Battery storage is scaling rapidly around energy shifting. Spain’s changing voltage-control rules show why the next measure of storage value may be less about installed MW and more about the flexibility a grid can actually use.
When Spain and Portugal lost power at 12:33 CEST on April 28, 2025, the size of Spain’s battery fleet quickly became part of the discussion.
The contrast was striking. By January 31, 2025, Spain had 32,043 MW of installed solar PV, making it the country’s largest generation technology by capacity, according to Red Eléctrica. Battery storage remained at a much earlier stage: Fundación Renovables, citing Red Eléctrica data, later reported 28 MW of installed battery-storage power in April 2025. Red Eléctrica: Spain’s installed solar PV capacity
Those numbers encouraged a simple explanation: Spain had built renewable generation faster than the flexibility needed to support it.
The final blackout investigation points elsewhere.
ENTSO-E’s Expert Panel found several interacting factors, including oscillations, shortcomings in voltage and reactive-power control, differences in voltage-regulation practices, rapid output reductions and generator disconnections. The resulting voltage increases contributed to cascading losses of generation. The panel did not identify insufficient battery-storage capacity as the cause of the blackout. ENTSO-E final report on the Iberian blackout
That distinction matters for the storage industry. Adding battery capacity can improve a power system without automatically providing every form of flexibility the system may need. Technical capability has to be available at the right time and location, visible to the system operator and supported by rules that allow it to be called on.
Spain is beginning to show what happens when that engineering problem becomes a market-design problem.
Battery Storage Is Scaling Around Energy Shifting
The global battery build-out is no longer marginal. The International Energy Agency estimates that 108 GW of battery-storage capacity was added worldwide in 2025, around 40% more than in 2024. Installed battery capacity is now eleven times its 2021 level. IEA Global Energy Review 2026: Battery storage
More revealing is the change in what new projects are primarily being built to do.
In 2015, around 40% of new battery capacity was classified by the IEA as energy shifting, while roughly 45% primarily targeted ancillary services such as frequency regulation and operating reserves. By 2025, energy shifting accounted for more than 90% of annual additions, while projects primarily targeting ancillary services represented about 7%.
That does not mean ancillary services are disappearing. Their absolute deployment has continued to grow, and more projects now combine several services and revenue streams. But the commercial centre of gravity has clearly moved toward storing electricity when it is abundant and releasing it when it is more valuable. Average duration among newly commissioned utility-scale projects increased from about two hours in 2023 to three hours in 2025. IEA: Battery storage is scaling up and taking on a larger system role
The economics are understandable. More solar creates larger periods of low-value or surplus generation and greater demand for storage that can move energy into higher-value hours. That gives developers a scalable revenue case and gives grids a way to reduce curtailment and cover peaks.
But project revenue and system need are not always the same thing.
A developer normally operates a battery around the opportunities available in its market. A system operator is concerned with whichever capability is scarce at a particular point on the network. Energy shifting may be the most valuable service on one day or in one market; voltage support, frequency response, congestion relief or another form of flexibility may matter more elsewhere.
Spain’s blackout makes that distinction unusually visible because the problems identified in the investigation were not primarily about moving energy from midday to evening. Voltage behaviour, reactive power, generator response and operational coordination were central to the event.
The relevant question is therefore not whether the industry is building too many energy-shifting batteries. It is whether market rules also make other useful capabilities worth keeping available.
Spain Is Starting to Put a Price on Flexibility
Spain was already reforming its voltage-control framework before the blackout.
Red Eléctrica’s proposed overhaul of operating procedure PO7.4 was publicly consulted between November 17, 2023 and February 17, 2024 and formally submitted to the CNMC in March 2024. A further CNMC consultation followed later that year. The June 2025 reform therefore cannot accurately be treated as a policy created in response to the April blackout. Official BOE resolution on PO7.4
What PO7.4 tried to change is more interesting than its timing.
The procedure was designed to make better operational use of reactive-power capability available from generation, demand and storage. It introduced a more demanding mode in which participating resources follow real-time voltage or reactive-power setpoints. Voluntary participation can be remunerated, and the framework allows additional reactive-power capacity to be procured where the system operator identifies a need.
An inverter may already possess reactive-power capability. That does not make the capability a grid service. The operator still needs to know it is available, request it, verify the response and give the asset owner a commercial reason to provide it.
The blackout increased the urgency around that work.
In October 2025, the CNMC introduced temporary measures aimed at improving voltage stabilisation. At that point, 161 installations — including combined-cycle, wind, solar PV, biomass, hydro and hybrid plants — had requested qualification testing for dynamic voltage control under the new PO7.4 framework. Official October 2025 voltage-stabilisation measures
The rules were adjusted again in June 2026.
