The Duck Curve Problem

What the Duck Curve Actually Shows

In 2013, California's grid operator CAISO published a chart that became one of the most discussed graphs in the energy industry. It showed projected net electricity demand (total demand minus solar generation) across a typical spring day. The shape looked like a duck: a belly dipping down in the midday hours as solar production peaked, then a steeply rising neck as afternoon demand climbed while solar dropped off. The "duck curve" was born.

A decade later, the duck has grown. Solar output on the CAISO grid hit a record 23,896 MW on August 19, 2026 (CAISO key statistics), and the state's three largest utilities alone serve more than 1.5 million households with rooftop solar. On sunny spring afternoons when air conditioning use is still low, solar generation routinely exceeds what the state can use domestically. The midday dip in net demand has gone from a curve to a valley, and the evening ramp-up has become steeper and faster than almost anything in the history of grid management.

The numbers tell the story. In 2024, California recorded negative wholesale electricity prices on multiple sunny spring days. The grid was producing more power than it could consume or export to neighboring states. Solar farms were curtailed (told to shut off panels that were generating clean energy) because there was simply nowhere to put it.

The Two Problems the Duck Curve Creates

Midday Oversupply

When solar production exceeds load, grid operators have a few options: export power to neighboring states via transmission lines, curtail solar generation, or store it. Transmission export capacity is limited. Curtailment wastes clean energy that was generated at essentially zero marginal cost. Storage is the ideal solution, but battery capacity on the grid is still limited relative to the scale of the surplus.

Negative prices are a symptom of oversupply without enough storage or export capacity. In competitive wholesale markets, generators sometimes pay to keep running because stopping and restarting a gas plant is more expensive than paying to offload power. When solar is generating large surpluses, wholesale prices can go negative, meaning generators are paying the grid to take their power. California saw this happen multiple times in spring 2024.

The Steep Evening Ramp

The harder problem is the neck of the duck: the steep rise in net demand from late afternoon into the evening. Solar output drops as the sun sets, but residential demand rises sharply as people come home, cook dinner, run appliances, and charge devices. The steepest ramps come in spring and fall, when solar is strong and air conditioning is light: CAISO's record is 24,567 MW of added net demand over three hours, set March 16, 2026 starting at 3:50 p.m. That's an enormous amount of generation capacity that must be available on short notice.

The conventional answer has been gas-fired peaker plants: fast-ramping combustion turbines that can be online within minutes. They're expensive to operate (natural gas prices are volatile), run relatively few hours a year, and emit carbon dioxide every time they run. They're exactly the kind of generation the energy transition is supposed to eliminate. But without adequate storage, they're still needed to manage the evening ramp.

Why This Is a Sign of Success, and a Real Problem

It's worth being clear about what the duck curve represents. An oversupply of solar at midday means solar energy has become abundant enough to fully serve daytime load in a major grid region. That's a remarkable achievement. Ten years ago, that level of solar penetration was a distant projection, not a daily operational reality.

But success at one level of the energy system creates stress at another. The duck curve is fundamentally a symptom of the mismatch between when solar generates and when people use electricity. Solving it requires either changing when people use power, storing the midday surplus, or building more transmission to move surplus power to where it's needed. All three approaches are being pursued simultaneously, but none of them is free or fast.

Solutions: What's Actually Working

Time-of-Use Electricity Rates

Time-of-use (TOU) rates charge more for electricity during peak evening hours and less during midday. They create a price signal for customers to shift flexible loads (dishwashers, EV charging, pool pumps, laundry) to the middle of the day when solar is abundant.

California utilities have moved aggressively toward TOU as a default rate structure. The evidence that it shifts load is solid. The limitation is that many households have limited flexibility: they're at work during the day and can't shift their evening cooking, bathing, and homework. Automated smart home devices help, but not every household has them. Understanding TOU rates is increasingly important for any homeowner with solar or a battery.

Grid-Scale Batteries

Four-hour grid batteries are well matched to the duck curve problem. They charge from midday solar surplus and discharge into the evening ramp. California's battery fleet grew from less than 700 MW in 2019 to 21,112 MW by August 2026, including nearly 16,000 MW of utility-scale batteries in California and about 2,000 MW in Nevada and Arizona serving the CAISO grid. Those batteries charge at midday and discharge into the evening ramp, though EIA notes that in spring more solar is often produced than can be used within a day. See the full analysis of grid-scale battery storage for how those systems work.

Vehicle-to-Grid Technology

An electric vehicle with a 60–100 kWh battery is, in principle, a mobile storage unit. If that vehicle charges from cheap midday solar and can discharge back to the grid in the evening, it directly addresses both sides of the duck curve. Vehicle-to-grid (V2G) technology is still mostly in pilot programs, but Ford's F-150 Lightning, Nissan Leaf, and several other vehicles now support bidirectional charging. At scale, V2G could represent tens of gigawatts of flexible grid storage across millions of vehicles.

Demand Response Programs

Demand response programs pay industrial and residential customers to reduce or shift consumption during peak hours. A large industrial facility that curtails an aluminum smelter or steel mill for three hours on a July evening provides as much grid relief as a moderately sized power plant. Residential demand response (particularly smart thermostat programs) can defer cooling load and flatten the evening demand spike. Texas used demand response successfully during summer 2023 to avoid the kind of grid crisis that occurred in 2021.

Curtailment: The Waste Problem

Curtailing solar (telling panels to stop generating) is a real and growing issue. CAISO curtailed 3.4 million megawatt-hours of utility-scale wind and solar output in 2024, 29% more than in 2023 and 93% of it solar, enough to supply about 315,000 average US homes for a year at EIA's average of about 10,800 kWh per home. That's clean energy being discarded because the grid can't use it. Every megawatt-hour curtailed represents a missed opportunity to displace fossil fuel generation.

Reducing curtailment requires more storage, more transmission, or more demand-shifting. None of those solutions happens overnight, but the policy and market incentives are increasingly aligned in that direction. The duck curve, uncomfortable as it is for grid operators, is ultimately an engineering problem with known solutions, just solutions that require substantial investment and time to deploy at scale.

Related reading: For van life and RV picks, see our van-life rankings. Demand Response Programs covers demand response programs.