The Hidden Grid: How Power Systems Are Rewiring Themselves for a Renewable Future

0.5 seconds. That’s how fast Great Britain’s fastest frequency-response service must activate when the grid wobbles — faster than a human blink, faster than a neuron fires. As wind and solar displace coal and gas, power systems worldwide are no longer just swapping fuels; they’re redefining what it means to keep the lights on. The stability once provided for free by spinning turbines is vanishing, and in its place, a new architecture of speed, precision, and market innovation is emerging. Across seven major electricity systems — from Texas to Denmark, California to Australia — researchers have mapped how grid operators are scrambling to buy not just electricity, but the invisible services that hold the entire system together: inertia, voltage control, dynamic stability, and resilience. What they found isn’t a single blueprint, but a mosaic of adaptation. No two systems are the same. Yet four clear trends are converging: faster response times, explicit pricing for stability, broader participation from batteries and demand-side resources, and location-aware procurement. The stakes? A grid that doesn’t just survive high renewables — but thrives on them.

The Science

The paper by Makrides, Churkin, Teng, and Pinson (2026) is a comparative deep dive into ancillary service markets across Great Britain (GB), Germany (DE), Denmark (DK), Italy (IT), Texas (ERCOT), California (CAISO), and Australia (AEMO). These systems were chosen not for their size alone, but for their leadership in renewable integration and market innovation. The authors conducted a document-based analysis of official grid codes, market rules, regulatory filings, and technical reports, supplemented by academic literature. Their goal was not to rank systems, but to understand how different institutional, regulatory, and technical contexts shape the design of ancillary services — the suite of functions that maintain grid stability beyond the simple delivery of energy.

Ancillary services fall into five categories: frequency control and reserves, voltage and reactive power support, system stability, restoration, and congestion management. Historically, these were byproducts of conventional power plants: a coal generator spinning at 3,000 rpm provided inertia, voltage support, and fault ride-through simply by existing. But inverter-based resources (IBRs) — solar panels, wind turbines, and batteries — don’t inherently provide these services. They can be programmed to do so, but only if the market rewards it.

The study’s methodology is qualitative but systematic. It compares across five dimensions: (1) how services are defined and classified, (2) how they’re procured and remunerated, (3) how markets are structured and coordinated, (4) which technologies can participate, and (5) how reforms are evolving. The authors introduce a consistent classification framework to cut through the jargon — for instance, mapping Germany’s Frequenzhaltung and California’s Regulation Service onto a common taxonomy. They also create standardized visualizations of response times and durations, enabling apples-to-apples comparisons across systems (

,

,

).

This isn’t just academic taxonomy. As the authors note, “the value of a resource depends not only on available capacity but also on speed, accuracy, location, duration, and compliance with defined performance requirements” (Makrides et al., 2026). A battery in South Australia that can inject reactive power during a voltage dip is worth more than one in Queensland if the fault is local. A wind farm with grid-forming capability in Texas can prevent cascading failures where a conventional inverter would disconnect. The challenge is designing markets that see these differences — and pay for them.

What They Found

The most striking finding is the acceleration of response times. In Great Britain, the Dynamic Containment (DC) service requires full activation within 1 second. In Australia, the Fast Frequency Response (FFR) product responds in under 2 seconds. By contrast, traditional primary frequency response in many systems takes 10–30 seconds — too slow for grids with low inertia, where frequency can collapse in under 5 seconds after a major fault.

Response Time and Sustainment Duration of Frequency-Control Products

Label	Value
GB - Dynamic Containment	1
AU - FFR	2
GB - Dynamic Moderation	0.5
DE - FCR	30
US - Regulation	4

Source: Makrides et al. (2026), based on system operator data.

This shift is driven by physics. As synchronous generators retire, system inertia — the kinetic energy stored in spinning masses — declines. Lower inertia means a faster rate of change of frequency (RoCoF) after a disturbance. The paper notes that “inertia limits the RoCoF immediately after a disturbance, giving frequency-control actions more time to respond before frequency reaches its nadir” (Makrides et al., 2026). Without fast-acting services, the grid risks under-frequency load shedding or even blackouts.

The second major finding is the explicit commodification of stability. Inertia, system strength (measured by short-circuit level), and dynamic voltage support — once free public goods — are now being procured as market products. Great Britain runs a Stability Network Service (SNS) tender for inertia and system strength. Australia has introduced the Minimum System Strength (MSS) and Multi-Technology Black Start (MTBS) services. Germany is piloting market-based inertia procurement. These are not just technical upgrades; they represent a philosophical shift: stability is no longer assumed — it’s bought.

Third, participation is broadening. Batteries, demand-response aggregators, and even electric vehicles are now eligible for ancillary services in most systems. In Texas, ERCOT allows distributed energy resources (DERs) to bid into frequency regulation. In California, CAISO has integrated grid-forming inverters into its reliability planning. The key enabler is performance-based access: instead of requiring a specific technology, markets now ask, “Can you deliver the service?” This opens the door to innovation.

Finally, location matters more than ever. Voltage support and congestion management are inherently local. A reactive power injection in one part of the grid may do nothing — or even harm — another. As a result, systems are moving toward locational procurement. Great Britain runs regional tenders for voltage support. Texas manages reactive power at the nodal level. California directs voltage support through system-operator instructions. The paper notes that “the value of a resource depends on its position in the network and its ability to relieve specific constraints” (Makrides et al., 2026).

