32 gigawatts. That’s the number the Commonwealth put on the table when it expanded the Capacity Investment Scheme, split roughly between 23GW of variable renewables and 9GW of dispatchable capacity, all meant to land by 2030. It is, by a distance, the largest single intervention in the National Electricity Market since privatisation, and it doesn’t look anything like the market-based competitive-tender language the government uses to describe it. The honest read is simpler: Canberra has become the counterparty of last resort for new generation and storage, because the merchant market stopped pricing that risk properly on its own.
I’ve spent a fair chunk of the past year going back through AEMO Services’ tender documentation and the Clean Energy Regulator’s project data trying to work out where the CIS money is actually landing, as distinct from where the press releases say it’s landing. The gap between the two is the story.
What the Capacity Investment Scheme actually promises #
The scheme, run by AEMO Services on behalf of the Department of Climate Change, Energy, the Environment and Water, works through periodic tenders in each region of the NEM. Developers bid a strike price. Winners get a contract, not a subsidy in the traditional sense, but a revenue underwrite that sits alongside their normal participation in the wholesale market.
That’s the bit people get wrong. Projects under CIS still sell their output into the NEM through AEMO’s five-minute dispatch process, same as everyone else — the mechanics of that dispatch cycle haven’t changed. The contract only bites at settlement. If a project’s actual market revenue falls below an agreed floor, the government tops it up. If revenue runs above an agreed cap, the project pays the difference back. It’s a two-sided contract for difference, not a one-way handout, and that distinction matters when critics call it a blank cheque. It isn’t. But it isn’t nothing either.
The mechanism: floors, caps, and who wears the risk #
Here’s where it gets interesting. The floor price protects developers and their financiers against a repeat of what happened through 2023 and 2024, when wholesale prices went from unpredictably high to unpredictably low within the same calendar year, and project finance for anything without a hedge became close to unbankable. The cap protects taxpayers from paying developers twice, once through the contract and again through a runaway wholesale price during a supply crunch. On paper it’s an elegant piece of risk-sharing.
In practice, the floor is where the argument is. Set it too low and nobody bids. Set it too high and you’ve built a subsidy scheme wearing a tender’s clothing, with the government absorbing merchant risk that private capital used to be paid to hold. Follow the money on this one and you find the answer isn’t ideological, it’s mechanical: private capital wasn’t holding that risk any more anyway, not at a price anyone could build a project around. The CIS didn’t crowd out a functioning market for merchant renewables. It replaced one that had already stopped clearing.
Follow the money: who’s actually signed up #
Tender rounds have now run across most NEM regions — NSW, Victoria, the combined southern states, and Queensland’s own allocation running in parallel with the state’s own pumped hydro and public-ownership push, which I covered separately in the piece on Queensland’s energy plan. The generation-stream results have been overwhelmingly solar and wind, which is unsurprising given where the cost curve sits. The dispatchable stream is where it gets more revealing.
Lithium-ion batteries have dominated the dispatchable allocations by a wide margin. Developers like Akaysha Energy, backed by BlackRock’s infrastructure arm and profiled in our piece on Akaysha’s build-out across the NEM, have been repeat winners across multiple rounds. Pumped hydro has picked up far less of the dispatchable allocation than its proponents would like, partly a timing problem — projects like Kidston, which I wrote about in Kidston pumped hydro: the mine that became a battery, take years longer to build than a battery does, and the CIS timetable rewards speed as much as durability. Whether that’s the right trade-off for a grid that will eventually need long-duration storage, not just four-hour batteries, is a genuinely open question I don’t think the scheme has answered yet.
The picking-winners problem #
We’ve argued before, in the piece Is the Capacity Investment Scheme quietly picking winners?, that the tender design has structural biases baked in, favouring technologies that can bid confidently on cost and timeline over ones that are strategically valuable but slower and lumpier to finance. That argument holds up better a year on, not worse. A scheme that runs on rolling regional tenders with fixed capacity targets per round will always reward whoever can turn up with a bankable, shovel-ready bid. That’s mostly batteries and mostly solar. It is not, so far, pumped hydro, and it is not pointing much money at grid-forming capability either, a gap I think matters more than the scheme’s architects have acknowledged — the case for that technology is laid out in our explainer on grid-forming inverters and why the grid suddenly needs them.
Mildly contrarian view, and I’ll own it: I don’t think the CIS is broken, but I do think the consensus that it’s a neutral, technology-agnostic tender is wrong. It’s technology-agnostic on paper and battery-heavy in practice, because batteries are what the finance market can price fastest right now. That’s not a scandal. It’s just not what the marketing implies.
Batteries, coal exits and the timing problem #
The scheme exists because the coal exit clock and the replacement-capacity clock aren’t running at the same speed, a mismatch we’ve tracked closely, including in Is Australia closing coal faster than it can replace it? Eraring’s extended life, discussed at length in our piece on how 2025 became 2027 for that station, is itself a symptom of the same underlying gap the CIS is trying to close from the other direction: build enough firmed capacity fast enough that ageing coal units can retire on a schedule that doesn’t require last-minute extensions negotiated under duress.
Whether batteries alone can do that job is a separate question, and one we’ve been sceptical about. Four-hour batteries solve an evening peak problem well. They don’t obviously solve a multi-day wind lull, and the debate over whether pumped hydro or big batteries actually firm the grid is not settled just because batteries have won most of the CIS contracts so far. Winning a tender and solving the problem the tender was designed for aren’t automatically the same thing.
Where the states actually fit #
The Commonwealth scheme runs alongside, and sometimes in tension with, state-based targets and schemes — NSW’s own renewable energy zone rollout, which I’ve written is running behind its own timetable in NSW’s energy transition, Victoria’s storage targets, and the varying ambition levels tracked in our scorecard of state renewable energy targets. AEMO’s own Integrated System Plan is meant to be the coordinating document that ties all of this together, but a national capacity scheme layered over eight or so competing state programmes is inherently messy, and nobody involved pretends otherwise off the record.
The honest read #
The Capacity Investment Scheme isn’t a subsidy in the pejorative sense some of its critics use, and it isn’t the free-market tender its architects sometimes claim either. It’s the Commonwealth doing what governments have always eventually done when private capital gets nervous about a public good: becoming the buyer of risk itself, formally, through a documented contract instead of an implicit bailout later. The Department’s own scheme page is fairly upfront about that framing if you read past the summary.
What I’d watch over the next couple of tender rounds isn’t the headline gigawatt tally, which will keep climbing and keep generating good news stories. It’s the split between batteries and everything with more than four hours of duration. If that split doesn’t move, the CIS will have solved the easy half of the firming problem and quietly left the hard half — long-duration storage, grid-forming capability, the stuff that doesn’t bid well into a fast tender — for somebody else’s budget, later. Rather like declaring victory at tea when the tail’s still got to bat.
— Marcus Wren, Editor
Photo by Artem Kniaz on Unsplash