Australia's grid runs near-fully renewable in simulation with 24 GW of storage

Australia's grid runs near-fully renewable in simulation with 24 GW of storage A new simulation pegs 98.5% renewable supply on Australia's main grid with modest storage, as May set fresh wind and solar records despite a generation drought. A simulation of Australia's National Electricity Market found that 24 GW and 125 GWh of storage, mostly four-hour and eight-hour batteries, would have delivered 98.5 per cent renewable electricity on the main grid over the past four years.³ That number matters because it puts a concrete storage figure on a question traders and TSOs keep circling: how much battery capacity it actually takes to push a large synchronous grid close to fully renewable. The remaining 1.5 per cent gap was filled by "other" sources, rarely between September and March but regularly from April to August, when sun and wind thin out.³ The seasonal pattern is the catch. Short-duration batteries handle the daily swing well, but they do not cover a winter stretch when renewable output sags for days at a time. That is the gap the 1.5 per cent represents, and it is the part of the problem cheap four-hour cells do not solve.³ Snowy Hydro made the same argument this week (2026-06-03) as it prepared the market for another blowout in Snowy 2.0 costs. "Short-term batteries are the sprint-runners, vital for daily balancing, but only Snowy 2.0 can run the marathon," the company's Dennis Barnes wrote, conceding that batteries are the best-value option for short-duration storage while insisting they cannot see the grid through longer lulls.³ The real-world data underneath the simulation has been moving fast. Rystad Energy reported that Australia's large-scale solar and wind assets generated 4.6 terawatt-hours in May (2026-05), up 10 per cent from 4.2 TWh in May 2025.² What makes that figure notable is the weather it was set against. May (2026-05) brought the worst wind and solar drought on the main grid since 2022, yet four new monthly records still fell.² Victoria led the country, generating 1,218 GWh of combined utility wind and solar, including a record 1,079 GWh from wind and 139 GWh from utility PV. Queensland set its own May wind record at 625 GWh.² Asset-level numbers show how uneven the fleet remains. The 175 MW White Rock Wind Farm near Glen Innes in New South Wales topped capacity factors at 45.6 per cent, with Western Australia's Mumbida and Badgingarra farms close behind at 44.6 and 43.6 per cent.² For utility solar, the best performers were almost all in Queensland, led by Pacific Blue's 100 MW Haughton Stage 1 in the Burdekin at 24.3 per cent AC capacity factor. The spread between a 45 per cent wind site and a 24 per cent solar site is exactly why storage sizing depends so heavily on geography.² The price signal points the same direction as the simulation. Rystad noted that wind and solar continue to push down spot prices on the electricity market while batteries smooth volatility, the combination that makes a battery-heavy build look viable rather than aspirational.² This was the backdrop in late January (2026-01-29) when demand on Australia's largest grid hit a fourth-quarter record and renewables supplied more power than fossil fuels across the country. The trend in the data is consistent even as the year-to-year weather swings widely.¹ For coal, the read is bearish at the margin. More renewable generation and falling spot prices erode the economics of thermal plant on the main grid, and the cross-market chain runs from softer Australian demand toward Newcastle coal and, downstream, the cargoes that feed Asian LNG buyers.² But the simulation is a backcast, not a deployment plan. It assumes 24 GW and 125 GWh already built and dispatching optimally, and it still leaves a winter gap that four-hour and eight-hour cells cannot close. That is the unresolved question.³ The thing to watch is whether the 1.5 per cent "other" share shrinks or hardens as more storage arrives, and whether long-duration projects like Snowy 2.0 land at a cost that lets them, rather than coal, fill the April-to-August lulls. The next Snowy 2.0 cost number will tell traders a lot about which way that bet breaks.³