The Duck Curve, Explained

The duck curve is what a day’s electricity demand looks like once you subtract the solar power generated at the same time. In a system with a lot of solar, the leftover demand, which is what everything else has to supply, sags deeply through the middle of the day and then climbs steeply as the sun sets and people come home. Drawn on a chart, the sag and the climb look like the belly and neck of a duck.

Net load, not demand

The distinction that makes the duck curve make sense is between demand and net load.

Demand is how much electricity consumers use. That pattern has been broadly stable for a century: low overnight, rising through the morning, a plateau during the working day, a peak in the early evening, then a decline.

Net load is demand minus the output of generators the operator cannot control, mainly wind and solar. It is the amount the rest of the fleet has to cover. When solar capacity is small, net load looks much like demand. As solar grows, a bite is taken out of the middle of the day, and net load stops resembling demand at all.

The duck appears in net load. Consumer behaviour has hardly changed. What changed is how much of that behaviour is met before the dispatchable plants are asked to do anything.

California, the classic example

The shape was popularised by grid planners in California, who published a series of projected net load curves in 2013 showing the midday dip deepening year after year. The projections turned out to be conservative: the dip arrived faster than expected.

The CAISO balancing authority covers most of California, and EIA data shows solar supplying about 26% of the electricity generated within it across 2025. That is an annual average across all hours, including every night, so the midday share on a clear spring day is far higher.

Two features of California make the effect especially sharp. Its solar fleet is very large relative to its demand, and a lot of it sits on rooftops behind customers’ meters, where it reduces measured demand rather than appearing as generation. Operator statistics generally capture utility scale plants, so the real midday dip is deeper than the published generation figures alone suggest.

The three problems the shape creates

Oversupply in the belly. On mild, sunny, low demand days, mainly in spring, there can be more generation available than the system needs. Prices fall towards zero or below, and some output is curtailed. Plants that cannot easily reduce output make this worse.

The evening ramp up the neck. This is the hardest part. As the sun goes down, solar output falls to zero over a couple of hours at the same time as household demand rises. The gap has to be filled quickly, which means thousands of megawatts of capacity starting or increasing output within a short window. Ramping capability, measured in megawatts per minute, becomes as important as capacity itself. Not every plant can move that fast, and those that can are often the more expensive ones.

A peak that moves later. In a system without much solar, the annual peak in net load tends to fall in the late afternoon. With a lot of solar, the peak shifts into the evening, after solar has faded. That matters for planning, because the question is no longer how much capacity is available at four in the afternoon but how much is available at eight in the evening, when solar contributes nothing.

How grids respond

Storage. Batteries are close to a purpose built answer: they charge in the belly, when power is abundant and cheap, and discharge up the neck, when it is scarce and dear. They earn money from that spread, which is why battery fleets have grown quickly in exactly the places where the duck is deepest.

Flexible generation. Plants that can start quickly and ramp hard, including some gas turbines and hydro units, are valued for the ramp rather than for energy over the year. Their capacity factors can be low while their contribution is essential.

Moving demand. Electric vehicle charging is the biggest opportunity, because a car parked at work all day can absorb midday surplus, while the same car charging the moment its owner gets home makes the neck steeper. Tariffs that reward daytime charging, water heating on timers and pre-cooling buildings before the evening all shift load into the belly.

Trade with neighbours. Sunset moves west, so a system can import from a region where the sun is still up, or export a midday surplus to a region that is cloudy or has less solar.

Building differently. Wind often produces more in the evening and overnight, so a mixed portfolio is flatter than either resource alone. Panels facing west produce less over a day but more in the late afternoon, which is worth more when the peak sits in the evening.

Where else it appears

The duck is not a Californian peculiarity. Any system that adds enough solar develops the same shape: parts of Australia, Spain, Chile and Hawaii show it, and so do a growing number of US states. Local conditions change the details. Where air conditioning drives a late afternoon peak, the neck starts higher. Where winter heating dominates, the deepest bellies come in spring and autumn rather than summer. Systems with substantial hydro have a ready made way to ride the ramp.

The one thing all of these have in common is that the challenge is not the total amount of solar energy. It is the timing. Electricity is only valuable at the moment it is delivered, and the duck curve is the clearest picture anyone has drawn of what that means.

Reading a net load chart

If you want to see the shape yourself, look at an hourly generation mix for a sunny region over a single clear day rather than at annual totals. Watch solar rise from nothing at dawn to its maximum around midday, then collapse through the afternoon. Watch gas, hydro and imports do the opposite. The mirror image between them is the duck, and it explains most of what is difficult and most of what is promising about a solar heavy grid.