Featured image: Electricity pylons illustrate a network whose stability depends on coordinated equipment. Contextual photograph; not the Kauai grid or AEMO’s trial site. Photo: Matthew Henry / Unsplash. Unsplash licence.
A battery’s megawatts tell us how quickly it can deliver power. They do not tell us how its electronics behave in the first moments after a generator trips or a fault appears on the network. That behaviour increasingly matters to grid reliability.
Grid-forming inverters address this fast stability problem. This is a background explainer, drawing on 2026 technical specifications and field reporting alongside ongoing Australian work. It is not an announcement that one new device has solved every challenge of a renewable-heavy grid.
The inverter helps establish an electrical reference
An inverter connects a source such as a battery or solar installation to an alternating-current grid. Conventional grid-following controls use the existing network signal as a reference. Grid-forming controls maintain an internal voltage reference and respond rapidly when the surrounding network changes.
The UNIFI consortium’s Version 3 specification, released on 30 January 2026, defines that behaviour on very short timescales and sets out performance requirements and tests. The important feature is the response of the device, rather than whether its software carries a fashionable label. Read the UNIFI specification.
Several such devices must still synchronise and share power with other resources. A stable reference does not make available energy or current unlimited. The underlying source and equipment ratings constrain what the inverter can deliver during a disturbance.
An island grid provides evidence beyond a simulation
On 8 January 2026, the National Laboratory of the Rockies reported work with Kauai’s electricity cooperative. A large generator trip in 2021 had triggered network oscillations. Researchers used recorded events, detailed modelling and hardware testing to identify interactions and examine different controls. Read the laboratory’s account of the Kauai work.
In 2023, the same large generator tripped again. The laboratory reports that, after grid-forming controls had been added, the earlier oscillation did not recur. That is useful operational evidence connected to a specific grid and event.
It does not establish that any grid-forming inverter will stabilise any network without engineering study. The strength of the example is the chain of evidence: observed behaviour, a model that reproduces it, hardware validation and a later field event.
Protection systems need a different kind of current
Grid stability and fault protection are related but different jobs. Stability concerns keeping electrical behaviour controlled after disturbances. Protection concerns identifying a fault and isolating the affected part of the system quickly enough. A device can contribute to one task without proving every requirement of the other.
AEMO has outlined a trial to investigate whether grid-forming inverters can provide protection-quality fault current. Its stated questions concern the current’s magnitude, duration and composition, and whether protection relays operate correctly under different conditions. Historically, the National Electricity Market has relied on synchronous machines for this service. AEMO’s fault-current trial description.
The published page sets out a procurement plan and trial objectives. It should not be treated as a report of successful results. A plan to test a capability is evidence that the capability matters, not confirmation that every candidate device already supplies it.
Specifications have to support cooperation between devices
AEMO is also reviewing technical requirements for grid-forming plants. Its consultation describes work to clarify definitions, account for capabilities and limitations, and validate proposed requirements with stakeholders. This is standards-development work, rather than a final blanket certification for all equipment. AEMO’s technical-requirements review.
Interoperability means devices can work together. In a grid, that includes sharing a response without amplifying each other’s oscillations or conflicting with protection. The relevant evidence must cover the plant’s intended configuration and operating conditions, rather than a single idealised test.
A useful project assessment would therefore distinguish energy capacity, fast control behaviour, fault response and restart capability. They serve different purposes and need their own evidence. Treating them as one feature called grid-forming would make procurement less precise.
The technology is important because power electronics can provide stabilising behaviour that grids once obtained mainly from spinning machines. The next task is rigorous integration: specify the service, test the equipment and verify that the whole network behaves as intended.
Leave a Reply