How the Merit-Order Curve Explains Power Price Spikes

Wholesale electricity prices are determined by a system called the merit-order curve, which turns power plants on from the cheapest to the most expensive to run. When consumer demand spikes, grid operators must fire up costly, inefficient fossil-fuel plants to avoid blackouts, and that final, most expensive plant sets the price for all electricity on the grid. Today, giant battery systems are fundamentally altering this curve by storing cheap solar power during the day and discharging it during peak hours, effectively cutting off the most expensive price spikes.

The Invisible Market Ruling Your Energy Bill

Electricity is a uniquely challenging commodity. Unlike wheat, crude oil, or copper, it is difficult and highly expensive to store in massive quantities, meaning supply and demand must be perfectly balanced every second of every day across thousands of miles of transmission lines. To maintain this delicate and vital balance, modern wholesale electricity markets rely on a pricing mechanism known as marginal pricing, governed by the "merit-order curve."

While the terminology sounds highly academic, the underlying concept is straightforward and mirrors other commodity markets. Imagine two wheat farmers supplying a town. One farmer utilizes modern, highly efficient equipment and can profitably produce a bushel of wheat for two dollars. The other farmer utilizes manual labor and requires ten dollars a bushel to break even. If the town only needs a small amount of wheat, they buy it exclusively from the efficient farmer for two dollars. But if the town is starving and needs every grain available to survive, both farmers must be called upon. The expensive farmer will not sell for less than ten dollars, or they would go bankrupt. Consequently, the market price for all wheat rises to ten dollars to ensure the required supply is met ¹.

Electricity markets function exactly the same way. Power generators are ranked not by how much capital it cost to build the physical plant, but by their "short-run marginal cost" - the exact cost required to generate one additional megawatt-hour (MWh) of electricity at a given moment ²³. The grid operator mathematically lines up all available power plants from the cheapest marginal cost to the most expensive. This cumulative lineup is the merit-order curve ¹³.

As homes and businesses consume power, the grid operator moves left to right along this curve, activating plants sequentially until demand is perfectly met. The final power plant turned on to satisfy the exact level of demand is known as the "marginal plant." Whatever it costs that specific plant to generate its electricity becomes the clearing price paid to every single generator supplying the grid during that specific trading interval ¹⁵⁶.

Visually, this curve looks like a staircase moving upward from left to right. The X-axis represents the cumulative gigawatts of capacity available, while the Y-axis represents the marginal cost in dollars per megawatt-hour. When a vertical line representing current consumer demand is drawn across the chart, where it intersects the staircase dictates the price. During times of low demand, the intersection happens low on the staircase. During extreme heatwaves or winter storms, the demand line pushes far to the right, hitting the highest steps on the curve.

The Uniform Price Debate

A frequent critique of this system - often raised during periods of high inflation or energy crises - is that it seems fundamentally unfair to pay a wind farm the same high price as a gas plant. Because a wind farm generates electricity for free once built, politicians and consumer advocates frequently question why electricity prices are tethered to volatile fossil fuel costs rather than the average cost of all generation combined ¹.

However, economists, grid operators, and market monitors overwhelmingly support the uniform marginal pricing model as the most efficient way to allocate resources ³⁴. If a wind farm were forced to accept a lower price simply because its fuel is free, the financial incentive to build more wind farms would evaporate. By paying all generators the marginal clearing price, the market heavily rewards the most efficient, low-cost technologies ¹³.

The profits earned by renewable generators during high-price hours are exactly what incentivize private developers to pour billions of dollars into building more clean energy infrastructure without relying entirely on government subsidies ¹. The European Union Agency for the Cooperation of Energy Regulators (ACER) recently concluded that the current wholesale electricity market design ensures efficient supply, and that shifting away from marginal pricing would distort market signals and ultimately increase costs for consumers ⁶⁴. In Europe alone, cross-border trade driven by these integrated marginal pricing markets delivers an estimated €34 billion per year in benefits for consumers by constantly seeking out the cheapest available generation across the continent ⁴⁵.

