Curtailment mitigation

Curtailment mitigation is the use of storage, transmission and market reform to capture renewable output the system would otherwise refuse. Curtailment itself, in the IEA's definition, occurs when the power system cannot absorb all generated power because of transmission capacity limitations, system stability requirements or supply-demand imbalances.

A battery mitigates it by charging during the surplus and discharging hours later when the network can take the energy — EIA describes CAISO batteries charging on excess midday solar and discharging as the sun goes down.

Read this alongside clipping and curtailment, which covers the loss mechanism itself; this entry covers what you do about it. The trap is that mitigation is easy to describe and hard to quantify: as of mid-2026 no TSO or regulator figure of the form "storage cut curtailment by X%" could be sourced.

Reviewed July 2026 by Sergey Syrvachev

Learn it in the Grid-Scale BESS course All BESS terms

New to BESS? Start free with the 7-email fundamentals course — no cost, no account.

What it is (precise)

Start with the thing being mitigated. The IEA states that curtailment occurs when the power system cannot absorb all generated power because of transmission capacity limitations, system stability requirements or supply-demand imbalances, and that while some curtailment is expected and inevitable, persistent or widespread curtailment often highlights gaps in planning, flexibility or infrastructure.

That last clause matters commercially: a site with structural curtailment is telling you something about the network it sits on, not just about one bad year. Reducing it, the IEA says, requires a comprehensive strategy involving transmission, flexibility and coordinated system planning. Storage is one leg of that, not the whole answer.

Then split the category, because the two halves behave differently. The IEA distinguishes technical curtailment — dispatching down renewable energy for network or system reasons — from dispatch-down driven by economic or market conditions, which its technical-curtailment data series excludes.

EIA describes the same split from the delivery side: output is reduced either through price signals or, rarely, through an order to reduce output, during periods of congestion, when power lines do not have enough capacity to deliver available energy, or oversupply, when generation exceeds customer demand. An instructed reduction is a network problem; a price-driven reduction is a market signal, and a battery monetises the second far more cleanly.

Storage mitigates curtailment by moving energy in time rather than discarding it. EIA's account of the CAISO case is the plain-English version: battery storage allows some renewable energy to be stored and used four to eight hours later in the day, charging on excess solar at midday and then discharging when the sun is going down, providing electricity during the hours it is most needed.

Note the boundary of that claim. Intra-day shifting works. EIA is explicit that in spring more solar energy is often produced than can be used within a day, and that without more transmission capacity or a long-term storage solution, high curtailments can still occur.

Why it matters in a real grid-scale project

The volumes are large enough to be worth modelling. In 2024 CAISO curtailed 3.4 million MWh of utility-scale wind and solar output, a 29% increase on the amount curtailed in 2023, with solar accounting for 93% of the total. That is the gross curtailed volume in one balancing area in one year — an upper bound on what all storage everywhere could theoretically absorb, not an addressable pool for one battery.

It is also a moving target in both directions: curtailment grows with variable renewable build-out and shrinks as flexibility and transmission are added. A business case that assumes a static curtailment volume across a 20-year term is assuming the network stops changing, which it will not.

The build-out response has been fast, and the two published capacity numbers do not agree. EIA's survey of recent and planned capacity changes puts CAISO battery capacity up 45% in 2024, from 8.0 GW to 11.6 GW.

CAISO's own Department of Market Monitoring counts 13,000 MW in the balancing area in December 2024, up from about 500 MW in 2020, with over half of that capacity physically paired with solar or wind — either sharing a point of interconnection under the co-located model, or registered as a single hybrid resource. The counting bases differ. Quote whichever figure you use with its source attached, and never present either as the number.

The mitigation shows up as load. CAISO's market monitor reports that from hours-ending 10 to 13, battery charging represented around 14.7% of load in the CAISO balancing area in 2024, and that during those hours batteries help reduce the need to curtail or export surplus solar energy at very low prices.

Batteries then provided an average of about 8.6% of the balancing area's energy in hours 17 to 21. The revenue side is less comfortable: net market revenue for batteries fell from about $78/kW-yr in 2023 to $53/kW-yr in 2024, which the monitor attributes largely to lower peak energy prices and lower loads.

Both lines rose in the same system in the same year — and as of mid-2026 no TSO or regulator publishes a figure of the form “storage reduced curtailment by X%”. The quantity this figure cannot draw is the point.

Neither bar explains the other, and the arithmetic does not run either way. Recapture is bounded by duration (a four-to-eight-hour shift), by permission, and by price — the curtailed-MWh total is an upper bound on what all storage everywhere could absorb, not an addressable pool for one battery. The value is a spread net of round-trip efficiency, not a count of recovered MWh: batteries charged about 14.7% of CAISO balancing-area load in hours-ending 10 to 13, precisely when that energy was worth very little. A widely circulated claim that curtailment would have been 67% higher without storage traces to no CAISO or EIA primary source.

Key facts

Curtailment (IEA definition): System cannot absorb all generated power: transmission limits, stability requirements, supply-demand imbalance
Two triggers (EIA): Congestion (lines lack delivery capacity) and oversupply (generation exceeds demand)
Technical vs economic: Technical = dispatch-down for network/system reasons; economic/market dispatch-down is counted separately (IEA)
CAISO curtailment, 2024: 3.4 million MWh of utility-scale wind and solar, +29% on 2023; solar was 93% of it
CAISO storage growth, 2024: 8.0 GW to 11.6 GW (+45%) per EIA survey; CAISO DMM counts 13,000 MW in December 2024
Batteries as midday load: About 14.7% of CAISO balancing-area load in hours-ending 10-13 during 2024 (CAISO DMM)
Time shift achieved: Charge on midday surplus, discharge four to eight hours later (EIA)
China curtailment: Volumes +55% in 2024; rates 4.1% wind and 3.2% solar PV; expected to stabilise around 5-6% (IEA)
The number that does not exist: As of mid-2026, no sourceable TSO or regulator figure for 'storage reduced curtailment by X%'

Typical values and standards

Outside CAISO, the IEA's China numbers are the most quotable. Renewable curtailment volumes increased roughly 55% in 2024, reaching 4.1% for wind and 3.2% for solar PV, and the IEA expects the figure to stabilise at around 5-6% thanks to expanding HVDC transmission infrastructure and more utility-scale and behind-the-meter battery storage.

