32 gigawatts. That’s the target capacity the Commonwealth’s Capacity Investment Scheme is meant to underwrite by 2030, split roughly across dispatchable storage and renewables, and the tender rounds run so far have already allocated a meaningful chunk of it. The scheme was sold as a neutral backstop, a way of smoothing revenue risk so projects get built without the taxpayer taking the full downside. Several tender rounds in, the pattern of who actually wins starts to look less like a neutral market mechanism and more like industrial policy wearing an auction’s clothing.
We asked the broad version of that question back in July, in whether the scheme is quietly picking winners. This piece is narrower. It looks at the tender clock itself: how the deadlines, bankability tests and regional volumes decide what can even enter a round, and what that timetable leaves on the outside.
That’s not necessarily a scandal. Industrial policy is a legitimate thing for a government to run.
But it’s worth calling it what it is, instead of pretending the CIS is simply letting the market sort itself out.
What the Capacity Investment Scheme does
We’ve covered the mechanics before in how the Capacity Investment Scheme actually works, so I won’t re-run the whole explainer. The short version: the Commonwealth, with AEMO‘s tender arm AEMO Services running the competitive rounds, tenders for capacity region by region, and successful bidders get a revenue floor and a ceiling. If wholesale and ancillary market revenue falls below the floor, the government tops it up. If it rises above the ceiling, the generator pays the difference back. It’s designed to de-risk investment without the government owning the asset or dictating technology.
In theory it’s technology neutral. In practice the tender design and the capacity targets set region by region have already decided quite a lot before a single bid is assessed. Canberra, through the Department of Climate Change, Energy, the Environment and Water, sets the volumes and the dispatchable/generation split for each round. That’s a policy choice dressed as a market parameter.
The battery tilt
Look at where the contracted capacity has actually landed. Big four-hour batteries have taken a disproportionate share of the dispatchable allocation in most rounds, well ahead of pumped hydro, long-duration storage, or gas peaking plant. Partly that’s because batteries are quick to build and the CIS rounds reward projects that can hit financial close and construction milestones inside a tight window. Partly it’s because the economics genuinely favour four-hour batteries right now, a point we’ve made in gas peakers versus big batteries in the 2026 firming maths.
Well-capitalised storage developers such as Akaysha Energy, profiled in our look at Akaysha Energy’s project book, and Neoen, whose shift from the original Hornsdale battery into a broader storage and wind portfolio we tracked in our profile of Neoen’s Australian build, are exactly the kind of bidder this design suits. There’s nothing wrong with strong developers winning on merit. But when the tender settings keep favouring the same profile of project, it tells you something about how the bar is set, not just about who’s good at clearing it.
What gets shut out
Pumped hydro barely features in CIS contracting to date, for the obvious reason that an eight-year construction timeline doesn’t fit a scheme built around near-term capacity targets. Snowy 2.0’s long saga, which we’ve followed in will Snowy 2.0 ever earn its keep, and Queensland’s Borumba project, covered in Borumba pumped hydro’s cost and timeline, sit mostly outside the CIS funding envelope and rely on government balance sheets instead: the Commonwealth’s for Snowy, Queensland’s for Borumba. That’s a structural exclusion, not a merit judgement. A technology that takes a decade to build is never going to compete on the same clock as a battery that takes eighteen months, no matter how useful it would be for winter firming in 2032.
Gas peakers have done a little better, picking up some contracts where state governments pushed hard for dispatchable firming alongside renewables, but the volumes are modest next to batteries. Long-duration storage technologies that aren’t lithium-ion, think iron-air, flow batteries, thermal storage, have barely registered. If you’re a developer with a genuinely novel storage chemistry, the CIS tender timeline and bankability requirements are close to a wall.
Regional splits as a policy lever
The other place the scheme shapes outcomes is the regional allocation. New South Wales has had the largest dispatchable target in several rounds, which tracks with the state’s accelerating coal exit, a mismatch we’ve written about at length in NSW’s coal exit outrunning its REZ build. That’s defensible. New South Wales does need the capacity, and soon.
