG benchmarks and Pacific trade flows, VLGC freight, China PDH and NGL export infrastructure. Independent market intelligence, built only on sources you can check.

Read More →