The Wasted Megawatt-Hour That Wasn’t Wasted
Every year, grid operators across the world deliberately instruct wind farms and solar plants to stop generating. In California alone, curtailed renewable energy reached approximately 2.4 million MWh in 2022 — enough to supply hundreds of thousands of homes. In China, curtailment rates in certain provinces have periodically exceeded 20% of available renewable output. The renewable energy curtailment reasons are not what the headline numbers suggest. On the surface, these figures appear to indict the entire project of grid decarbonisation. Why build renewable capacity only to switch it off?
The answer exposes a fault line in how the energy transition is understood — not just by the public, but often by policymakers and industry observers. Curtailment needs to be split into two mechanisms, because the two are driven by entirely different causes. Conflating them leads to bad policy, misallocated capital, and a fundamental misunderstanding of how power systems actually work.
The Two Curtailments: Wires and Markets
Renewable curtailment — the deliberate reduction of output from wind or solar generation that could otherwise have been produced — falls into two distinct categories. The distinction matters more than most regulatory discussions acknowledge.
Transmission-constrained curtailment occurs when generation exceeds the physical capacity of the wires connecting supply to demand. A wind farm in West Texas may have abundant resource availability, but if the transmission lines to Dallas and Houston are saturated, the operator has no choice but to reduce output. This is a physical constraint — the electrons literally have nowhere to go without compromising system stability or thermal limits on conductors.
Economic curtailment is different. It occurs when market prices fall below a generator’s variable operating costs — or, in the case of zero-marginal-cost renewables, when prices turn negative enough that it becomes cheaper for the system to pay generators to stop producing than to continue. This is not a physical constraint. It is a market signal indicating that at a particular moment, supply exceeds demand at any positive price.
The distinction matters. Constraint-based curtailment should trouble grid planners because it reveals that the network cannot deliver available generation to load — a physical shortfall no amount of market design can fix. Economic curtailment, in contrast, is in many circumstances a sign that markets are functioning exactly as designed: when supply exceeds demand at any positive price, paying generators to back down is the efficient outcome.
When the Wires Can’t Carry It
Constraint-based curtailment is fundamentally an infrastructure failure — though not necessarily one that justifies immediate investment. Transmission networks were overwhelmingly designed around large, centralised thermal plants located near population centres or fuel sources. Wind and solar resources are geographically dispersed and often remote: the best wind regimes in the United States run through the Great Plains, far from coastal load centres. China’s richest solar resources lie in the western provinces of Xinjiang and Gansu, thousands of kilometres from the manufacturing hubs of the eastern seaboard.
The consequence is what grid operators call congestion — a condition where the least-cost generation cannot physically reach demand. When this happens, system operators must either curtail the remote renewable generation or dispatch more expensive, closer-in resources. Either outcome raises system costs. Either outcome wastes something. The difference is that congestion represents a structural mismatch between generation location and transmission capacity, one that persists across hours and seasons rather than resolving with market prices.
What makes constraint-based curtailment particularly frustrating is its predictability. Network operators can typically forecast congestion hours or days in advance. The constraint is the lack of infrastructure to act on that knowledge. Building new transmission, however, involves permitting timelines that stretch across years, if not decades, and capital costs that run into the billions. The result is a persistent gap between where clean energy exists and where it can be consumed.
When Too Much Generation Is the Problem
Economic curtailment paints a different picture. On a sunny spring Sunday in California, solar generation can easily exceed total system demand during midday hours. With no fuel cost to save, solar plants have every incentive to bid into the market at near-zero prices. As more solar enters the supply stack, wholesale prices collapse — sometimes below zero. At that point, some generators choose to curtail rather than pay to remain online.
This is not a malfunction. It is the market communicating that, during those specific hours, additional generation has zero or negative value to the system. The alternative — keeping every solar plant running regardless — would require either paying generators to produce unwanted energy or forcing other plants offline in ways that create reliability risks. Neither approach would be economically efficient.
The International Energy Agency has noted that some level of curtailment is inherent to systems with high renewable penetration. As variable renewable energy shares rise, periods of oversupply become more frequent. Curtailment serves as a pressure release valve — one that avoids more costly interventions such as forced thermal plant cycling or voltage excursions that could cascade into wider instability.
The Counterintuitive Efficiency of Wasted Megawatt-Hours
Here is the claim that most unsettles casual observers: building enough transmission to eliminate all curtailment would almost certainly be a poor investment.
Transmission infrastructure is expensive, and its value depends on utilisation. A line designed to carry every possible megawatt-hour from a remote wind farm — including the infrequent peaks that occur only during the windiest hours of the windiest months — would spend most of its operating life significantly underutilised. The incremental cost of capturing those last few percentage points of generation would exceed the value of the energy recovered.
This principle is well-established in generation planning but less intuitively grasped in transmission. No one expects a peaking gas plant to run 8,760 hours per year; its economics depend on running during the highest-value hours. Similarly, a transmission line should be sized for the value it delivers across typical conditions, not for the extreme tail of renewable output that occurs only rarely. Some level of curtailment is therefore economically optimal — a feature of rational system design rather than evidence of failure.
Regions that have accepted this have generally made better investment decisions. Texas, through its Competitive Renewable Energy Zones programme, built transmission to connect West Texas wind to load centres — but the design accepted that some curtailment would persist during the highest-output periods. The alternative would have required substantially larger investment for diminishing returns.