The revised PO7.4 and PO14.4 lowered participation barriers by introducing a fixed-setpoint mode for eligible generation and storage facilities of at least 1 MW, including resources connected to distribution networks. More importantly, the CNMC increased remuneration for voluntary dynamic voltage control, introduced payment for periods when resources provide reactive power without producing active energy, and added compensation linked to installed capacity to help cover technology-adaptation costs. CNMC: 2026 voltage-control procedure changes
Those details reveal where the economics are moving.
Providing flexibility can impose costs even when little or no energy is delivered. A battery may need to preserve inverter headroom, maintain a particular operating range, install additional controls or forgo another revenue opportunity. If the market wants that availability, the owner needs a reason to reserve it.
Spain is beginning to price some of those trade-offs explicitly.
Red Eléctrica reported in June 2026 that 21 GW of installed capacity was already providing the new PO7.4 voltage-control service, including 8 GW from renewables, cogeneration and waste-to-energy resources. That is not a battery-storage figure. Its significance is broader: a pool of existing electrical capability is being converted into an operational service through qualification, control and market rules. Red Eléctrica: 21 GW participating in voltage control
For storage developers, the implication is less about one Spanish voltage-control product than about the relationship between asset design and future revenue. If markets increasingly pay for location-specific availability or inverter capability, decisions that once looked like engineering details can start affecting project economics.
Why LCOS Stops Short of System Value
Levelized cost of storage remains useful. It answers a legitimate investment question: under defined assumptions, what does it cost to build and operate a storage system and deliver energy over its lifetime?
Research on LCOS also makes clear that its result depends heavily on assumptions about capital cost, charging cost, degradation, efficiency, cycling and dispatch. Used consistently, it helps compare storage technologies and configurations.
But cost and value are not interchangeable.
A low-LCOS battery is not automatically the most valuable resource for a particular grid problem. Location can matter. Power and duration can matter. So can whether a project has committed part of its capacity to another service or whether the system operator is able to dispatch it at all. Academic work on LCOS has long noted that limited dispatchability can distort conclusions drawn from cost metrics alone.
The Spanish reforms make the limitation concrete.
LCOS cannot tell a grid operator what it should pay a battery owner to reserve inverter capability rather than maximise energy-market revenue. It does not determine the value of maintaining availability in a constrained part of the network. Nor does it decide whether adapting controls to meet a new technical requirement will produce enough additional revenue to justify the investment.
Those are market-design and opportunity-cost questions.
They also introduce a useful constraint on the storage narrative: not every flexibility problem belongs to batteries. Spain’s voltage-control framework covers generation, demand and storage because several types of resources can provide the service. In another system, network reinforcement, generation controls, synchronous equipment or demand response may prove more efficient.
System value belongs to the capability being provided, not automatically to a particular technology.
That is why the shift matters commercially. Battery hardware costs can continue falling while the value of two otherwise similar projects diverges because one sits in the right location, qualifies for a scarce service or can stack revenue streams that the other cannot.
LCOS remains part of the investment case. It is no longer enough to describe the whole value proposition.
What to Watch Next
Spain’s battery fleet is growing quickly. Fundación Renovables reported, using Red Eléctrica data, that installed battery-storage power increased from 28 MW in April 2025 to 193 MW a year later — a 589% increase from a very small base. Fundación Renovables: Spain battery capacity growth
That headline number matters, but the next set of numbers may tell us more.
How many storage projects qualify for services beyond energy shifting? How much are they paid to remain available? Do new revenue streams change inverter sizing, state-of-charge strategies or project location? And as the system operator publishes more information about zonal flexibility needs, does that begin to affect where storage is actually built?
Those are measurable questions.
They also offer a better way to interpret the storage market than assuming every additional megawatt carries the same system value.
Battery deployment is accelerating, and multi-hour energy shifting has become its dominant application. At the same time, power systems with rising shares of converter-based resources are placing more attention on flexibility and system-strength requirements. The IEA identifies storage alongside grid-forming inverters, synchronous condensers, demand response, flexible generation and stronger networks as parts of that wider toolkit. IEA Electricity 2026: Flexibility
Spain does not prove that the global battery market has already shifted from buying capacity to buying “usable flexibility.” It offers a more useful early signal: technical capability is becoming more valuable when it can be qualified, dispatched and paid for as a defined system service.
A megawatt on an installed-capacity register tells us that an asset exists.
The harder question — and increasingly the commercial one — is what the grid can actually ask that megawatt to do.