Other mechanisms exist alongside this wholesale market to protect consumers from extreme volatility. In Ireland's Single Electricity Market (SEM), for instance, the government uses Contracts for Difference (CfDs) under the Renewable Electricity Support Scheme (RESS). If wholesale prices fall too low, generators receive a top-up payment to ensure financial viability. However, when market prices soar above a guaranteed strike price, renewable generators must return the excess revenues to a public fund, directly limiting excessive price increases for retail consumers ⁵.

Stacking the Deck: The Generation Merit Order

To understand how and why wholesale prices move, one must examine the specific technologies that make up the dispatch stack. Power plants are generally divided into variable renewables, continuous baseload generation, and dispatchable peaking plants. Their physical position in the merit-order lineup is dictated almost entirely by their variable operating and maintenance (O&M) costs, which are primarily driven by fuel ².

Fixed costs - the billions of dollars spent pouring concrete, installing wind turbines, or financing long-term debt - are completely ignored in this daily operational lineup ²³. Once a power plant is built and connected to the grid, the only question that matters for grid dispatch is how cheaply it can run today.

Zero-Cost Sunshine and Intermittent Renewables

At the very front of the line are variable renewable energy sources like wind and solar photovoltaic (PV) arrays. Once the panels and turbines are installed, the fuel - wind and sunshine - is entirely free. Their short-run marginal costs are virtually zero, consisting only of minor variable wear-and-tear expenses ³. Consequently, whenever the sun is shining and the wind is blowing, these generators bid into the market at zero, guaranteeing they are dispatched first ³⁶.

This front-of-the-line status is bolstered by the fact that renewables are now the cheapest form of new electricity generation to build. In 2024, the global weighted average levelized cost of electricity (LCOE) for new onshore wind fell to just $0.034 per kilowatt-hour (kWh), closely followed by solar PV at $0.043/kWh ⁷⁸. Astonishingly, 91% of all newly commissioned utility-scale renewable projects globally delivered electricity at a lower cost than the cheapest new fossil fuel-fired alternative ⁷⁸. As a result, solar and wind energy combined are projected to rise to over 19% of global electricity generation by 2026, physically pushing more expensive generation further down the merit order ⁹.

The Steady Baseload: Nuclear Power

Following renewables in the dispatch order are nuclear power plants. While incredibly expensive and time-consuming to construct - often suffering from severe cost overruns - nuclear reactors are incredibly cheap to operate once they are online. Uranium fuel costs are a tiny fraction of total generating expenses, giving nuclear a highly competitive position near the bottom of the curve ¹⁰¹¹.

For comparison, the fuel cost for a nuclear plant is typically between one-quarter and one-fifth of the fuel cost for a gas combined-cycle plant ¹⁰. Lazard's 2024 Levelized Cost of Energy report assumes a fuel cost of just $0.85 per million British thermal units (MMBtu) for nuclear generation, far below fossil fuels ¹²¹³. Because nuclear plants rely on complex thermal dynamics and are difficult to ramp up and down quickly, they typically run continuously at maximum capacity to supply steady, reliable baseload power ¹⁰. Driven by restarts in Japan and new reactors in China and the United States, global nuclear generation is forecast to hit record highs through 2025 and 2026 ⁹.

The Fossil Fuel Workhorses: Coal and CCGT

Next in line are the traditional fossil fuel workhorses: coal-fired steam plants and combined-cycle natural gas turbines (CCGT). The marginal costs for these plants are highly sensitive to global commodity markets ¹⁰¹⁴.

The relationship between coal and natural gas creates a dynamic middle section of the merit-order curve. If natural gas prices plummet due to high production, gas plants become cheaper to run and move ahead of coal in the merit order, a phenomenon known as coal-to-gas switching. Conversely, if gas prices surge due to geopolitical conflicts or supply chain bottlenecks, coal becomes relatively cheaper and is dispatched earlier ⁹¹⁴.

The U.S. Energy Information Administration (EIA) forecasts the average cost of coal delivered to the U.S. electric power sector to hover around $2.36 to $2.42 per MMBtu through 2026, while natural gas prices experience much sharper volatility . CCGT plants generally feature higher thermal efficiency than older coal plants, allowing them to extract more megawatt-hours of electricity per unit of fuel burned. Despite occasional fluctuations, the broader trend is moving away from coal; global coal-fired generation is expected to drop below 33% of the total generation mix for the first time in a century by 2025 ⁹.