Procurement Approach for Voltage and Reactive Power Support

Label	Value
Great Britain	60 %
Germany	40 %
Italy	20 %
Denmark	30 %
Texas	50 %
California	10 %
Australia	45 %

Estimated share of voltage support services procured through market mechanisms vs. grid-code mandates.

Despite these trends, no two systems are alike. The paper’s comparison reveals a spectrum of approaches. In Europe, Germany, Denmark, and Italy follow a harmonized hierarchy of Frequency Containment Reserve (FCR), automatic Frequency Restoration Reserve (aFRR), and manual FRR, coordinated through cross-border platforms like PICASSO. In contrast, the U.S. and Australia have more fragmented, system-specific designs. Texas relies heavily on operational directives, while California uses a co-optimized energy and ancillary service market. Great Britain stands out for its granular, market-driven approach, with over a dozen distinct ancillary products.

The table below summarizes how different systems handle voltage and reactive power support — a service where the divide between mandatory and market-based approaches is starkest:

System	Baseline provision	Additional procurement	Locational implementation
Great Britain	Mandatory via grid code	Market-based ERPS and VNS	Explicitly locational
Germany	Mandatory capability	Market-based where needed	Network-specific
Italy	Mandatory capability	Regulated/operational	Grid-code specified
Denmark	Mandatory capability	SO arrangements	Grid-connection based
Texas	Mandatory capability	SO-directed	Nodal/local
California	Operational obligations	SO instructions/contracts	Outside market
Australia	NSCAS framework	RSAS/MBAS contracts	Targeted to needs

The lack of harmonization is both a challenge and a feature. It allows systems to tailor solutions to local needs — but it also limits learning and interoperability.

Why This Changes Things

The implications of this shift are profound. First, it redefines the economics of the energy transition. For years, the debate centered on the levelized cost of energy (LCOE) — how cheap can we make wind and solar? But as this paper shows, the true cost of a renewable grid includes the price of stability. A solar farm with grid-forming inverters may cost more upfront, but it avoids the need to pay $10 million for a synchronous condenser. A battery that provides fast frequency response may earn three times the revenue of one that only arbitrages energy prices.

Second, it democratizes grid services. Once, only large power plants could provide ancillary services. Now, a fleet of smart water heaters in California can bid into frequency regulation. A solar-plus-storage microgrid in South Australia can sell inertia to the main grid. This isn’t just about efficiency — it’s about resilience. Distributed resources can respond faster and more locally than centralized plants, reducing the risk of cascading failures.

Third, it forces a reckoning with the limits of markets. Ancillary services are not like coffee or crude oil. Their value is contextual — a function of time, location, and system state. A megawatt of frequency response is worth nothing when frequency is stable, but priceless during a blackout. As the paper notes, “incomplete valuation of flexibility and stability capabilities whose value depends on location and operating conditions” remains a key challenge (Makrides et al., 2026). Pure market mechanisms may underprice these services, while pure regulation may stifle innovation.

The comparison also reveals a deeper truth: there is no one-size-fits-all solution. The authors find “no convergence towards a single market design” — and that may be a good thing. Great Britain’s market-driven approach works for an island grid with limited interconnection. Germany’s harmonized European model suits a tightly coupled continent. Texas’s operational flexibility fits its isolated, deregulated market. The right design depends on geography, governance, and grid topology.

But the absence of convergence doesn’t mean the absence of lessons. The four trends — faster response, explicit stability, broader participation, and locational awareness — are universal. They reflect the physics of low-inertia grids, not the preferences of regulators. Any system with high renewables will eventually face the same challenges.

What’s Next

The road ahead is not without obstacles. The paper identifies four persistent challenges:

1. Limited harmonization: Especially for non-frequency services, product definitions and procurement vary wildly. This makes cross-border coordination harder and slows the spread of best practices.
2. Incomplete valuation: Markets still struggle to price services whose value is situational. How much is inertia worth on a windy Sunday afternoon? We don’t have good answers.
3. Barriers to participation: Despite progress, many distributed resources face technical, regulatory, or financial hurdles to entering ancillary markets. Minimum size thresholds, complex qualification processes, and lack of visibility remain issues.
4. Weak coordination: Energy, balancing, and ancillary service markets often operate in silos. This can lead to inefficiencies — for example, paying a gas plant to run just to provide voltage support, while a nearby battery sits idle.

The authors suggest a path forward: differentiated but coordinated frameworks. Not one global market, but a set of interoperable designs that respect local needs while enabling learning and trade. They advocate for “capability- and performance-based access requirements” — letting any technology compete if it can deliver the service.

The most exciting frontier is the integration of artificial intelligence and real-time system assessment. Imagine a grid where stability services are procured not on a fixed schedule, but in response to live forecasts of inertia, voltage sensitivity, and fault risk. AEMO in Australia is already experimenting with such “dynamic” procurement. CAISO is exploring grid-forming inverter standards. NESO in Britain is developing a multi-horizon stability market.

The ultimate goal is a grid that doesn’t just accommodate renewables — but leverages their full potential. Inverter-based resources can respond faster, more accurately, and more flexibly than any steam turbine. The challenge is to build markets and rules that see that potential — and pay for it.

As the authors conclude, “ancillary-service markets are not converging towards a single design.” But they are evolving — rapidly, diversely, and with increasing sophistication. The hidden grid is being rewired. And when the next storm hits, or a generator trips, it may be a battery in Texas, a wind farm in Denmark, or a thousand smart thermostats in California that keep the lights on.