Mary Valley Road out of Gympie still runs past dairy country before it climbs toward the Borumba Dam turn-off, and if you didn’t know what was planned up there you’d drive straight past it. That’s the thing about pumped hydro sites generally: the interesting part is underground, or in this case, inside a mountain. I went up for a look earlier this year when Queensland Hydro was running site tours, and the scale of the excavation already underway is the first thing that tells you this isn’t a project anyone can quietly shelve.
Borumba pumped hydro is a 2GW, roughly 24-hour-duration pumped storage scheme being built by Queensland Hydro, the state-owned entity set up under the Palaszczuk and now Crisafulli governments to deliver the storage leg of Queensland’s energy plan. It sits on the existing Borumba Dam, about 160km north of Brisbane near Gympie, and it’s meant to pair with the wind and solar going into the state’s renewable energy zones. The concept is straightforward enough to sketch on a napkin: build an upper reservoir, connect it to the existing dam via tunnels and an underground power station, and let water run downhill through turbines when the grid needs megawatts, then pump it back up when there’s cheap solar to burn. The execution is where it gets expensive and slow, and that’s really the whole story of this project at this point.
The cost has moved, and moved a lot #
Let’s be careful with that number, because Borumba’s budget has shifted more than once and different documents quote different scopes. The original 2022 announcement under the previous Labor government put the project at around $14.2 billion. By the time Queensland Hydro’s own project updates and state budget papers caught up with ground conditions, contractor pricing and a redesign of the underground works, that figure had moved well past $18 billion, and the state government under David Crisafulli has been candid in public statements that further cost pressure is likely. I’d treat any single number here as a snapshot rather than a destination. Pumped hydro projects of this size (tunnel-heavy, underground powerhouse, remote site) tend to re-price every time soil testing turns up something the original geotechnical survey didn’t quite capture.
What’s worth separating out, because gentailers and critics alike blur it constantly, is power versus energy. Borumba’s 2GW figure is a capacity number. It tells you the maximum rate at which the plant can push electricity into the grid at a single instant. The roughly 24-hour duration figure is an energy number. It tells you how long the plant can sustain that output before the upper reservoir runs dry. Multiply the two and you get something in the order of 48,000 MWh of total storage, which is an enormous number next to the battery fleet AEMO currently tracks in its quarterly generation information data. Most grid-scale batteries built in Australia over the last three years sit at one to four hours of duration. Borumba is being built to do a fundamentally different job: multi-day firming through a run of cloudy, windless days, not smoothing a single evening peak.
Where construction actually sits #
Early works have been underway for a while now: access roads, the main works camp, and tunnelling for the underground power station cavern. Queensland Hydro’s own progress updates describe excavation of the headrace and tailrace tunnels as the critical path item, which makes sense given this is the part of the project with the least room to accelerate. You can throw more contractors at a site office. You can’t easily throw more diggers into a single tunnel face without them getting in each other’s way.
The commissioning timeline has slipped from the original target. Current public guidance points toward first power somewhere in the early 2030s rather than the mid-2020s date floated at announcement, and I wouldn’t be surprised if that moves again before concrete is poured for the main powerhouse. None of that makes Borumba unusual among big pumped hydro builds – Snowy 2.0 has had its own well-documented timeline slippage, discussed at length in our look at whether Snowy 2.0 will ever earn its keep, but it does mean anyone pencilling Borumba into supply-demand balance models for 2028 or 2029 is working off a date that Queensland Hydro itself doesn’t claim anymore.
Why Queensland needs it specifically #
Queensland’s grid has a coal-heavy generation mix retiring on a schedule that doesn’t wait for storage to catch up. We’ve covered the mechanics of that mismatch in NSW’s coal exit and REZ rollout gap, and the same tension applies north of the border, just with a state government that’s chosen to build the storage itself rather than rely entirely on the market. That’s the real significance of Borumba: it’s one of the two load-bearing pillars of Queensland’s own energy plan, alongside the transmission build connecting renewable zones to demand centres. We went through that framing in detail in Queensland’s energy plan rests on two pillars and one cliff face, and Borumba is, bluntly, the pillar most exposed to construction risk.
Pumped hydro’s advantage over batteries isn’t speed to market – batteries win that comparison comfortably, as we’ve set out in our firming comparison between gas peakers and big batteries. Pumped hydro’s advantage is duration at a cost per megawatt-hour of storage that gets cheaper the longer you can run it, because you’re not paying for more battery cells, you’re paying for more water in an already-built reservoir. Borumba’s roughly 24-hour duration is the point. Queensland doesn’t need another two-hour battery; AEMO’s fleet data already shows the NEM has plenty of short-duration storage and comparatively little of the long kind.
The comparison nobody quite likes making #
Borumba invites an obvious comparison to Genex Power’s Kidston project further north, another pumped hydro scheme built into a disused gold mine, which we tracked in Kidston’s slow walk to commissioning. Kidston is smaller – 250MW against Borumba’s 2GW, and privately developed rather than state-owned, which makes it a useful stress test of the public-versus-private ownership question we’ve examined separately in public versus private ownership of generation. Kidston has had its own schedule slips and financing complications along the way, and it’s materially smaller in scope. If a 250MW project with private capital discipline behind it still ran into delay, a 2GW state-owned project with a bigger tunnel programme was never going to be a straight-line build. I don’t think that’s a knock on Queensland Hydro specifically. It’s just what this class of infrastructure does.
A genuinely contrarian view on the cost overruns
Here’s my mildly unpopular take: the cost blowout gets treated in press coverage as a scandal, when for a project of this engineering complexity it’s closer to the base rate. Civil works of this scale – deep rock excavation, underground powerhouse construction, environmental approvals near a working dam catchment – routinely reprice by 30 to 50 per cent between concept and construction, even for well-run public utilities overseas. What would actually worry me is a cost estimate that hadn’t moved at all four years after a $14 billion headline figure was first floated. That would suggest nobody had done enough ground-truthing to find the expensive surprises yet.
What it means for the market around it #
If and when Borumba commissions, its real market role will be arbitrage and reliability at scale: buying cheap, abundant daytime solar and selling it back during the evening ramp and through multi-day lulls, the exact gap that gas peakers currently fill, as we’ve described in what firming actually means. Queensland’s gas peaker fleet isn’t going anywhere in the near term regardless of Borumba’s progress, because 2GW of pumped hydro commissioning in the early 2030s doesn’t retire existing peaking capacity on day one. What it should do, eventually, is reduce how often those peakers run flat out during stressed periods, which matters for both wholesale price volatility and the state’s emissions trajectory under the Safeguard Mechanism framework tracked by the Clean Energy Regulator.
The project’s financing structure also matters more than it gets credit for. Because Borumba sits on the Queensland government’s own balance sheet rather than being merchant-financed, it doesn’t face the same revenue-certainty pressure that’s shaped how private battery and pumped hydro developers have approached the Capacity Investment Scheme tender process. That’s a genuine point of difference worth noting next time someone compares Borumba’s economics directly to a privately financed battery. They’re not raising capital against the same risk profile at all.
I’ll admit I went up to Borumba half-expecting a stalled site with a sign and not much else, the way some of these announcements age. What’s actually there (tunnel boring progress, a genuine workforce, visible dirt moving) tells a different story to the headline cost figures alone. The project is real, it’s underway, and it’s also years later and billions of dollars more expensive than the first press release suggested. Both things are true at once, and anyone writing about Queensland’s energy transition needs to hold both in their head rather than picking the version that suits the argument they already wanted to make.
– Anjali Rao, Grid & Storage Correspondent
Photo by American Public Power Association on Unsplash