The Peakers: Emergency Gas Generators

At the very back of the line sit the "peakers." These are often older, less efficient fossil fuel plants, or open-cycle gas turbines (OCGT) that act essentially like giant jet engines hooked to a generator ⁶¹⁵. They are extremely expensive to run and highly inefficient, but they possess one crucial advantage: they can start up in minutes.

Grid operators rely on peaker plants as an absolute last resort when demand suddenly surges and all other generation on the grid is maxed out ¹⁵¹⁶. Because they sit at the highest point of the merit-order curve, whenever a peaker plant is turned on, it sets the clearing price for the entire market at a severe premium.

The Duck Curve and the Chaos of the Peak

The modern electricity grid is increasingly defined by a visual and physical phenomenon known as the "duck curve." As nations aggressively build out utility-scale solar capacity, the sheer volume of zero-marginal-cost energy is physically pushing the merit-order curve to the right during daylight hours ³⁶.

During the middle of the day, solar panels flood the grid, creating a deep trough in net demand - the belly of the duck. When solar generation is abundant, it displaces higher-cost fossil fuel plants entirely. If demand is moderate and solar and wind are producing at maximum capacity, the grid may not need to turn on a single gas or coal plant. In this scenario, the marginal plant setting the price is a wind or solar farm, and the wholesale price of electricity crashes toward zero ³⁵⁶.

The Rise of Negative Pricing

This dynamic has given rise to an increasingly common phenomenon: negative electricity prices. When the grid is flooded with more renewable energy than consumers can use, and interconnection cables to neighboring grids are maxed out, power plants will actually pay the grid operator to take their electricity ³¹⁷.

This occurs because certain baseload generators, like nuclear or coal plants, face significant technical limitations or severe financial penalties for shutting down completely and restarting later. They would rather pay a small penalty to keep running through a brief period of oversupply than undergo a complex, costly shutdown sequence ⁶¹⁷.

The frequency of negative pricing is surging globally as renewables scale. In 2024, wholesale prices dipped below zero for roughly 25% of all hours in South Australia, and 15% of all hours in Southern California ¹⁸. In Europe, Finland led the continent by experiencing negative prices 8% of the time, while Sweden, the Netherlands, and Germany saw negative prices between 5% and 7% of the year ¹⁸. While most negative prices remain moderate - averaging between -$2 to -$30 per MWh - they signal a profound lack of system flexibility during periods of abundant generation ¹⁸.

The Evening Scramble

While abundant sunshine forces prices to zero at noon, the exact opposite occurs when the sun sets. As evening approaches, solar production rapidly drops to zero at the exact moment people return home from work, turn on their air conditioners, plug in their vehicles, and begin cooking ¹⁶¹⁹.

This creates a terrifying mathematical reality for grid operators. Within a span of two to three hours, they must find massive amounts of replacement electricity to match the sudden, vertical spike in net demand - the neck of the duck ²⁵²⁶.

As the grid operator scrambles to find power, they march rapidly to the far right side of the merit-order curve. They exhaust the cheap renewables, blow past the baseload nuclear and coal, and are forced to activate the gas peaker plants ¹¹⁶. Because the clearing price is set by the most expensive unit running, the activation of just one inefficient peaker plant causes the price of every megawatt-hour on the grid to instantly skyrocket ¹⁵. During these critical evening hours, wholesale prices can multiply by five or ten times their daytime average ¹⁶¹⁹. A failure to secure enough power at any price results in rolling blackouts.

Enter the Giant Batteries: Flattening the Curve

For decades, the only viable way to manage evening demand spikes was to build more gas peaker plants and accept the high prices. However, a relatively new technology is fundamentally disrupting the merit-order curve: grid-scale battery energy storage systems (BESS) ¹⁵¹⁷.