Queensland and Victoria get smaller dispatchable allocations relative to their coal exit timelines, partly because both states run their own capacity mechanisms alongside the federal scheme, Queensland through its public generation build-out and Victoria through the reconstituted SEC, a point covered in Victoria’s SEC revival and its offshore wind test. South Australia, already running on a very high renewable share, has needed less. Western Australia is handled through separate tenders for its own Wholesale Electricity Market, because the SWIS is a disconnected grid with its own rules, a point worth remembering whenever someone quotes a single national CIS figure as if every region were the same market; they aren’t, as we set out in WA’s SWIS grid: an island with no way to call for help.
None of that is secret. AEMO Services publishes the tender round results and DCCEEW sets out the methodology behind regional targets on its own website. But “published” and “widely understood” are different things, and I’d wager most people who have an opinion on the CIS haven’t read a single tender outcome document.
Why construction-readiness beats novelty
Here’s the part that deserves more scrutiny than it’s had. A scheme built around competitive tenders should, in theory, produce turnover among winners as new entrants undercut incumbents on price. Several rounds in, that’s not obviously what’s happening. Developers with strong balance sheets, existing grid connection agreements, and in-house engineering teams look structurally favoured, while smaller or first-time bidders struggle to get projects to the stage where they’re bankable enough to tender competitively in the first place.
That’s not a conspiracy. It’s what happens when a tender rewards construction-readiness above almost everything else. A developer with land secured, connection studies done, and a balance sheet that can absorb delay without the project collapsing has a structural advantage that has nothing to do with the quality of the underlying technology. Follow where the risk actually sits in these contracts and you find the government has effectively chosen to underwrite scale and experience over novelty.
Genex Power’s experience is instructive here. Its Kidston project, now under J-Power’s ownership and covered in our profile of the Kidston bet under J-Power, took years longer than planned to reach financial close, partly because smaller listed developers carry more balance sheet risk than a Neoen or an Akaysha backed by larger institutional capital. The CIS tender process, whatever its intentions, rewards exactly the kind of corporate heft that smaller developers don’t have.
Is that actually a problem?
I’ll be honest: I’m not convinced picking winners is automatically bad policy here. Australia needs capacity built fast, and a scheme that rewards construction-readiness over novelty is at least rational given the timeline pressure from coal closures, a pace mismatch we’ve tracked in coal exits versus replacement builds. If the choice is between a perfectly neutral market mechanism that takes longer to produce capacity and a slightly tilted one that gets batteries built on schedule, I’d take the tilted one most days.
Where I’d push back is on the public framing. The scheme is routinely described, including in some government material, as technology-neutral and market-based. It’s neither, not entirely. It’s a government department choosing targets, splits and timelines that favour certain technologies and certain balance sheets, then running a competitive process within those constraints. Calling that a market mechanism undersells the amount of discretion built into the design, and it makes it harder for the public to ask the obvious follow-up question: are those discretionary choices actually the right ones?
Better public data would help. A single round-by-round breakdown of contracted technology, region and developer, published with the same rigour AEMO applies to dispatch data, would make the scheme far easier to scrutinise. At the moment, you have to piece the picture together from ASX announcements and state government press releases, which is a strange way to track a multi-billion dollar public underwriting scheme.
What a longer clock would change
None of this suggests anything improper in how the CIS is run. It means it’s a policy instrument doing what policy instruments do: encoding choices about what gets built, where, and by whom, then presenting those choices as the neutral output of competition. That’s worth saying plainly, because the alternative, treating every CIS round as pure market discovery, lets both government and industry avoid the harder conversation about whether the current tilt toward quick, lithium-ion, large-balance-sheet projects is actually the best use of taxpayer-backed revenue support over the next five years.
I’ll leave the technical umpiring to AEMO. But like a Test selector picking the same middle order every series because they know how to graft a score rather than because they’re clearly the best six players in the country, the CIS keeps backing what it already knows works. Fair enough, up to a point. Just don’t call it an open contest. A separate, slower track for long-build technologies such as pumped hydro and long-duration storage, with its own volumes and milestones, would let the scheme back speed and diversity at the same time, instead of asking a decade-long project to race an eighteen-month battery on the same timetable.
– Marcus Wren, Editor
Photo by Patti Black on Unsplash