Grid congestion: why Europe pays billions to move power it already has

In 2025, German grid operators curtailed 9,379 GWh of renewable electricity because the lines could not take it, about 3.5% of all renewable output that year. On the other side of the bottleneck they paid conventional plants to run. The total bill for managing congestion came to €3.071 billion, according to the Bundesnetzagentur’s figures published on SMARD on 30 March 2026. That is the price of a transmission network laid out around big power stations near the southern and western load centres, which now has to carry wind from the north coast.

What congestion actually is

European power markets mostly treat each country, or each bidding zone, as a copper plate: inside the zone, any generator can sell to any buyer as if the wires were infinite. The physical grid does not agree. Electricity follows the path of least impedance, not the contract, and a sale from a Schleswig-Holstein wind farm to a factory in Bavaria loads every line between them in proportion to how the network is built.

When the market’s result would push a line, cable or transformer beyond its thermal limit, or would leave the system unable to survive the loss of one element (the N-1 rule), the grid is congested. The power exists, the buyer exists, and the wires in between cannot carry the trade.

Congestion comes in two shapes. Transmission congestion is the German case: bulk power cannot flow from where it is made to where it is used. Distribution congestion is the Dutch case, covered below: local networks and substations are simply full, so new users cannot connect at all.

Redispatch and what it costs

The operator’s first tool is redispatch. After the market clears, the operator orders specific plants to change output: down behind the bottleneck, up in front of it. The trade still happens on paper; the physics is rearranged underneath it, and the generators are compensated.

Germany publishes the most detailed numbers in Europe. For 2025:

Item	2025	2024
Total congestion management cost	€3,071 million	€2,954 million
Redispatch with conventional plants	€1,176 million	€1,233 million
Compensation for curtailed renewables	€433 million	€554 million
Grid reserve power plants	€1,359 million	€1,068 million
Countertrading	€102 million	4% lower than 2025
Redispatch volume	30,319 GWh	30,318 GWh
Curtailed renewable energy	9,379 GWh	9,389 GWh

Two things stand out. Volumes were flat, but the cost of the reserve plants kept on standby for the grid rose by more than a quarter. And the mix of curtailment shifted: offshore wind curtailment fell by 27%, while curtailed solar almost doubled to 2,704 GWh as rooftop and field installations kept arriving faster than the local networks.

Across the EU, ACER’s 2025 monitoring report puts the cost of remedial actions at €4.3 billion in 2024, for 60 TWh of redispatch and countertrading, about the annual electricity consumption of Austria. The same report found that operators in the Core region, which spans most of central Europe, made only 54% of capacity on their most congested lines available for cross-border trade in 2024, against a legal minimum of 70% from the end of 2025. Congestion inside countries therefore also shrinks the space for trade between them.

Borders congest too, and there the cost shows up as price gaps instead of redispatch bills. By Red Eléctrica’s count, two hours in every three on the Spain–France border were congested during 2024 (67.6%), and Iberian prices regularly split from French ones as a result; the guide to the April 2025 Iberian blackout looks at what that thin connection meant on a bad day. Every one of the cross-border links in the directory of European power links exists, in part, to narrow one of these gaps.

Countertrading is the cross-border cousin of redispatch: an operator buys power on one side of a border and sells it on the other to relieve a line, without naming specific plants.

Curtailment: the renewable side of the bill

Curtailment is the part of redispatch that the public notices. A wind farm is told to stop spinning on a windy day, and its owner is paid for the energy it did not produce. In Germany in 2025 that compensation was €433 million, down 22% on 2024 despite almost identical curtailed volumes.

It is tempting to call curtailment waste, and in part it is. But a small share is also cheaper than building a line sized for the windiest hour of the year. The problem in Germany is not that 3.5% is lost; it is that the share is concentrated on the same northern nodes year after year, because the large north-south lines are still under construction.

The Dutch congestion maps

The Netherlands has turned congestion into a map. The national capacity map run by Netbeheer Nederland, the association of Dutch grid operators, colours each area by whether there is transport capacity left for taking power from the grid and, separately, for feeding it in. Look up a business park in Noord-Holland and the answer for a new large connection is likely to be no.

The waiting lists are long. In a letter to parliament on 4 February 2026, the climate and green growth minister gave the figures: 14,044 requests for offtake capacity with regional operators, totalling 9 GW, and 212 requests with the national operator TenneT, totalling 38 GW. TenneT has said that in a large part of Noord-Holland no extra space can be made for large consumers for at least a decade. The eight measures in the letter lean heavily on flexibility: higher risk acceptance by operators, better forecasts, revised flexible contracts, regional tenders for flexibility, negotiations with the 50 largest users, flexible connections as the standard for batteries and charging, and around €500 million a year of contracted flexibility. The government expects them to free 5–10 GW by 2030.

This is distribution congestion, and it hurts differently from the German kind. It does not show up as a redispatch bill. It shows up as a heat pump factory that cannot expand, a bus depot that cannot electrify, or a solar park that is built but cannot feed in.

Remedies

New lines and cables

The structural answer is more transmission. Germany’s SuedLink, two 2 GW, 525 kV underground DC cables running about 700 km from Schleswig-Holstein to Baden-Württemberg and Bavaria, was fully permitted in October 2025 and is due to connect in 2028. It is built as HVDC because a controllable DC link can push a set amount of power south regardless of how the AC network around it would share the flow. Cross-border links such as NordLink help too, by giving northern German wind an outlet to Norwegian hydro storage. Offshore, the plan is to stop running each wind farm’s cable to the nearest beach and to land power where the grid can take it, the idea behind the hubs described in the guide to energy islands.

The trouble with new lines is time. A line takes a decade from plan to power, and the German redispatch bill is the price of the years in between.

Dynamic line rating

The cheapest megawatt is one that the existing line can already carry. Overhead lines are rated for a conservative assumption, typically a hot summer day with no wind, because the limit is how far the conductor sags as it heats. On a cold, windy day the same conductor stays much cooler and can carry more current safely. Dynamic line rating uses sensors on the line and weather forecasts to set the limit hour by hour.

The BEST PATHS project worked on exactly this. Its demonstration on repowering AC corridors, led by 50Hertz, built a prototype DLR system based on low-cost sensors, alongside research on high-temperature low-sag conductors that can run hotter without sagging further. Both approaches get more out of an existing right of way, which matters because a permit for a new corridor is the slowest item in any grid plan.

DLR has limits. It gains most on windy days, which often coincide with the wind-driven congestion it is meant to relieve, but it does nothing for cables, transformers or substations, and it cannot turn a structural shortfall into a surplus.

Storage and flexible connections

Batteries placed on the right side of a bottleneck can soak up power that would otherwise be curtailed and release it later. Flexible connections do something similar on the demand side: a user accepts a contract that lets the operator cut or cap its load in the few congested hours, in exchange for being connected years earlier. The Dutch plan relies on both. The catch is that a battery only relieves congestion if it is placed and operated for the grid; one that simply chases wholesale prices can make a local bottleneck worse.

Market design

A final remedy is to price the physics: smaller bidding zones or locational prices, so that a megawatt behind a bottleneck is worth less than one in front of it. It is the most contested option in Europe and the one most likely to be decided by politics, not engineering.

The European Commission’s European Grids Package, presented on 10 December 2025, proposes faster permitting and better cross-border cost sharing for new lines. Whether it shortens the ten-year wait for SuedLink’s successors will be visible first in the German figures for 2027, the first full year before the cables come online.