Keep the units straight when repeating this: the 55% is a change in volume, while 4.1%, 3.2% and 5-6% are rates. The IEA also credits market reform — requiring wind and solar plants to participate in wholesale markets — with improving dispatch efficiency and limiting curtailment growth.

There is no standard figure for how much curtailment storage removes, and you should be suspicious of anyone offering one. As of mid-2026, no authoritative statement of the form "storage reduced curtailment by X%" from a transmission system operator or a regulator could be sourced.

The closest sourced quantities are indirect: CAISO's market monitor reporting 14.7% of balancing-area load charged by batteries in hours-ending 10 to 13, and the IEA attributing China's expected stabilisation at 5-6% partly to battery storage. A widely circulated claim that CAISO curtailment would have been 67% higher without storage appears in trade press but traces to no CAISO or EIA primary source, so keep it out of a board pack.

No standard governs curtailment mitigation as such either; it is a dispatch and market outcome, not a product with a test method. What is governed sits around it — the interconnection agreement that defines your export limit and any curtailment obligation, the market rules that decide whether a reduction is instructed or price-driven, and the metering that determines whose energy was curtailed in a co-located configuration.

Note also that curtailment costs sit outside the headline cost benchmarks: Lazard states that congestion, curtailment and other integration-related costs have not been examined within the scope of its levelized cost analyses.

How it shows up in specs, studies and contracts

In a co-located project the first contractual question is whose energy the battery is charging with. CAISO's market monitor notes that over half of the storage capacity in its balancing area is physically paired with solar or wind, either sharing a point of interconnection under the co-located model or registered as a single hybrid resource.

Those two structures settle and meter differently. Get the charging-source definition written down — grid charging, on-site renewable charging, or either — because offtake terms, incentive eligibility and the whole curtailment-recapture calculation hang off it, and a shared point of interconnection means the battery's export competes with the generator's for the same headroom.

In the energy yield study, curtailment appears as a loss line, and the modelling basis has to be stated. Ask whether the study models technical curtailment — dispatching down for network or system reasons, in the IEA's terms — or the economic, price-driven dispatch-down, or both, because published curtailment series often exclude the economic half.

Ask what year's network topology is assumed and whether queued projects upstream are in it. And ask what the battery is permitted to do during an event: where the interconnection agreement caps site export, charging the battery from the co-located array is often allowed when exporting is not, and that asymmetry is where the value lives.

Quantifying that value means pricing the shifted energy, not counting recovered MWh. The energy you recapture was worth very little at the moment it would have been curtailed — CAISO's monitor describes batteries reducing the need to curtail or export surplus solar at very low prices — and it is worth whatever the evening market pays when you release it.

The mitigation value is therefore a spread, net of round-trip efficiency, exactly as in energy arbitrage. Model it against hourly prices at the specific node rather than an annual average, and carry the fact that net market revenue for CAISO batteries fell from about $78/kW-yr to $53/kW-yr between 2023 and 2024.

Common pitfalls

The first pitfall is conflating the two loss mechanisms. Clipping is an inverter or interconnection limit inside your own plant; curtailment is the system telling you to stop, through a dispatch instruction or through price.

A battery on the DC side can recapture clipped energy, but recapturing curtailed energy also requires that charging is permitted during the event and that the site has somewhere to put the energy. The two also fail differently: clipping is predictable from the DC/AC ratio and the irradiance profile, while curtailment depends on network state and market outcomes that appear on no datasheet.

The second is treating curtailment volumes as a durable revenue pool. The IEA's own framing is that persistent or widespread curtailment highlights gaps in planning, flexibility or infrastructure, and that reducing it requires transmission, flexibility and coordinated system planning together.

Those fixes are being built: China's curtailment rate is expected to stabilise at around 5-6% partly because of expanding HVDC transmission, and EIA notes CAISO promoting the addition of flexible resources for the same reason. If your business case depends on curtailment staying high, you are betting against the system operator's stated plan. That is a legitimate bet, but state it as one.

The third is the duration mismatch. Intra-day shifting is what a 2- to 4-hour battery does, and EIA describes exactly that: storing renewable energy and releasing it four to eight hours later in the day.

Spring oversupply is a different problem, and EIA is direct that in spring more solar energy is often produced than can be used within a day, so high curtailments can still occur without more transmission capacity or a long-term storage solution. Sizing a battery on annual curtailed MWh, without checking how that energy is distributed across days and seasons, overstates what a 4-hour asset can actually recapture.

Sources & standards

Common misconception

Pair a battery with a curtailed renewable plant and you recover the curtailed energy, so the annual curtailed MWh is the size of the prize.

In reality: Recapture is bounded by duration, by permission and by price. EIA describes intra-day shifting of four to eight hours and states that in spring more solar is often produced than can be used within a day, so high curtailment persists without more transmission or a long-term storage solution. The recovered energy is also worth only the spread between the near-zero price at the moment of curtailment and the price when you release it, net of round-trip efficiency — not the full market value of the MWh. And as of mid-2026 there is no sourceable TSO or regulator figure quantifying how much curtailment storage actually removes.

Visuals & further reading

Clipping / curtailment Glossary

Go deeper