Batteries operate outside the traditional rules of electricity generation because they do not actually produce new energy; they merely shift existing energy across time ¹⁶¹⁷. Driven by massive manufacturing scale-ups and improved material efficiencies, the cost of utility-scale battery storage fell to $192/kWh of capacity in 2024 - a 93% decline since 2010 ⁸. Independent analysis by Ember estimates that the actual levelized cost of storage (LCOS) to move electricity across time has plummeted to roughly $65/MWh, making batteries vastly more economical than building new combustion peaker plants ²⁰²¹.

Batteries participate in the wholesale market through extreme price arbitrage ¹⁵¹⁶. They act as a massive sponge during the middle of the day, buying up excess solar power when the merit-order curve is flat and prices are zero or negative. Then, during the evening demand spike, they discharge that stored power back onto the grid ¹⁵¹⁶¹⁷.

This temporal shift fundamentally alters the shape of the market. By discharging massive amounts of cheap, stored solar energy during the peak, batteries prevent the grid operator from having to move to the far right of the merit-order curve ¹⁷²⁵. The expensive gas peakers are simply no longer needed to meet demand, which forces the clearing price down for the entire market ¹⁶¹⁷.

Real-World Proof of Market Disruption

The theoretical impact of batteries is no longer just a projection; it is highly visible in live electricity markets worldwide. Early studies in Europe showed that between 2020 and 2022, adding just 100 MW of storage reduced daily price spreads by an average of five percent ¹⁷. But the true scale of the disruption became evident in 2024 and 2025 in markets heavily saturated with solar power.

In California's Independent System Operator (CAISO) market, the evening surge in natural gas generation has historically been a reliable, daily event. However, by April 2024, the sharp evening gas peak had eroded into a flat plateau. Battery capacity, which grew to 13,000 MW by the end of 2024, absorbed the massive ramps in solar output, directly displacing natural gas generation precisely when it used to be most profitable ¹⁵²⁵²⁶. As a result, net market revenues for batteries in CAISO decreased from an average of $78 per kilowatt-year in 2023 to $53 in 2024, largely driven by lower peak energy prices caused by their own success at flattening the curve ¹⁵.

To keep these assets viable, markets like CAISO employ Bid Cost Recovery (BCR) mechanisms. Originally designed to help thermal generators cover the costs of starting up and running at minimum load, BCR is now applied to batteries to compensate them for opportunity costs. Between January and August 2024, CAISO batteries earned over $12 million in BCR payments, ensuring they remain profitable even as they suppress the volatility they rely on for arbitrage ¹⁵²².

Similarly, in Australia's National Electricity Market (NEM), grid-scale battery capacity more than doubled over the 12 months leading to the first quarter of 2026. This massive influx allowed batteries to become the single most frequent price-setting technology in the entire market, setting the wholesale price in roughly 32% of all trading intervals. By capping the high-end prices during demand spikes and absorbing excess power during the day, batteries helped drive a 12% year-over-year decline in average wholesale electricity prices across the Australian market ²³.

Regional Grid Dynamics: CAISO, PJM, and ENTSO-E

While the merit-order curve is a universal economic principle, how it manifests depends heavily on regional infrastructure, policies, and market design. The struggles and solutions in three major grid areas illustrate the complexities of managing modern power demands.

CAISO: Managing the Solar Surge and AI Load

In California, grid operators are grappling with extreme solar saturation and rapid load growth. The 2024-2025 CAISO Transmission Plan recommends 31 infrastructure projects costing an estimated $4.8 billion to accommodate forecasted demand ²⁴²⁵. Much of this load growth is concentrated in the Greater San Francisco Bay Area, driven by the electrification of transportation and an anticipated surge in energy-intensive data centers used for artificial intelligence ²⁴²⁵.

To manage this growth, CAISO relies heavily on its Western Energy Imbalance Market (WEIM), which allows it to trade electricity seamlessly with neighboring states in real-time. During peak solar hours, CAISO exports excess generation to the Desert Southwest; during non-solar hours, it imports hydro and wind power from the Pacific Northwest, constantly adjusting the regional merit order to find the cheapest electrons ¹⁵²⁶.

PJM Interconnection: Capacity Market Strains

In the Eastern United States, PJM Interconnection - which coordinates wholesale electricity for over 51 million people across 13 states and Washington D.C. - faces a different set of challenges. PJM relies heavily on natural gas, coal, and nuclear power, and is experiencing severe strains in its capacity market ²⁷²⁸²⁹.

