en.Wedoany.com Reported - The Australian Energy Market Operator (AEMO) released the final 2026 General Power System Risk Review (GPSRR) on July 31, confirming that battery energy storage systems (BESS) equipped with grid-forming inverters have not yet been proven to meet the highest level of system strength requirements in Australia's National Electricity Market (NEM), but still listed them as a potential replacement for synchronous machines following the exit of coal-fired generation.
The GPSRR is an annual review required under Clause 5.20A of the National Electricity Rules, prospectively assessing low-probability, high-impact risks facing the grid. The 2026 edition focuses on four priority risks: increased large load connections from data centers, non-credible system strength risks arising from synchronous machine retirements, large non-credible generation and grid incident events, and voltage control risks.
Regarding system strength, the report concludes that with the retirement of coal-fired units, the timely delivery of synchronous condensers and transmission network expansion remain the primary means of maintaining fault levels; grid-forming BESS is explicitly identified as an alternative or supplementary solution under active consideration. The report states that it will continue to assess in a timely manner the contribution of synchronous condensers, grid-forming battery energy storage systems, or other alternatives to NEM system strength.
The report distinguishes between two levels of system strength supply. Grid-forming BESS has been proven to support voltage waveform stability, which is a component of system strength historically provided by synchronous machines; what has not yet been confirmed is protection-quality fault current, a higher standard than voltage waveform stability and a requirement for meeting the minimum system strength level specified in the National Electricity Rules.
AEMO plans to procure "Type 2 Transitional Services" to test under real grid conditions whether grid-forming inverters can provide protection-quality fault current, verifying whether the fault current they produce has sufficient magnitude, duration, and composition to support the reliable operation of power system protection devices.
This distinction has already had direct commercial implications. Transgrid's latest material change of circumstances assessment shows that synchronous condenser costs have overrun by 38%, with total Phase 1 project costs reaching A$1.13 billion (US$920 million), compared with an initial average estimate of A$163 million per site. In response, Transgrid has proposed replacing two of the five planned Phase 2 synchronous condensers with a 900MW grid-forming BESS, pending confirmation of whether the technology can deliver a reliable contribution at the minimum system strength level.
The GPSRR analyzes system strength risks on a scenario-specific basis. AEMO simulated the consequences of the non-credible loss of two large synchronous units, assuming up to three additional large synchronous units are in planned or unplanned outages, using the expected retirements of Eraring Power Station in New South Wales and Yallourn Power Station in Victoria as stress-test scenarios. For New South Wales, the conclusion is relatively optimistic: if synchronous condensers and transmission expansion are delivered on time, even with three large synchronous units already out of service, the simultaneous non-credible loss of two coal-fired units can be managed through operational dispatch without further action. If these projects are delayed beyond Eraring's retirement, the situation deteriorates markedly—without synchronous condensers in place, under multiple outage conditions there may not be enough units to restore system security after certain non-credible contingencies.
Victoria faces greater pressure. If the Hazelwood synchronous condensers are in place before Yallourn's retirement, with three coal-fired units already out of service, restoring system security after the non-credible loss of two Loy Yang units would require up to eight fast-start gas units to synchronize within 30 minutes, a target the report considers difficult to achieve. If the synchronous condensers are delayed until after Yallourn's retirement, the number of fast-start units required would increase to 16, and AEMO acknowledges this scenario may not be achievable within the expected timeframe. The report therefore calls for improved outage coordination to reduce the duration of simultaneous multi-unit outages and for expanded operational procedures to cover specific non-credible contingency events.
As of the report's data cutoff date of July 1, 2026, three events involving the simultaneous loss of two large synchronous units had occurred in the 2025-26 fiscal year: the loss of Yallourn Units 3 and 4 on October 16, 2025, the loss of Callide C3 and C4 units on January 15, 2026, and the loss of both Vales Point units on February 2, 2026. The GPSRR cites these events to note that such low-probability contingencies are not hypothetical, and the timeline for system strength delivery is already quite urgent.
Battery energy storage has reshaped NEM operations faster than any other single technology. In the second quarter of 2026, Australia's grid-scale battery storage capacity surpassed 9,000MW; with rapid expansion of existing facilities, NEM battery price spreads narrowed by 85% year-on-year to an average of A$51/MWh. In the project pipeline, grid-forming capability has become the dominant architecture—74% of the 33.2GW NEM battery storage pipeline uses grid-forming inverters, reflecting both the commercial incentives of system strength contracts and the accumulation of substantial field data on real-world inverter behavior for AEMO.
The proposed fault current testing will build on this field data. AEMO states it will obtain and analyze fault current data from operational grid-forming BESS to understand their response during faults; the Engineering Roadmap and Type 2 Transitional Services will provide the framework for this work. AEMO had previously listed grid-forming BESS as a 2026 priority action, describing it as set to become the "operational heartbeat of the New South Wales power system" after synchronous generation retires. The GPSRR now adds a formal risk assessment dimension to this priority designation, setting out the conditions required for the grid-forming BESS role to expand from efficient-level system strength to the minimum level.
The report also discusses the separate risk posed by data centers as inverter-based loads. AEMO modeling shows that by 2030, without voltage ride-through standards, a single fault on the 330kV grid near western Sydney could cause approximately 1,500MW of data center load to disconnect. Under AEMO's "Step Change" scenario projections, data centers will account for 6% of NEM demand in 2029-30, rising to 10% by 2050. The report supports the Australian Energy Market Commission's (AEMC) proposed "Package 2 load access standards" to address voltage ride-through, active power recovery, and ramp rate risks.
The GPSRR also draws lessons from the April 2025 Iberian Peninsula blackout, where inadequate voltage control led to cascading failures in Spain and Portugal. AEMO recommends extending voltage control risk studies to non-credible contingency events during periods of minimum system load and low fault levels. Power system oscillation risks are listed as a topic for future GPSRR editions or the Transition Plan for System Security; forced oscillations caused by AI training loads, which can cause data center demand to fluctuate within ±60% of facility ratings within seconds, are identified as a risk of subsynchronous interaction with turbine shafts and interregional power flows.
