Snowy 2.0 has been under construction since tunnelling began in earnest in 2021, and five years on, Australians are still waiting. The project’s estimated completion date has moved more than once — the current position, as reflected in Snowy Hydro’s own public communications, is first power sometime in 2028, with full commissioning later still. The cost estimate, which started life at $2 billion, passed $12 billion some time ago. Let’s be careful with that number: Snowy Hydro has not published a single audited final-cost figure, and the Commonwealth’s own Infrastructure and Project Financing Agency has flagged the evolving scope. What is public is that the project is late, it is expensive, and the federal government — which owns Snowy Hydro outright — is carrying all of it.
That is the thing about being a government-owned corporation. The upside is patient capital. The downside is that every cost blowout lands in the public account eventually, and there is no market discipline to concentrate minds the way a capital raise or a credit downgrade might.
What Snowy Hydro actually is #
Most Australians know the Snowy scheme as a postwar engineering feat — 16 major dams, seven power stations, 145 kilometres of tunnels and aqueducts spread across the Kosciuszko National Park ranges. That infrastructure underpins roughly 4,100 MW of generation capacity across the existing scheme, though the dispatchable output at any moment depends heavily on water storage levels, seasonal inflows, and how aggressively the company has been running the machines. Power and energy are different things, and the Snowy scheme illustrates that distinction as well as any asset in the NEM: 4,000-plus MW of rated capacity means very little in a dry year when the reservoirs are low.
Beyond hydro, Snowy Hydro owns the Colongra gas peaker north of Sydney — four open-cycle gas turbines at 667 MW combined — and more recently brought the Kurri Kurri gas plant online in the Hunter Valley. Kurri Kurri was commissioned in 2023 after a protracted procurement and construction saga that drew sustained criticism from independent analysts and consumer advocates, who argued the plant was unnecessary given the pace of battery storage deployment and that its costs would flow through to consumers. The AER’s annual market review data shows Kurri Kurri has run at low capacity factors since commissioning, which was broadly predictable from the project’s design: it is a reliability backstop, not a baseload plant. Whether a backstop at that cost was the right call is a question I’d argue has not been adequately answered in public.
And then there is the retail business. Snowy Hydro owns Red Energy and Lumo Energy, which between them serve hundreds of thousands of residential and small business customers across the NEM. Red and Lumo operate as separate brands but share back-office infrastructure; Red markets itself on renewable credentials whilst Lumo has historically competed on price. Both sit under the same wholesale risk exposure as any vertically integrated gentailer, and both are subject to the Default Market Offer framework — worth understanding if you want context on what drives retail margins. The Default Market Offer and your power bill is a good place to start on that.
Snowy 2.0: what the delays actually mean for the grid #
AEMO’s 2024 Integrated System Plan — the ISP, which is the document I’d point anyone to first — modelled Snowy 2.0 as a significant piece of the Step Change and Progressive Change scenarios. The ISP treats the project’s 2,000 MW of pumped-hydro capacity and roughly 350,000 MWh of storage energy as available by the late 2020s. That is a big number: 350,000 MWh would be, by a considerable margin, the largest long-duration storage asset in the NEM.
But the ISP was built on an assumed timeline that has since slipped. If first power is 2028 and full commissioning runs to 2029 or beyond, projects that were sequenced around Snowy 2.0’s firming capacity will need other solutions in the interim. The transmission build — the HumeLink and EnergyConnect lines that partly justify the project — is running on its own complicated timeline. I’ve written before about how pumped hydro and batteries are not perfect substitutes, and the pumped hydro vs big batteries piece covers the technical tradeoffs in detail. The short version is that Snowy 2.0’s long-duration capability — days of storage, not hours — is genuinely hard to replicate with lithium-ion at any foreseeable cost trajectory. The delays matter, but they do not make the project wrong in principle.
What they do is create a window. Battery developers like Akaysha Energy are moving fast into exactly the gap Snowy 2.0 was meant to fill. The Akaysha Energy profile gives a sense of how quickly four-hour BESS capacity is being deployed. Four hours and 350,000 MWh of pumped hydro are solving different problems, but the gap between now and Snowy 2.0’s commissioning is real, and private capital is reading it.