While the daily energy market buys electricity based on the merit-order curve, the capacity market acts like an insurance policy, paying generators simply to be available in the future. In PJM's 2025/2026 capacity auction, prices cleared significantly higher than expected. The Independent Market Monitor for PJM noted that the results were driven by market design changes - such as shifting the availability metric from Equivalent Demand Forced Outage Rate (EFORd) to Effective Load Carrying Capability (ELCC) - and the ongoing retirement of fossil fuel plants ³⁰³¹. With gigawatts of older coal plants like the 180 MW Warrior Run facility retiring, and interconnection queues facing years-long delays to approve new renewable projects, the localized supply is tightening, which puts upward pressure on prices ²⁸³⁰.

ENTSO-E: The Power of Continental Coupling

In Europe, the European Network of Transmission System Operators for Electricity (ENTSO-E) manages the largest interconnected electrical grid in the world, representing 40 TSOs across 36 countries ³²³³. Europe's strategy for managing price spikes relies heavily on advanced market coupling algorithms, specifically the Single Day-Ahead Coupling (SDAC) and Single Intraday Coupling (SIDC) ³⁴.

Using the EUPHEMIA algorithm, SDAC calculates electricity prices across Europe while implicitly allocating cross-border capacity ³⁴. Instead of each country relying solely on its own internal merit-order curve, flow-based market allocation allows cheaper electricity to flow seamlessly across borders. If Germany is experiencing a demand spike, but cheap wind power is available in Denmark, the algorithm routes that power south, preventing Germany from having to turn on expensive local gas peakers ³⁵. ACER confirmed that this coordinated flow-based approach is vital for resilience, particularly during weather anomalies like "Dunkelflaute" - extended periods of low wind and high demand - which drove price surges in the Core European region to twice the monthly average in late 2024 ³⁵.

Why Wholesale Drops Don't Always Lower Retail Bills

If wholesale electricity prices are periodically suppressed by cheap renewables, cross-border trading, and battery arbitrage, consumers frequently wonder why their monthly utility bills continue to rise. According to the EIA, U.S. residential electricity prices averaged 18.2 cents per kWh in 2026, an increase of nearly 5% from the previous year .

The disconnect lies in the fundamental difference between wholesale generation costs and retail delivery costs. The price dictated by the merit-order curve only accounts for the raw electricity generated at the plant ³⁶. To physically get that power to a residential home or commercial business, local distribution utilities must pay for extensive planning, maintenance, and vast transmission networks ³⁷⁴⁵.

Recently, the costs to maintain and expand this physical infrastructure have skyrocketed. Inflation and rising material costs have made basic system upgrades significantly more expensive. Furthermore, utilities are forced to spend billions on infrastructure hardening - burying lines, replacing aging towers, and purchasing liability insurance - to protect the grid against extreme weather and mitigate the risk of catastrophic wildfires ³⁷.

Additionally, the rapid expansion of hyperscale data centers is forcing utilities to build entirely new high-voltage transmission lines. While these massive consumers ultimately pay for their energy, the upfront capital costs to upgrade the local grid are frequently socialized across all ratepayers in a given territory. Consequently, even if the wholesale price of power drops to zero during the day, the fixed costs of maintaining the physical grid ensure that retail bills remain high ³⁷⁴⁵³⁸.

The Shift to Time-of-Use (TOU) Pricing

To prevent the grid from buckling under peak demand - and to limit their financial exposure to the extreme wholesale price spikes at the top of the merit-order curve - utilities globally are increasingly shifting consumers away from traditional flat-rate retail pricing ³⁹⁴⁰.

Historically, homeowners paid the exact same price for electricity whether they ran their high-wattage clothes dryer at 2:00 AM or 6:00 PM ⁴¹⁴². This model shielded consumers from wholesale volatility, but it also provided zero incentive for them to conserve energy when the grid was under severe stress. Today, utilities and regulators are mandating or highly incentivizing Time-of-Use (TOU) tariffs ³⁹⁴².

