Why Does Clean Electricity Get Switched Off? Renewable Curtailment Explained

Renewable energy curtailment explained with solar, wind, transmission, battery storage, and flexible demand

A bright, windy day should be perfect for renewable energy.

So why would a grid operator ever tell a solar farm or wind farm to produce less electricity?

A recent KBS News report used Jeju Island to show this paradox. The same problem now appears in California, Chile, Germany, and Great Britain.

The problem is not too much clean energy. It is too little flexibility around it.

Quick Answer

Renewable curtailment happens when solar or wind could produce more electricity, but the power system cannot safely or economically use all of it at that moment. The bottleneck may be demand, transmission, storage, other generators, or several of these at once.

First, What Does “Curtailment” Mean?

Curtailment is often described as “throwing away electricity.” Usually, the electricity is never generated. A solar inverter reduces output. A wind turbine changes how it operates. The sunlight or wind is still there, but some of its potential electricity is left unused.

Four-step explanation of how renewable curtailment happens

Figure 1. Strong renewable output becomes curtailment when demand, grid capacity, storage, or other system flexibility cannot absorb it.

A simple example makes the problem easier to see.

At noon	Available power
Electricity demand	500 MW
Other generation still operating	250 MW
Available solar and wind	400 MW
Total possible supply	650 MW

The system needs 500 MW, but 650 MW could be produced. If other generators cannot reduce output enough, the grid cannot export the surplus, and storage cannot absorb it, then some generation must be cut.

That is the core idea: building generators is only one part of building an energy system. The electricity also needs somewhere to go.

Jeju Was an Early Warning

Jeju makes the problem easy to see because an island has obvious physical limits. The island can have strong wind and solar output while local demand is low. High-voltage direct-current links connect Jeju with the mainland, but every cable has a capacity limit. Storage helps, but it also has limits.

The important development is that this is no longer only a Jeju story.

On May 10, 2026, the Korea Power Exchange said it curtailed non-centrally dispatched generators on the mainland from 10 a.m. to 5 p.m. It cited grid constraints and supply-demand constraints. Minimum demand that day was 38,474 MW at noon, when solar output tends to be strong.^[1]

KPX reported another mainland curtailment event on May 30 under similar spring conditions.^[2]

Jeju was not a strange exception. It was an early warning of what happens when renewable generation grows faster than grid flexibility.

This is closely related to the broader grid bottleneck in the Electric Age. More electrification can create more clean energy, but it also puts more pressure on the network that connects supply and demand.

Renewable Power Can Arrive at the Wrong Time—or in the Wrong Place

Most curtailment stories can be understood with two questions:

Did the electricity arrive at the wrong time?
Did it arrive in the wrong place?

California: the wrong time

California made the “duck curve” famous. Solar generation rises around midday, pushing net load down. Then the sun sets, solar output falls, and the rest of the system must ramp up quickly for the evening.

The duck curve is not a blackout curve. It is a flexibility test: can the system store midday electricity, move it elsewhere, reduce other generation, or shift demand into the sunny hours?

In April 2026, California ISO reported about 1.42 TWh of wind and solar energy curtailed, with solar accounting for most of it.^[3]

Chile: the wrong place

The Atacama Desert has extraordinary solar resources, but the best place to generate electricity is not automatically the place where people and industry need it.

Chile's renewable-energy association ACERA said renewable curtailment exceeded 6 TWh in 2025, linking the problem to transmission, storage, and regulatory gaps.^[4]

A solar project can have excellent sunlight and low production costs, yet still lose value when the grid cannot accept its output during its most productive hours.

The same bottleneck can hit factories, cities, and AI data centers. Transformers, cables, substations, and construction can become the real constraint.

Germany and Great Britain: the cost of congestion

Germany and Great Britain show what happens after the mismatch appears: keeping the system balanced can become expensive.

In Germany, renewable redispatch reached about 9.379 TWh in 2025, equal to about 3.5% of total renewable generation. More than 96% of renewable generation still reached end users, but preliminary congestion-management costs reached about €3.071 billion.^[5]^[6]

In Great Britain, strong wind output in Scotland and northern Britain can exceed the ability of north-south transmission corridors to move electricity toward demand. In April 2026, NESO said balancing costs were running at about £2 billion a year, with more than 60% linked to thermal constraints.^[7]

The numbers differ, but the physics is the same: generation, demand, and network capacity do not always meet in the same place and hour.

