On paper, Australia’s firming story is accelerating: batteries are scaling, prices are responding, and system operators are formalising new ‘grid security’ toolkits.
In practice, February 2026 is making one thing unmissable for risk committees: firming performance is being defined less by nameplate duration and more by how assets behave under compound system stress – volatility windows, Frequency Control Ancillary Services (FCAS) incentives, and the growing separation between energy and the services that keep the grid stable, according to National Electricity Market (NEM) analysis from Global Roam.
A real-world warning shot: South Australia’s batteries ran close to empty
A late-January event in SA showed how quickly fleet-level duration can disappear when the market stays stressed for hours.
Global-Roam Market Analyst Dan Lee’s post-mortem plotted the Australian Energy Market Operator’s (AEMO) new energy-storage field for SA and found aggregate stored energy peaked at 1,325 MWh (17:05 NEM time) and fell to a low of 66 MWh (21:45) during a sustained high-price window. He notes this was unsurprising given much of SA’s battery fleet is around two hours of duration, while high prices persisted for more than three hours.
This is a reminder that market events can compress usable firming into a narrower window than shorthand suggests, particularly when many units chase the same incentives at the same time.
1.1 Stored energy SA batteries on 27 Jan via Global Roam Dan Lee
The NEM’s price behaviour is itself becoming a central firming risk driver.
Rystad Energy research characterises the NEM as among the most volatile electricity markets globally, driven by large price spreads and supply variability.
Rystad modelling and industry reports show several gigawatts of battery storage and renewable capacity in various stages of construction and commissioning, indicating continued growth in dispatchable assets.
For boards, the risk logic is immediate (and observable in market outcomes, not theory): higher volatility can improve arbitrage value, but it also increases the probability of fleet-wide synchronised behaviour – rapid discharge into the same price window, and diminished capacity later in the evening if stress persists.
A lot of the divergence between modelled firming and operational reality turns on FCAS incentives.
AEMO defines Frequency Controlled Ancillary Services (FCAS) as services that balance power over short intervals (shorter than the dispatch interval). AEMO’s FCAS guide explains FCAS is procured so the power system can be operated safely, securely and reliably.
The risk point for investors is that FCAS and energy arbitrage can compete for the same asset capability at the same time, and FCAS requirements and costs can vary materially. AEMO notes ancillary service costs depend on prices offered and the quantity required, and “can vary substantially from period to period.”
In other words, if your firming model assumes a battery is available for energy whenever you need it, you are implicitly making a view about services-market opportunity cost – and that view can be wrong on the very days stress arrives.
Duration language remains pervasive in public project framing – and that framing can obscure what matters under stress.
The Tomago Battery (500 MW/2,000 MWh) has reached Final Investment Decision and construction is underway, with operations expected in late 2027.
Transgrid has been explicit that batteries are being pulled into non-energy roles as well to stabilise the system.
Transgrid states that big batteries can be configured in ‘grid-forming’ mode to stabilise the high-voltage network in major disturbances such as lightning strikes or generator malfunctions.
AEMO’s grid-forming BESS connection fact sheet notes grid-forming technology can be designed to mimic the behaviour of a synchronous machine.
The implication is that ‘4-hour’ is not a guarantee of availability for the specific service the system needs at the specific moment it needs it – especially when the asset is participating across multiple markets and constraints.
The cleanest signal that operational reality is diverging from old assumptions is the explicit procurement of grid stability capability.
Transgrid has signalled plans for periodic tenders to procure grid-forming battery capacity as part of a broader system strength portfolio alongside synchronous condensers, aiming to progressively scale system strength solutions.
“Synchronous condensers are large spinning machines that mimic the grid-stabilising role of coal generators, enabling the NSW power system to more rapidly accommodate renewable energy generation, delivering cleaner and cheaper electricity to consumers,” Transgrid Executive General Manager Network Jason Krstanoski explained.
GE Vernova Power Conversion and Storage Business Leader, Edgardo Torres, said the technology will enhance grid stability and reliability, thereby strengthening NSW’s transmission network and ultimately supporting Australia’s goal of achieving a more secure and decarbonised power system
As coal retires, the ‘byproduct’ stability services that used to come bundled with energy must be replaced deliberately with contracts, tenders, and portfolios.
AEMO’s most recent Quarterly Energy Dynamics release for Q4 2025 makes the near-term inflection visible: battery discharge “nearly tripling to an average of 268 MW”, supported by 3,796 MW of new battery capacity added since late 2024. Renewables and storage supplied more than half of NEM electricity for the first time during Q4.
So yes, batteries are already moving averages and price outcomes. But the SA depletion event and Transgrid’s system-strength procurement show why the risk lens is shifting: the market benefit is real, and so is the operational edge-case risk.
The ESOO is still often read as a question of energy adequacy – whether there are enough megawatts to meet peak demand. But the emerging risk is different. The grid increasingly depends on discrete system services – inertia, system strength, fast frequency response – that are no longer bundled with energy. Under stress, a battery may exist on paper, but its behaviour depends on incentives, constraints and competing service markets.
The actionable question becomes:
If volatility persists for three hours, constraints tighten, and FCAS prices spike – what, exactly, is your firming asset incentivised and technically able to do?
That’s not a philosophical question. South Australia has already shown what fleet depletion looks like under sustained price stress. New South Wales is already contracting for system strength as a distinct product. And AEMO’s own reporting shows storage is scaling fast enough that these interactions will be more frequent, not less.