Beetaloo Energy targets Q4 2026 first gas on AI data centre opportunity (Darwin LNG, Australia).
Australia's Beetaloo Basin has spent the better part of two decades as a "someday" story—enormous resource numbers, slow appraisal, and a market that struggled to price molecules that weren't yet flowing. That is starting to change. Beetaloo Energy Australia Limited, formerly Empire Energy Group Limited, is now weeks away from its Carpentaria pilot delivering first gas, and has layered a second growth angle onto the story: a memorandum of understanding (MOU) with US oilfield services major Halliburton to help power a proposed AI data centre near Darwin. Managing Director Alex Underwood set out both threads in a recent interview, alongside an update on financing, well economics, and the broader competitive backdrop for Australian gas.
Beetaloo holds two large acreage positions either side of the Beetaloo Basin in the Northern Territory. The Carpentaria project (EP187), on the eastern side, covers around 110,000 continuous acres and is the near-term production asset; a much larger position on the western side spans roughly 1.2 million continuous acres. Independently assessed prospective resource across the portfolio stands at approximately 47 trillion cubic feet (Tcf) of gas equivalent, with around 1.7 Tcf converted to 2C contingent resource.
Three wells are connected to the pilot pad roughly 5km from the Carpentaria gas plant, with the plant construction and flowline now largely complete. Underwood confirmed the company is targeting first gas in the fourth quarter of 2026—later than the 2025 timeline the company had previously flagged, reflecting the practical realities of construction and commissioning rather than any change in resource or contract position.
Underwood was candid about the cost side of the ledger. The company's most recent well cost north of A$50 million to drill and hydraulically fracture—a high figure by US shale standards, driven largely by the cost of transporting equipment and materials across Australia. Sand alone accounted for A$6 million on the last well, of which A$5 million was trucking. Using a rule-of-thumb gas price of A$10–12 per gigajoule and an estimated 10 petajoules of recoverable gas per well, Underwood put gross revenue potential at around A$100 million per well over its life, generating modest but net-present-value-positive returns during the pilot phase.
The bigger opportunity, he said, lies in cost reduction as drilling and stimulation move to a year-round cadence—a pattern already seen in the Queensland coal seam gas industry, where current wells cost around 20% of the earliest ones. Beetaloo's own modelling suggests well costs could fall by at least half under year-round operations, at which point Underwood expects returns of 30–50% IRR (internal rate of return).
Beetaloo has backing from Macquarie Bank, including a midstream infrastructure facility funding the Carpentaria gas plant build, and raised approximately A$70 million in equity earlier this year. The company recently reported total available liquidity of around A$125 million, split roughly evenly between cash and undrawn facilities—which Underwood described as the strongest financial position in the company's history heading into first gas.
A binding 10-year gas sales agreement with the Northern Territory Government underpins near-term revenue, priced on a fixed basis with a CPI (consumer price index) escalator. Beetaloo expects to supply 10 terajoules per day (TJ/d) into the McArthur River mine pipeline this year, rising to a further 15 TJ/d next year once the Northern Territory Government funds a reversal of pipeline flow to serve the local market.
The most significant new development is a non-binding MOU signed with Halliburton, focused on the power generation component of a proposed data centre project—Beetaloo Digital—on government-granted land at Weddell, near Darwin's industrial precinct. [Editor's note: land size stated as 85 hectares in the interview audio but 185 hectares per the company's press release—flagged for verification.]
The company is assembling a consortium spanning gas supply, solar generation, pipeline construction and power generation, with Halliburton contributing expertise from its own behind-the-meter power generation work—including a recent tour of the Stargate data centre project in the US. Underwood noted a rule of thumb of around 200 TJ/d of gas required per gigawatt of installed data centre capacity, translating to roughly $1 billion of EBITDA (earnings before interest, tax, depreciation and amortisation) a year per gigawatt for gas suppliers into that market. Beetaloo's role, he stressed, is confined to gas supply rather than data centre ownership or downstream power revenue.
Beyond the AI opportunity, Underwood pointed to structural demand from Australia's east coast, where declining Bass Strait output and rising LNG (liquefied natural gas) export needs are expected to draw on Beetaloo Basin gas. Neighbour Santos has publicly flagged the Beetaloo as a meaningful contributor to its Gladstone LNG project by the early 2030s, while APA Group is reportedly planning a pipeline capable of moving 1,000 TJ/d or more east. Darwin itself already exports LNG equal to roughly 11% of Japan's gas needs, with expansion capacity at both existing terminals.
Underwood also addressed the renewables question directly, noting that hydrocarbons have held steady at around 91% of Australia's total primary energy mix over the past 20 years despite the renewables build-out—underscoring, in his view, gas's durability as the backbone of the system.
The Beetaloo story now sits at the intersection of two macro themes reshaping global energy demand: a structural east-coast Australian gas shortfall, and the rapidly escalating power requirements of AI infrastructure. That dynamic, already visible in the US through behind-the-meter projects and reopened nuclear capacity, has arrived in Australia, where the federal government now expects large data centres to source their own power rather than draw on a constrained grid. For gas developers with the resource base and location to supply that demand, the AI buildout represents a potential second market layered on top of traditional domestic and LNG export demand—though, as with any consortium-dependent proposal, the economics remain conceptual until financing and offtake are locked down.