The need for modern grids

The grid is the critical enabling infrastructure on which the energy transition depends for security, decarbonisation and affordability, as economies around the world increasingly rely on electricity. Electrification of end uses is a core pillar of the strategy selected by many jurisdictions to strengthen their energy security and cut emissions, making electricity the backbone of the transition.

Demand for electricity is rising rapidly. Electricity demand has grown at roughly twice the rate of overall energy demand over the past decade, driven in part by electrification, data centres, cooling and industrial growth. This ratio continues to widen, while new renewable generation, often built far from demand centres, is adding further need for transmission. To meet these needs, 25 million kilometres1 of lines are expected to be added under the IEA’s Current Policy Scenario (CPS), equivalent to boosting grid capacity by 30% by 2035. This is why the current period is termed the Age of Electricity: electricity demand is decoupling from energy demand, and electricity is fast becoming the backbone of the energy system.

This transformation is also changing the conditions the grid has to cope with, not only how much it carries. Distributed energy resources such as rooftop solar, batteries, electric vehicles (EVs) and flexible demand have turned many grids from unidirectional systems into systems with dynamic two-way flows, while large inverter-based power sources and vast data centres2 represent a new, more concentrated form of generation and demand. Both trends are increasingly evident at the distribution level as much as at the transmission level, creating new challenges for voltage management, transformer loading, feeder capacity and protection settings, and both can produce sharp peaks that grids historically built for steady, predictable conditions are not designed to absorb without being actively managed.

Where the grid cannot keep pace, the consequences are well documented: lengthening connection queues, new housing and industrial development competing with data centres and generation projects for scarce grid capacity, and, in the most severe cases, service disruption. Left unaddressed, these circumstances risk slowing electrification plans just as they need to accelerate, with knock-on costs for competitiveness and climate goals.

Stability adds a further dimension to this complexity: operators need to ensure not only that flows stay within the thermal limits, but that the system retains enough strength to ride through disturbances. As converter-connected resources displace synchronous generators, several stabilising properties traditionally provided by synchronous machines, including inertia, voltage support, fault current contribution and damping (collectively known as system strength), come under pressure. Some of these, such as inertia, are inherently linked to rotating mass; others, such as voltage support and damping, can also be delivered by appropriately designed converter-connected resources or dedicated equipment. These variables are faster-moving and harder to observe and model than thermal limits, raising the bar for the quality of data and validation that grid operators need to manage them. This is particularly important as AI-based tools are increasingly used to support grid operations, because their performance depends heavily on the quality of the data and models they rely on.

Meeting this growth in demand will require higher grid capacity. The question this report addresses is how to achieve this capacity growth given that expansion, that is the buildout of new infrastructure, is struggling to keep pace.

New grid infrastructure is failing to keep pace with the growth in new supply and demand-side assets, even as record volumes of new supply and demand are being connected. Lengthy permitting procedures, supply-chain bottlenecks, labour shortages and rising costs all weigh on the pace of new lines.

This becomes evident in two ways. Connection queues are lengthening: at least 1 700 GW of renewables and 600 GW of utility-scale battery projects at an advanced stage3 were awaiting connection in 2025. And network congestion4 is becoming more frequent and costly for grid users: in 2024, grid congestion cost USD 12 billion in the United States and EUR 4.3 billion in the European Union, with Germany alone accounting for over half of the EU total. Curtailment of surplus renewables also carries an environmental cost as fossil-based production is not displaced; curtailment reached over 10 TWh in the European Union in 2024 (enough to power around three million homes for a year), while India recorded 2.3 TWh of curtailed solar generation in 2025.

Grid expansion alone is not the economically efficient answer. Building grids to eliminate all congestion would mean matching peak needs at all times, leading to low average grid usage and uneconomical outcomes; given the low load factor of many renewable and demand assets, a certain level of congestion is in fact efficient. As a result, a growing share of total grid spending will go towards the tools and resources required to dynamically operate the system, meaning that how well the existing network is used has become just as important as how much new capacity it is built.