The Age of Electricity is underpinned by rising investments in new resources. These include growing converter-based variable solar PV and wind, battery storage systems, as well as spatially and temporally concentrated demand from EVs, heat pumps and large loads like data centres. Combined with the expansion and upgrade of transmission and distribution grids, substantial increases in the flexibility of power systems are required for secure and cost-effective integration of generation, load and storage technologies that characterise this new era.

Last year’s report, Electricity 2025, focused on measures to enhance flexibility on the supply side of electricity markets by reducing the various technical, regulatory or contractual inflexibilities affecting generation. This year, we focus on demand-side applications for providing and improving flexibility. In addition, we have updated our analysis on supply side flexibility to include further examination of falling “capture rates” for solar and wind generation and the spectacular growth of battery storage systems amid declining costs in various markets.

Finally, markets with high shares of solar and wind are challenged by periods of overabundance, when supply has the potential to exceed demand for grid electricity, necessitating operators to implement measures to balance the system. Different options are addressed, from increasing control of output from small solar installations, to higher battery deployment, to refining price signals, including offering free electricity to customers.

With the share of variable renewable energy (VRE) sources in electricity supply rising rapidly in many regions and end-use sectors such as heating and transport becoming progressively more electrified, demand flexibility has become an essential component for power systems. In this evolving energy landscape, demand response (DR) programmes are becoming an increasingly important flexibility tool, enabling households and businesses to shift or shed their electricity use in response to grid or market signals.

In exchange, DR plans offer financial incentives, typically through contractual arrangements or participation in programmes with utilities, aggregators or system operators, or by direct participation in power markets in the case of large industrial consumers. For end-users, DR flexibility contracts provide inducements in the form of direct payments, rebates, or bill credits for curbing consumption during high-priced, peak-demand periods. When deployed at scale, DR providers can reduce peak capacity requirements, defer grid investment, lower integration costs for renewables, and strengthen resilience during system stress (see IEA’s The Value of Demand Flexibility report, December 2025).

The potential for demand response is enormous. Yet, despite its substantial system-wide and consumer benefits, DR implementation globally largely remains untapped. Our analysis shows that, as of 2024, only around 100 GW of demand response is utilised on a global basis. For example, aluminium production accounts for around 160 GW of peak electricity demand in the world, but only a small share of this is currently leveraged for demand response. Similarly, residential air conditioning (AC) contributes around 600 GW, yet flexible demand response from AC loads remains marginal. Unlocking a larger share of the overall demand response potential would be possible by applying various technologies already available today, as well as by adjusting market frameworks and regulatory mechanisms accordingly across different sectors.

Demand response potential can be characterised in different ways, depending on assumptions on technology availability, deployment levels, costs, and real‑world participation, including adoption barriers. Currently utilised demand response refers to the capacity already active in markets or programmes. However, in many cases, only part of the economically viable potential is realised because market barriers, regulatory frictions, and behavioural inertia prevent full participation.

Unlocking more of this economic potential becomes possible when adoption constraints and market barriers are reduced or removed. Going beyond this scope towards the technically enabled potential requires valuing flexibility appropriately. Technically enabled potential includes all loads that already possess the monitoring and control capabilities needed for participation, though this remains significantly below the theoretical potential. Theoretical potential encompasses all electricity demand that could, in principle, be shifted or reduced, assuming full deployment of enabling technologies and subject only to physical and end‑use service constraints.

Currently, the industrial sector provides the largest share of demand response across all sectors, with around 75 GW of utilisation. Industries across many regions currently participate in explicit DR where large consumers commit to reduce consumption during periods of system stress, typically in exchange for financial incentives such as capacity payments or activation payments when load is curtailed. Large industrial facilities with batch-based processes or on-site flexibility options often have processes that can be curtailed or shifted for short durations with relatively limited impact on output, allowing for a meaningful portion of their technical flexibility to be mobilised.

Much greater potential exists for industrial DR. Many industrial processes, especially those that rely on heat or operate in batches, can shift their electricity use more routinely via automation or thermal storage. This includes cold-storage and compressor-driven systems in manufacturing as well as melting and heat-treatment furnaces in metal processing. Overall, DR-enabling measures can raise the share of industrial load that can be adjusted with limited impact on production, yet they are not fully deployed today, leaving a high share of flexibility unused. 

In the buildings sector, demand response utilisation is estimated at around 30 GW globally. Many building loads, such as space heating, cooling and water heating, can respond to price signals or automated controls when equipped with controllable technologies, giving the sector large long-term DR potential. However, only a small portion of this potential is currently being realised. In practice, flexibility can often be delivered through short-duration measures, such as brief compressor cycling or small thermostat set-point adjustments, with limited impact on occupant comfort, underscoring that the DR potential in buildings is far greater than current utilisation.

Limited market penetration of enabling technologies, such as smart meters, controllable appliances and home energy management systems, together with low awareness and behavioural inertia, mean that much of the demand flexibility in the buildings sector has yet to be mobilised. In addition, in many markets consumers are not directly exposed, or only partially, to short-term wholesale price signals and there are regulatory barriers that restrict aggregation of small loads to limit participation in DR. This will require more attractive financial business cases for households as well as acceptance of automated control and incentives for utilities and aggregators to invest in the necessary infrastructure and capabilities to enable participation in DR programmes.