Why Different Markets Reach Different Answers
The acceptable level of curtailment is not universal. It depends on geography, market structure, and political priorities — which helps explain why countries approach the same technical problem so differently.
China has historically experienced some of the world’s highest curtailment rates, driven overwhelmingly by transmission constraints. The physical distance between renewable resources and demand centres is enormous. More importantly, the institutional separation between generation developers, grid operators, and provincial governments has often meant that wind and solar capacity was built faster than the transmission infrastructure required to deliver it. Curtailment rates in Gansu and Xinjiang exceeded 30% during certain periods in the mid-2010s, reflecting a planning failure rather than an economic optimisation.
Germany faces a different problem. Its transmission constraints run north-to-south: wind generation is concentrated in the northern Länder, while industrial demand clusters in Bavaria and Baden-Württemberg. The north-south transmission corridors have been delayed by permitting and public opposition, creating persistent congestion. Curtailment costs — both for renewable generators compensated for lost output and for redispatch measures — run into the hundreds of millions of euros annually.
California illustrates the economic variant. The state’s rapid solar buildout has created a now-famous “duck curve” — mid-day net load that plunges as solar floods the system, followed by a steep ramp as the sun sets. Economic curtailment during spring months, when demand is moderate and solar output high, has become a structural feature of the CAISO market. The state has responded by investing in battery storage and regional coordination, implicitly accepting that some level of oversupply curtailment will persist.
These regional differences underscore a point often lost in energy transition debates: the same megawatt-hour of curtailed energy can reflect entirely different underlying problems depending on where and why it occurred. A high curtailment figure in China’s western provinces may indicate planning inefficiency. A high curtailment figure in California may indicate a market working correctly to manage oversupply.
When Curtailment Becomes a Policy Signal
The line between acceptable and problematic curtailment shifts with time and technology. What makes curtailment a useful diagnostic tool is that it reveals where system constraints are binding — and therefore where investment, whether in transmission, storage, or demand-side flexibility, would generate the highest returns.
Persistent, geographically concentrated curtailment typically signals transmission inadequacy. If a particular wind resource zone experiences curtailment rates above 10–15% year after year, the economic case for additional transmission capacity becomes increasingly difficult to ignore. Brief, widespread curtailment during low-demand periods, by contrast, often points toward storage or demand response as more cost-effective solutions than new wires.
This diagnostic value is lost when curtailment data is aggregated into national averages. A country reporting 5% total curtailment may have regions at 25% and others near zero. The average obscures the investment signal. Regulators and system planners who treat curtailment as an embarrassment to be minimised — rather than as information to be interpreted — routinely misallocate resources.
What Should Concern the Industry
Two developments merit close attention. The first is the interaction between curtailment and interconnection queues. In markets where generators face long delays for transmission interconnection, developers may site projects in areas where curtailment risk is already elevated — not because the resource is ideal, but because the queue is shorter. This dynamic can concentrate curtailment geographically in ways that lock in inefficiency.
The second is the treatment of curtailed energy in renewable energy certificates and carbon accounting. When a wind farm is curtailed, the associated environmental attributes are not generated. If policy targets are expressed in terms of installed capacity rather than delivered energy, curtailment creates a gap between nominal and actual decarbonisation. This gap can be substantial in high-curtailment regions and deserves greater scrutiny in policy design.
Neither of these concerns invalidates the basic point: some curtailment is rational, and pursuing zero curtailment would impose costs far exceeding its benefits. But they do suggest that the rules governing how curtailment is measured, reported, and priced have consequences that extend well beyond operational inconvenience.
Looking Ahead
Curtailment is unlikely to disappear from high-renewable systems. As variable generation penetrations rise toward 50%, 60%, and beyond — levels already observed in South Australia and parts of Europe — periods of oversupply will become more frequent, not less. The question is not whether curtailment will occur, but whether the system can manage it efficiently and whether the associated costs fall equitably.
Several trends will shape the answer. Energy storage, particularly battery systems with four to eight hours of duration, can absorb oversupply during peak solar hours and shift that energy into evening demand periods — directly reducing economic curtailment. Regional market integration enlarges the geographic footprint over which supply and demand can balance, turning what would be local oversupply into export opportunity. Demand flexibility, from electric vehicle charging to industrial processes, can shift consumption into periods that would otherwise require curtailment.
Each of these solutions has its own cost curve, its own implementation timeline, and its own technical limitations. None will eliminate curtailment entirely. The systems that manage the transition most cost-effectively will be those that distinguish clearly between the two types of curtailment, invest where the returns justify it, and accept — without embarrassment — that some wasted megawatt-hours were never really wasted at all.
References
- International Energy Agency — World Energy Outlook and renewables integration analysis, providing data on curtailment rates and the economic rationale for accepting some level of oversupply in high-VRE systems.
- CAISO — Annual curtailment reports and market data illustrating the distinction between economic and constraint-based curtailment in California’s solar-heavy system.
- National Renewable Energy Laboratory (NREL) — Research on the economic efficiency of curtailment thresholds, including modelling that demonstrates the diminishing returns of transmission investments designed to capture extreme renewable output events.
- China National Energy Administration — Provincial curtailment statistics and the documented gap between renewable capacity additions and transmission infrastructure in western provinces.
- ENTSO-E — European transmission system data on north-south congestion in Germany and the associated redispatch and curtailment costs.
- ERCOT — Analysis of the Competitive Renewable Energy Zones programme as a case study in transmission investment that explicitly accepted residual curtailment as economically optimal.
