By AGILE Consulting Engineers, Solar PV and Battery Energy Storage Systems (BESS) specialists.
Inertia in the National Electricity Market (NEM) has fallen from around 93,000 megawatt seconds in 2017 to roughly 84,000 megawatt seconds by 2025, as synchronous generating units retire and non-synchronous plant, including solar and batteries, takes their place. A virtual power plant (VPP) built from thousands of small inverter-based devices sits right in the middle of that shift. Done properly, it is part of the answer to declining system security. Done carelessly, it is one more source of unpredictable, non-synchronous behaviour on a grid that already has less margin than it used to.
Table of Contents
- System strength, inertia and frequency response, briefly
- Why declining inertia matters for VPP design
- MASS, verification and why a VPP has to prove itself
- Response time and control design as a security feature
- The tension between distributed control and centralised oversight
- AEMO’s Engineering Framework and where this is heading
- What this means for VPP and BESS design decisions
- What to do next
- FAQ
System strength, inertia and frequency response, briefly
Three related but distinct concepts sit underneath most conversations about grid security in a high-renewables NEM. Inertia is the stored kinetic energy in the spinning mass of synchronous generators and large motors, which naturally resists sudden changes in system frequency in the first second or two after a disturbance. System strength describes how much a network can maintain stable voltage during and after a disturbance, which matters most in areas with a high concentration of inverter-based generation and relatively few nearby synchronous machines. Frequency response, delivered through Frequency Control Ancillary Services (FCAS), is the market mechanism that pays generators, loads and increasingly aggregated DER to actively correct frequency deviations after the first, inertia-driven moment has passed.
None of these are abstract concerns for AEMO. The Australian Energy Market Operator’s 2025 Transition Plan for System Security, released in December 2025, now combines what used to be separate system strength, inertia and reserve gap reporting into a single unified assessment, reflecting how closely these three factors interact as the generation mix changes.
Why declining inertia matters for VPP design
A conventional coal or gas turbine contributes inertia simply by existing and spinning, whether or not it is actively being dispatched for energy. A battery energy storage system (BESS) or a rooftop solar inverter contributes none of that automatically. It can be made to synthesise a fast, frequency-responsive contribution through its power conversion system (PCS) and control software, but that behaviour has to be engineered in deliberately, tested, and verified, rather than assumed as a free byproduct of the hardware being connected.
This is precisely where VPP design intersects with system security rather than sitting apart from it. A fleet of thousands of household batteries coordinated by a distributed energy resource management system (DERMS) can, in principle, deliver a meaningful, fast-acting frequency response, because the aggregate response time of well-designed inverter-based resources can be faster than a spinning turbine’s governor response. The catch is the word “can.” Whether it actually does depends on how the fleet’s control logic, communication latency and dispatch verification are engineered, not on the batteries existing at all.
MASS, verification and why a VPP has to prove itself
The Market Ancillary Services Specification (MASS) is the rulebook that defines how a participant proves it can deliver a contracted FCAS response, covering measurement accuracy, response timing and the metering standard used to verify delivery. A conventional generator has decades of established metering and testing practice behind it. A VPP built from thousands of geographically dispersed, individually owned devices has to prove the same thing at fleet scale, which is a materially harder verification problem.
AEMO’s VPP Demonstrations program, run in partnership with the Australian Renewable Energy Agency (ARENA) and several other bodies, spent years working through exactly this question, testing whether a relaxed version of MASS could still give AEMO confidence that an aggregated fleet would deliver contracted FCAS. One outcome from that work was a specification allowing power flow and frequency measurement at one-second time resolution for trial participants delivering fast FCAS, tighter than what many legacy metering setups could support. The lesson generalises well beyond the trial. A VPP’s dispatch verification, telemetry resolution and testing regime are not paperwork sitting alongside the engineering. They are the engineering that determines whether the fleet is actually trustworthy from a system security standpoint.
Response time and control design as a security feature
Response time is not a single number for a VPP. It is a chain: sensing at the device level, communication from the device to the aggregator’s energy management system (EMS), a dispatch decision at the DERMS layer, a command sent back down to the device, and finally the PCS acting on that command. Each link adds latency, and each link is a potential point of failure or delay that a purely centralised, single-generator system does not have to manage in the same way.
Good VPP control design treats this chain as a safety-relevant system, not just a commercial dispatch pathway. That means conservative assumptions about the slowest devices in the fleet rather than the fastest, fail-safe default behaviour if communication is lost, and enough margin in the certified response time that the fleet does not become a source of oscillation or delayed correction during a genuine frequency event. None of this is exotic control theory. It is the same discipline applied to conventional generator governor tuning, just distributed across thousands of endpoints instead of one turbine hall.