In the transport sector, utilised DR capacity remains comparatively small at less than 5 GW. However, the EV fleet in many regions is expanding rapidly, and if managed through smart charging systems that can respond automatically to price signals or external controls, the transport sector could become an important provider of demand response.

Battery storage has become one of the most versatile tools for providing short-term power system flexibility. They can support the integration of wind and solar power by responding quickly to provide system balancing and grid support services, contribute to security of supply through capacity provision, and shift renewable generation to periods of high demand. As a result, they help defer or reduce the need for some network upgrades. Battery storage can also help greatly with the secure and cost-effective integration of new types of loads such as EVs, heat pumps and data centres, where the consumption can be highly correlated across location and time.  

Costs have declined significantly in recent years, with battery storage project costs falling by about 40% in 2024 to around USD 150/kWh, underpinning a strong increase in deployment. In 2024, utility-scale battery storage additions reached 63 GW, marking another record year and bringing total installed capacity to 124 GW. The contribution of utility-scale batteries in meeting peak demand is increasing in many power systems. In California, the ratio of installed utility-scale battery storage to peak load stood at almost 25% in 2024, and around 15% in South Australia and the United Kingdom, which were all less than 5% in 2019.

Large project pipelines for utility-scale batteries reflect strong investor interest, but many projects face multi-year delays in securing grid connection and permitting, including planning approval processes and addressing local opposition related to issues such as fire safety. Projects also encounter uncertain and volatile revenue streams, or struggle to access financing on suitable terms.

Global battery storage additions in the power sector surged in 2024, with around 63 GW of new utility-scale capacity installed. A key driver of this expansion has been the rapid decline in lithium-ion battery pack costs. In 2024, average pack prices fell by around 20%, followed by a further decrease of about 8% in 2025.

Average capacity of utility-scale battery storage has been increasing in many markets, in line with the evolving market needs for flexibility and capacity as VRE penetration increases. In the United States, average capacity more than doubled from about 15 MW in 2021 to around 35 MW in 2024. This trend is also seen in Europe, where a 1 000 MW (4 000 MWh) battery storage system is being built by LEAG and Fluence in Germany, set to become the largest battery storage project on the continent. In Australia, the Eraring battery storage system is set to expand from a capacity of 460 MW to 700 MW (and a volume of 2 800 MWh) in 2027, which will make it the country’s largest utility-scale project. Saudi Arabia commissioned a 500 MW (2 000 MWh) battery storage facility in Bisha in 2025, adding to the growing number of large-scale projects worldwide.

China has become the largest market globally for battery storage, both in terms of annual additions and cumulative installed capacity. China added about 42 GW (101 GWh) of new-type energy storage capacity1 in 2024, corresponding to an average duration of around 2.3 hours. By end-2024, Inner Mongolia was the leading province in China, with around 10.2 GW of new-type energy storage. About 66 GW (189 GWh) was also added in 2025, bringing the cumulative capacity to approximately 145 GW by the end of the year. By the end of 2025, standalone energy storage accounted for 58% of cumulative installations.

The rapid declines in costs are particularly apparent in batteries with higher storage durations. In Great Britain, 98% of new capacity added in Q4 2024 had a duration of 2 hours or more. In Australia’s National Electricity Market, 95% of capacity post-2024 is designed for 2 hours or more, and is expected to increase average duration from 1.5 hours in 2024 to 2.5 hours in 2027. In China, the average duration of cumulative new-type energy storage installations increased from about 2.1 hours in 2021 to around 2.6 hours in 2025. California stands out for its concentration of battery storage with a 4-hour duration, in part due to Resource Adequacy rules that assign capacity value based on sustained output over this period. This reflects a trend in some markets, where new battery projects are being built with longer durations to provide greater energy shifting potential.

Beyond these market-driven trends, several countries are introducing state-backed procurement schemes to support longer-duration battery storage. In Italy, the first MACSE auction has already contracted 10 GWh of utility-scale battery storage for delivery in 2028, including about 1.3 GWh from battery storage with durations of 8 hours or more. Italy's auction mechanism aims to procure 50 GWh of BESS capacity by 2030 to support renewable energy integration. Great Britain has introduced a long-duration energy storage cap and floor scheme requiring projects to discharge at full power for at least eight hours, with an indicative capacity range of 2.7 to 7.7 GWh by 2035.

Excess electricity generation (also referred to as incompressibility or overproduction) amid high VRE output in low-demand periods, more common during holidays and weekends in off-peak months with longer daytime, can pose flexibility challenges with the need for additional downward flexibility. This is because electricity supply and demand needs to be continuously balanced for system stability. In some cases, such conditions may lead to reverse power flows from distribution grids to transmission grids, congestion and local overvoltage issues, requiring adequate planning and corresponding operational measures.

Smaller systems with the ability to export a higher share of their generation to other regions regularly see very low and even negative net load levels (e.g. Denmark, South Australia). Whereas net load levels in larger systems with lower shares of power exports are typically higher (e.g. Brazil, the United Kingdom, Germany), though negative net loads may also be observed (as in the example of Germany).

Greater supply‑side flexibility, more flexible loads, as well as increased storage for energy shifting and for ancillary services can help manage low grid load events. More widescale capability to control distributed PV and the ability to curtail them in the case of contingencies is another important lever. As the share of synchronous generators declines in the electricity mix, system strength can be reinforced by various technical measures such as deploying grid‑forming inverters, and by installing synchronous condensers.