Australia's battery boom is crushing the returns that justified it
Storage is winning the evening peak and setting prices more often, but arbitrage revenue is collapsing even as renewables set records.
BNEF's third-quarter 2026 Australia Power Market Quarterly report, published on 20 August 2026, showed that utility-scale batteries supplied an average 8.4% of power demand at 6pm across Australia's National Electricity Market in the second quarter of 2026, up from 3.1% a year earlier, while gas fell to 6.8% from 12.6% over the same period. Batteries also set wholesale prices in 39% of intervals, roughly double the 19% share a year earlier. Storage is no longer a bit player in the evening peak.6
The returns tell a different story. Average intraday arbitrage across the NEM fell to $103 per MWh in the second quarter, a 79% year-on-year decline, and BNEF put utility-scale battery arbitrage revenue at $53m for the quarter, down 57%. A separate figure in the same report shows arbitrage per unit falling 84% to $60/MWh. More batteries are chasing a shrinking spread.6
Wind and solar combined supplied 35% of NEM generation in the second quarter, up from 31% a year earlier, and renewables passed 50% of NEM electricity over a full quarter for the first time in late 2025. But average power prices fell between 40% and 65% year on year across the market in the same quarter. Falling prices are the mechanism by which renewables win share, and they are also what erodes the storage economics that make high penetration manageable.6,5
Frequency control tells the same story. FCAS revenue for batteries came in at $4.8m, down 51% year on year, as capacity growth saturated the market. The ancillary-services revenue stream that once underwrote early battery business cases is being competed away faster than the energy arbitrage stream.6
The counterargument is that falling per-unit revenue is fine if volume growth outpaces it. Batteries are clearly gaining share of the evening peak, and the dispatch data show they are displacing gas in the exact window where the NEM's residual demand is highest. If batteries are setting prices 39% of the time, they are capturing a larger slice of a smaller pie. Whether that arithmetic works depends on the cost of capital against which the next wave of projects is financed, and BNEF's forecast is that realized prices and battery returns will likely remain under growing pressure as renewable and storage capacity expands.6
On the renewable side, the records keep coming. Queensland hit a maximum instantaneous renewables-plus-storage share of 79.5% at 11:20 on Sunday 31 May 2026, the last day of autumn, up 1.1 percentage points from the previous high. NEM-wide wind production hit an all-time record of 10,349MW in the 21:20 dispatch interval on Wednesday 1 July 2026, 190MW above the prior record set on 25 July 2025, and it happened with roughly 450MW of wind still curtailed. Australia's pipeline of probable clean energy projects jumped about 30% to as much as 32.3GW after government tenders for nearly 10GW of capacity, the biggest surge on record.2,43
That pipeline needs to be priced carefully. More supply into a market already showing 40% to 65% year-on-year price declines, with transmission, supply-chain and planning bottlenecks flagged as the constraint on hitting the 82% renewable target by 2030, points to further compression of capture prices. Developers underwriting new projects at September 2026 forward curves may be underwriting a market that looks nothing like the one the assets will actually operate in. Montel EnAppSys data showing EU renewable generation at a record 384.9TWh in the first quarter of 2026, up 14.5% on the same period in 2025, confirms this is a global pattern, not an Australian quirk.5,61
The quieter signal sits in one asset. Tailem Bend 2, Vena Energy's hybrid battery in South Australia, earned A$27/MWh/day in average FCAS market value in the second quarter, the highest of any battery in the NEM. A hybrid asset earning outsized ancillary revenue while pure arbitrage collapses suggests the value is migrating to configurations that can stack services. That is a portfolio-design problem for every developer who built a single-revenue-stream battery on 2024 assumptions.6
What would falsify the bearish read on storage economics is a quarter where arbitrage revenue stabilizes or rises even as battery capacity grows. The BNEF data for the third quarter of 2026 will be the first real test, given the capacity added through the middle of the year. Another quarter of per-unit arbitrage declines into the $50 to $60/MWh range with FCAS still compressing would confirm it. Watch Tailem Bend 2's FCAS ranking and whether more developers pivot to hybrid designs. If hybrid becomes the default, the standalone battery thesis is already under pressure and the market has not yet repriced the fleet already built.6