Global comparison of renewable curtailment in South Korea, California, Chile, Germany, and Great Britain

Figure 2. Different accounting systems, similar physical pattern: renewable output, demand, and network capacity do not always meet in the same place and hour.

Why Not Just Turn Down Coal, Gas, or Nuclear Instead?

Sometimes they are turned down. But power plants do not all behave the same way.

Thermal generators can have minimum stable operating levels, ramp-rate limits, start-up times, and minimum up-or-down times. Some plants may also be needed for reserves or other reliability services.

NREL research found that these thermal-generator limits can strongly affect renewable curtailment at intermediate levels of solar penetration.^[8]

Solar and wind are also relatively easy to reduce quickly. But the actual dispatch choice depends on grid physics, market rules, contracts, reliability needs, and cost.

Is Curtailment Always Bad?

Not necessarily.

Trying to eliminate every last unit of curtailment could require enough transmission and storage to handle a few rare extreme hours. That infrastructure may sit underused for much of the year.

NREL calls this part of the curtailment paradox. In a high-renewable system, some curtailment can be economically rational, and curtailed renewable capacity can even help provide operating reserves.^[8]

A peer-reviewed global review reached a similar conclusion: the long-term goal should be optimal curtailment, not necessarily zero curtailment.^[9]

The key question is not “Was any clean energy curtailed?” It is “Has curtailment become frequent and large enough to show a structural problem in the grid?”

Three Ways to Give Clean Electricity Somewhere to Go

Once the problem is framed this way, the solutions become easier to remember.

1. Move electricity through space

Transmission lines, substations, transformers, and interconnectors connect renewable-rich regions with larger demand centers.

2. Move electricity through time

Batteries, pumped hydro, and other storage can absorb electricity when supply is high and release it later.

3. Move demand to the abundant hours

Flexible demand can shift EV charging, industrial processes, water heating, electrolyzers, and some computing work toward hours when electricity is abundant.

Power lines move electricity through space. Storage moves it through time. Flexible demand moves consumption toward the hours when electricity is abundant.

Why This Can Reach Your Electricity Bill

Curtailment sounds like a control-room problem, but the cost does not stay there. Someone pays for new transmission, batteries, and balancing actions. Developers also price curtailment risk into future projects.

On the other hand, periods of excess renewable output can create very cheap wholesale electricity. That can create a new opportunity for storage and flexible demand.

And This Is Where the Story Connects to AI

AI data centers are becoming some of the largest new electricity loads. They need high availability, so they cannot simply run only when solar is abundant.

But location, power contracts, batteries, and the scheduling of some non-urgent computing work may become more valuable as electricity supply varies by hour and region.

Could some of tomorrow's largest electricity users become part of the solution to renewable curtailment?

The answer will depend on reliability requirements, network location, pricing, storage, and how flexible the computing workload really is.

That is why the AI boom cannot be understood by counting GPUs alone. The physical infrastructure behind AI must also be built.

What to Watch Next

Curtailment rate: Is unused renewable potential growing faster than renewable generation?
Transmission completion: Are planned lines, substations, transformers, and interconnectors actually entering service?
Storage in operation: How many gigawatts and gigawatt-hours are working, not just announced?
Flexible demand: Are EVs, industry, heating, hydrogen, or computing shifting demand into renewable-rich hours?
Congestion cost: Are redispatch and balancing costs rising even when most renewable electricity still reaches customers?

The Bigger Lesson

The first phase of the renewable transition was about building more solar panels and wind turbines.

The next phase is harder: building a system that can use the electricity when and where it arrives.

California shows the time problem. Chile shows the location problem. Germany and Great Britain show the cost of congestion. Jeju shows that the same transition is now visible in Korea.

The question is no longer only how much clean electricity we can generate. It is how much we can actually move, store, and use.

Key Vocabulary & Phrases

Curtailment: A reduction in the electricity a generator could otherwise produce because the system cannot or should not accept all of it at that moment.
Grid congestion: A condition in which part of the power network cannot carry all the electricity that generators and consumers want to move through it.
Net load: The demand left for other grid resources after variable renewable generation is taken into account.
Redispatch: An action that changes generation in different locations to relieve a grid constraint.
Flexible demand: Electricity use that can move to another time without seriously reducing the service it provides.
System integration: The work of making generation, networks, storage, controls, and demand operate as one reliable system.

The Contexta

Search This Blog