The tension between distributed control and centralised oversight
Here is the genuine tension in the current NEM architecture. AEMO needs system-wide visibility to manage frequency, voltage and system strength in real time, and that visibility has historically come from a comparatively small number of large, well-instrumented generators reporting into central SCADA and market systems. A VPP fleet inverts that picture: enormous numbers of small devices, individually invisible to AEMO, whose aggregate behaviour is what matters for system security, but whose individual behaviour is only visible to the aggregator’s own platform.
This is not a flaw unique to VPPs so much as a structural feature of a high-DER grid that AEMO, the Australian Energy Market Commission and distribution network businesses are actively working through, including via the DER Register, which aims to give networks and AEMO better visibility of what DER capacity actually exists and where. A VPP operator that treats its own telemetry and reporting obligations as a genuine contribution to system visibility, rather than a compliance cost to be minimised, is aligning its commercial interest with the grid’s actual security needs. One that treats visibility reporting as an afterthought is quietly working against the thing that makes its own fleet valuable in the first place.
AEMO’s Engineering Framework and where this is heading
AEMO has been building out a broader Engineering Framework and an associated Frequency Control Workplan to prioritise the technical work needed to keep frequency control reliable as the generation mix keeps shifting, with reporting on this work folded into the Transition Plan for System Security from 2026 onwards. The direction of travel is toward more explicit, more frequently updated technical requirements for anything providing frequency response or contributing to system strength, VPPs included, rather than a static rulebook written for a synchronous-generator-dominated grid.
For anyone designing or specifying VPP-capable assets today, the practical implication is to design for where these requirements are heading, not just where they sit today. A fleet certified against yesterday’s minimum verification standard is a weaker long-term asset than one built with headroom to meet tighter measurement, latency and reporting requirements as they arrive.
What this means for VPP and BESS design decisions
Three practical design questions follow from all of this. First, does the PCS and control firmware in the proposed asset actually support the fast, verifiable frequency response behaviour the commercial model assumes, under real grid disturbance conditions, not just a lab test. Second, is the telemetry chain, from device to DERMS to AEMO-facing reporting, built to the resolution and latency that a MASS-grade FCAS registration will eventually require, or does it rely on legacy metering that will need retrofitting. Third, does the fleet’s control philosophy fail safely if communications degrade, in a way that protects rather than undermines system security during the exact conditions a frequency event creates. These are not questions a commercial term sheet answers. They are engineering questions that need to be worked through against the physical and control characteristics of the specific hardware being proposed.
What to do next
If you are specifying a VPP-capable BESS fleet, or a network connection that needs to account for a growing volume of distributed, frequency-responsive assets behind it, this is the point where an engineering review of the control and verification architecture is worth doing early, before assets are ordered or a connection agreement is locked in. We have worked through exactly this kind of review with teams before they committed to hardware.
FAQ
Why does declining inertia matter if batteries can respond faster than turbines anyway?
A battery can respond quickly, but only if its control system, communications and dispatch verification are specifically engineered to do so and proven under real conditions. Declining inertia narrows the margin for error, which makes that engineering work more important, not less.
What is the Market Ancillary Services Specification (MASS)?
MASS is AEMO’s rulebook defining how a participant, including an aggregated VPP fleet, proves it can deliver a contracted Frequency Control Ancillary Services (FCAS) response, covering measurement accuracy, response timing and verification standards.
Does AEMO have direct visibility of individual household batteries in a VPP?
Generally no. AEMO relies on the aggregator’s own reporting and, increasingly, on broader visibility initiatives such as the DER Register, rather than direct telemetry from every individual device.
What is the difference between system strength and inertia?
Inertia resists sudden frequency changes in the first second or two after a disturbance, largely through the physical spinning mass of synchronous plant. System strength refers to the network’s ability to maintain stable voltage during and after a disturbance, and is a separate, though related, characteristic.
Can a VPP make grid security worse rather than better?
Yes, if its control design, verification and telemetry are not engineered to a fleet-scale standard equivalent to conventional generator requirements. A poorly designed aggregated fleet can introduce unpredictable or delayed behaviour during a frequency event.
What is AEMO’s Transition Plan for System Security?
It is AEMO’s unified report, released in December 2025, combining system strength, inertia and reserve gap assessments that were previously reported separately, reflecting how interconnected these system security factors have become.
