Australia's Big Battery Build Outruns the Duck Curve, Reshaping NEM Price Signals
Australia's 2GW battery build in 2025 makes it the world's third-largest storage market, changing how gas peakers and data centres set NEM prices.
Australia added a record 2 gigawatts of utility-scale battery capacity during 2025, a 233 percent jump on the prior year that pushed the country past every market except China and the United States, according to the Clean Energy Council's latest tally.1 The build-out took Australia to third place globally on installed big battery capacity, a ranking that would have seemed implausible when the first grid-scale lithium projects were struggling through commissioning just a few years ago.1
The National Electricity Market is being repriced from the bottom up. AEMO now counts 2.8 GW of behind-the-meter batteries responding to price signals without central dispatch, equivalent to the capacity of Eraring, the country's largest coal plant.2 The new fleet absorbs solar oversupply and competes directly with gas peakers for the evening ramp, and increasingly displaces thermal plant altogether rather than simply supplementing it.1
The consequence is visible in the market's volatility profile. Intra-day price swings are compressing in some periods while inter-day and event-driven volatility becomes more important, a shift that reflects the storage fleet responding faster than any thermal unit can.2 Gas-fired generation is losing its role as the default flexible supplier, and the economics of building new gas peakers are being questioned at the same time that data centre demand is rising.4
Data centres are the wildcard. Average NEM demand from data centres through the first quarter of this year was nearly 600 MW, and 11 facilities totalling 5.4 GW of ultimate load were working through transmission connection agreements.3 The national science agency's latest GenCost report says batteries have become Australia's preferred technology for flexible generation precisely because data centre growth is driving up the costs of gas-fired plant.4
The project pipeline shows how far the storage build has come. AGL's 500 MW Liddell Battery, with one gigawatt-hour of capacity, is in the tally, with its first 250 MW stage in commissioning.1 Equis and Victoria's State Energy Corporation brought the 600 MW, 1600 MWh first stage of the Melbourne Renewable Energy Hub online.1 Akaysha's Ulinda Park battery in Queensland started trading in December with a 55 MW first phase.1
The committed pipeline is larger still. Another 4.3 GW and 13.5 GWh of big battery capacity was financially committed during 2025, representing $4.8 billion of investment, a 67 percent increase on the prior year's level.1 At that rate of commitment, the operational fleet could roughly double again within two years, which would put further downward pressure on peak prices in the evening window.
The evening remains the system's biggest challenge. Solar generation peaks around midday and collapses in the late afternoon, forcing the grid to find fast-ramping capacity just as household demand climbs.6 The duck curve is not new, but the storage response is. Batteries are now absorbing the midday surplus and discharging into the evening ramp, which changes the revenue calculus for any new peaker considering a merchant position in the NEM.
Government projections still flag supply risks in later years as ageing coal plants exit potentially faster than new renewables and storage come online, and AEMO continues to warn about the pace of the transition.5 But the battery build is ahead of most expectations, and the agency's own statements of opportunities have shifted to weigh demand-side response more heavily than in earlier editions.3
The next signal to watch is how the 4.3 GW of committed capacity lands. If the financial commitments convert to operational projects on schedule, the NEM's evening price spikes will keep compressing, and the case for new gas peakers weakens further. If commissioning slips, the window reopens for thermal plant to earn scarcity rents, and the investment thesis for storage loses some of its momentum.1,2