Can demonstrating inertia to change be valuable in a fast-changing world?

The Spanish electrical power network collapse in the end of April 2025 bears similarities with the 2016 South Australia network collapse.

Firstly, it illustrates that successful electrical power system operations are not usually celebrated. The absence of news is the good news.

And it takes immense effort to plan and operate these systems reliably without interruptions. Power systems are inherently instable, and they can collapse within seconds.

The resilience of the power systems is tested when events or contingencies take place. These can be environmental events, or unexpected large equipment outages, and generally a combination of both taking place at the same time.

Like on the flight deck of a jetliner, response time is of the essence. Power systems do not have much of it, only a couple of seconds in the best cases.

There is one index, among many, that identifies how fast the situation is getting out of control: the ‘ROCOF’. That is, the rate of change of frequency of a network. It measures how fast the frequency is falling. It is not dissimilar to the decent speed of a plane: it is ideally not too fast if you encounter a bundle of unexpected issues.

The faster the ‘ROCOF’, the less time you have to take measures to bring back stability and avoid losing the whole power system.

That is where rotating machines come into the picture and offer support: the energy stored in large spinning equipment such as alternators and turbines will buy time, by instantaneously releasing their kinetic energy when their rotation speed decreases at the same pace the electrical frequency does.

About 20 years ago, on many the large power systems the ‘ROCOF’ was about 1 Hz/s in most scenarios: the frequency was not dropping by more than a single Hertz per second. That assumption was a cornerstone to define the blueprint to deploy on bad days.

With the progressive integration of renewable and power electronic based generation, rotating machines are gradually accounting for a lower percentage of power generation in the network. Given this, frequency tends to be harder to control, even during normal operations, as illustrated by the east-coast Australian system data that is shown below.

Stealthily, the rotating inertia lowered in many grids, and the potential ROCOF increased. There are no major signs of an issue until we face an unplanned resilience test.

The figure below shows the frequency during the 2016 South-Australia event, and how the collapse took place within a second, and how fast the frequency sunk.

The next figure is the Spanish system frequency during a similar event.

There are similarities with the Spanish event: a network with high penetration of renewable generation, on the periphery of larger system having a large amount of inertia. When the interconnection with the large system is lost, the ‘ROCOF’ can increase drastically, sometimes above 5 times the 1 Hz/s mark.

 What shall we do about this?

Keep all the thermal generation online and slow down the integration of renewable generation?

 Not necessarily - there are other ways.

Networks in Ireland and South Australia have passed resilience tests over the last few years, despite high levels of renewable energy in the generation mix and despite their location on the periphery of larger systems.

As pointed out by many already, as a response to the Spanish system collapse in 2025, large storage systems powered by batteries can provide (synthetic) inertia: they can rapidly and autonomously release energy when the frequency drops, and lower the ‘ROCOF’ to more traditional levels.

And an underrated feature of large battery systems providing inertia is that they offer flexibility on the amount of inertia provided, it is a setting in a control loop, not a rotating mass characteristic (masses and shafts tend to be also more prone to participate to sub-synchronous oscillations).

Emulated inertia projects in Australia have demonstrated these features as reliable over the last 10 years: In Western Australia in islanded large mining systems, and in South Australia.

Alternative techniques to provide a boost of energy when the frequency changes too rapidly are also considered. The power electronic based generation and storage technology can also offer responses that are circa 50 times faster than thermal and nuclear generation responses. Some of these responses rely on Fast Frequency Response (FFR): A large amount of energy is injected within 50-100ms to the grid, as a one shot, when needed. The “FFR” is used to tame the ‘ROCOF’.

Considering South-Australia, the concept of Emergency Under Frequency Response (EUFR) was introduced by the Australian Energy Market Operator. The EUFR quantification is a dynamic process, taking into consideration the demand and generation mix (and the network inertia in particular). The EUFR includes the response from Under Frequency Load Shedding (UFLS), the frequency response from energy storage and renewable generation which can also contribute to arrest a fast frequency decline. Multiple contingency events are selected on their probabilistic representativity to determine adequate reserves dynamically.

In Europe, an ENTSO-E study  "Inertia and Rate of Change of Frequency"  identified that the UFLS response might not prevent anymore network collapse following events resulting in ‘ROCOF’ higher than 1 Hz/s.

In Ireland, EirGrid identified ROCOF as a concern and developed the “Look Ahead Security Assessment Tool” providing a real-time assessment of system security and future generation.

FFR services are procured when needed and synthetic inertia supply is considered as a future service.

Beyond proven and emerging solutions, several key takeaways stand out:

·         Managing a power system is inherently complex - and success is often invisible. We’re implementing major changes on the fly after more than 150 years of relatively stable strategy. Energy operators and regulators may face criticism for moving too slowly—or too fast—but it’s important to recognise the scale and complexity of the challenge they’re navigating.

·         Higher ‘ROCOF’ values are putting traditional go-to solutions under pressure. As the rate of change of frequency exceeds the historical 0.5 to 1 Hz/s range in some regions, the effectiveness of legacy responses like under-frequency load shedding is diminishing. A broader toolkit of solutions is now required to manage these risks.

·         Should inertia - real or synthetic - be monetised? If inertia is both valuable and in limited supply, should a market mechanism be introduced? It’s a question worth exploring from an economic standpoint.

·         Power system studies may be undervalued, yet they’re more critical than ever. These dynamic simulations are the equivalent of flight simulators for pilots—vital for preparing operators to handle unexpected events. As systems evolve, the role of such studies will only grow in importance to define and refine emergency protocols.

·         Most crucially, knowledge sharing across geographies is key. The similarities between the South Australian and Spanish events highlight the need for global collaboration. Practical system knowledge is evolving faster than at any point in history—and sharing it is essential to building resilience.

AECOM is a global infrastructure consulting firm that brings together deep local insight with international expertise.

Our strength lies in integration—connecting technical excellence, sector knowledge, and global experience to deliver holistic solutions.

As the energy transition accelerates, we believe knowledge sharing across borders is not just valuable - it is essential. It is one of the ways we deliver smarter and more resilient outcomes for our clients.