TOU tariffs divide the day into specific pricing blocks that directly mirror the realities of the wholesale merit-order curve. In the United Kingdom, the rollout of Market-wide Half-Hourly Settlement (MHHS) in 2026 is fundamentally reshaping how organizations pay for power, shifting them entirely away from flat rates toward dynamic time-of-use billing ⁴⁰.

Note: Exact timing, terminology, and percentages vary significantly by local utility and seasonal weather patterns ³⁹⁴¹⁴².

The financial gaps between these tiers are growing rapidly and deliberately. In states like California, peak pricing can reach 74 cents per kWh, while off-peak rates drop to just 21 cents ³⁹. In Colorado, peak rates are strictly structured to be 2.7 times higher than off-peak rates ⁴³. Even solar feed-in tariffs - the rate utilities pay homeowners for exporting rooftop solar back to the grid - are dropping sharply during the day to reflect the low wholesale value of midday power. In Queensland, Australia, feed-in tariffs dropped from 8.6 cents to 6.1 cents per kWh to reflect the flooded daytime market ⁴⁴.

For households and businesses that do not change their behavior, TOU tariffs can result in severe bill shock, acting as a direct financial penalty for relying on the grid when expensive gas peakers are running ⁴².

Beating the Peak with Smart Home Technology

While TOU tariffs introduce price volatility to the consumer level, they also offer massive savings opportunities for flexible households. By utilizing the exact same economic logic as grid-scale batteries, consumers can slash their bills by temporally shifting their energy consumption away from the top of the merit-order curve and into the troughs ⁴¹⁴².

Basic behavioral changes deliver substantial returns. Simply setting an electric vehicle (EV) charger to activate only after midnight, scheduling pool pumps to run during the midday solar peak, and delaying dishwashers until late evening can easily save hundreds of dollars annually ³⁹⁴¹.

Increasingly, this load-shifting process is being automated by the Internet of Things (IoT). Smart thermostats integrated with Artificial Intelligence track indoor temperatures, outside weather patterns, and the times a family comes and goes. These devices automatically pre-cool or pre-heat the house during off-peak hours when power is incredibly cheap, and then smoothly throttle the HVAC system down just as the expensive peak hours begin ⁵³⁴⁵⁴⁶. Field studies and data from Energy.gov show that optimized smart HVAC systems can reduce heating and cooling costs by 10% to 15% annually ⁵³⁴⁷.

Virtual Power Plants and Demand Response

The impact of these devices extends far beyond individual household savings. When thousands of smart thermostats, EV chargers, and residential batteries are aggregated together by utilities or third-party operators, they form a Virtual Power Plant (VPP) ⁴⁵⁴⁸.

VPPs operate through "Demand Response" programs. Instead of asking a power plant to generate more electricity during a crisis, the utility sends a digital signal to the VPP to temporarily reduce consumption ²⁹⁴⁵⁴⁹. By simultaneously lowering the air conditioning output in tens of thousands of homes by just a few degrees during a demand spike, a VPP physically lowers the total demand line on the merit-order curve ⁵³⁴⁹.

This allows the grid operator to turn off a highly polluting, expensive gas peaker plant - or avoid turning it on in the first place ²⁹⁵³. Participants are financially compensated for their flexibility, and the overall clearing price for the wholesale market is suppressed, benefiting every ratepayer on the network. Despite the obvious benefits, adoption barriers remain; while smart thermostat adoption doubled over eight years to reach 16% of U.S. internet households, only about 20% of those households are currently enrolled in utility demand response programs, highlighting a massive opportunity for future grid efficiency ⁴⁸.

Bottom line

The merit-order curve is an invisible but economically ruthless mechanism that ensures the cheapest available electricity is used first, but allows the single most expensive power plant required to set the final market price for everyone. While sudden demand spikes have historically caused severe price surges by forcing grid operators to rely on costly gas peaker plants, the rapid deployment of grid-scale batteries and cross-border trading algorithms are effectively shaving off these peaks and stabilizing wholesale costs. For the average consumer, the impact of these volatile wholesale dynamics is increasingly felt through Time-of-Use tariffs, heavily rewarding those who use smart home technology to shift their power consumption away from the expensive evening rush.