The tunnel problem #
The specifics of what went wrong underground deserve more than a paragraph, but the short version is this. The main access and transfer tunnels involve difficult geology in the Snowy Mountains — mixed rock conditions that required changes to the tunnel-boring machine specification mid-contract. The Italian contractor Webuild (formerly Salini Impregilo) and its joint-venture partner Clough ran into ground conditions harder than pre-contract investigations had indicated. Clough went into voluntary administration in December 2022. Webuild continued under restructured arrangements. Progress has resumed, but the schedule damage was done well before any recovery could be banked.
Snowy Hydro’s public reporting on progress is, to put it gently, optimistic in tone. The federal government has not commissioned a public independent review of the cost overruns in the way that, say, Infrastructure Victoria does with major state projects. That is a gap in accountability that I think is worth naming plainly.
Kurri Kurri: a plant looking for a role #
The Kurri Kurri gas plant sits on the old Alcan aluminium smelter site in the Hunter Valley — a site with existing grid connection and a history of heavy industry that made it a logical choice logistically, even if the policy logic around the plant itself has always been contested. The federal government directed Snowy Hydro to build Kurri Kurri as a reliability measure following concerns about gas supply adequacy and the planned closure of AGL’s Liddell coal station, which went offline in April 2023.
The argument for the plant was always about reliability, not economics. The argument against it — made by the Australian Energy Market Operator in early consultations and by consumer groups — was that the Capacity Investment Scheme and accelerating battery deployment would provide adequate firming at lower long-run cost. The CIS is worth understanding here; the Capacity Investment Scheme piece covers the structure and the tension between picking technology winners versus technology-neutral support.
Kurri Kurri is now a sunk cost. Whether it runs 5% of the year or 15% of the year, the capital is spent. The policy debate has moved on, but it is a useful data point when assessing how Snowy Hydro’s government ownership shapes its decision-making. Commercial operators rarely build generation assets on a minister’s instruction. Snowy Hydro did, and will carry the asset on its books regardless.
Red Energy, Lumo, and the retail squeeze #
The retail arms matter to Snowy Hydro’s financial position more than the company tends to emphasise publicly. Red Energy and Lumo together give Snowy Hydro a hedge against wholesale price volatility — when spot prices spike, the generation fleet benefits; when they crash, the retail load absorbs cheap energy. That vertical integration logic is the same one Origin Energy and AGL run, as the AGL coal exit profile explains.
But the retail market is competitive and the Default Market Offer cap constrains upside. The AER’s retail market reports show customer churn remains high across the NEM; brand loyalty in electricity retail is thin. Red’s renewable positioning gives it some differentiation, but in a market where rooftop solar penetration keeps climbing — Australia’s rooftop solar trajectory is worth a look for context — the volume of electricity flowing through retail meters is structurally under pressure anyway.
Where this leaves the company #
Snowy Hydro occupies a peculiar position in the NEM. It is the largest government-owned generator, it controls the most consequential storage project under construction in Australia, and it runs retail brands that compete directly with private gentailers. It does not publish ASX disclosures. Its annual reports go to the responsible minister and then, eventually, to the parliament. The scrutiny that comes with quarterly earnings calls and analyst briefings does not apply.
I’ve been watching Snowy 2.0’s timeline shift for a few years now. The honest read, from where I sit, is that the project will eventually deliver something the NEM genuinely needs — long-duration, dispatchable, low-emissions firming capacity. But the journey has been more expensive and slower than the original business case, and the public has not been given the tools to hold anyone fully accountable for the gap. Governments that own energy assets and want public trust have to show their working. On Snowy 2.0, that is still a work in progress.
The tunnelling continues. The target date is 2028. I’ll believe it when AEMO’s Integrated System Plan stops treating it as a future assumption and starts treating it as a commissioned asset. And if you want a benchmark for how the AER’s retail performance data tracks Red and Lumo against their competitors, that dataset is updated annually and is worth bookmarking.
— Anjali Rao, Grid & Storage Correspondent
Photo by Jaël Vallée on Unsplash