Spark Spreads: Power Plant Economics and Generation Margins

Published: June 23, 2026

Article by Ryan O'Connell, CFA, FRM

The spark spread is the key profitability metric for gas-fired power generation. It measures the gross margin a power plant earns by converting natural gas into electricity — and helps determine whether a plant should dispatch (run) or stay idle. Understanding spark spreads is essential for energy traders, power plant operators, and anyone analyzing electricity market economics.

What Is a Spark Spread?

A spark spread represents the difference between the revenue from selling electricity and the cost of the natural gas fuel required to generate it. In simple terms, it’s the gross profit margin for a gas-fired power plant before accounting for other operating costs.

The coal equivalent of the spark spread is called the dark spread, which measures the margin for coal-fired generation. Both metrics help grid operators and traders understand which plants are economically viable to run at any given time.

Spark spreads vary significantly across regions due to differences in natural gas prices, electricity market structures, and plant efficiency. A plant that’s profitable in one market may be uneconomical in another.

Heat Rate: Converting Gas to Power

The heat rate measures how efficiently a power plant converts fuel into electricity. It’s expressed in British thermal units (Btu) per kilowatt-hour (KWh) or, equivalently, million Btu (MMBtu) per megawatt-hour (MWh). The lower the heat rate, the more efficient the plant — and the higher its spark spread for any given set of prices.

The Spark Spread Formula

The spark spread calculation converts gas and electricity prices to comparable units:

Where:

• P_electricity — wholesale electricity price in dollars per megawatt-hour ($/MWh)
• P_gas — natural gas price in dollars per million Btu ($/MMBtu)
• Heat Rate — plant efficiency in MMBtu per MWh (or equivalently, thousand Btu per KWh divided by 1,000)

The result is the gross margin in $/MWh that the plant earns on fuel alone. This is sometimes called the gross spark spread to distinguish it from the net or operating spark spread, which subtracts variable O&M and other costs.

Spark Spread Example

Now consider a less efficient NRG Energy peaker plant in the same market with a heat rate of 10.0 MMBtu/MWh:

Dark Spreads and Coal-Fired Generation

The dark spread is the coal equivalent of the spark spread. It measures the gross margin for coal-fired power plants:

Coal is typically priced per short ton, so conversion to $/MMBtu is required before applying the formula. For bituminous coal with approximately 20-25 MMBtu per ton, divide the $/ton price by the energy content to get $/MMBtu. Coal plants generally have higher heat rates than efficient gas plants, but coal fuel costs are often lower per MMBtu, which historically made coal competitive for baseload generation.

The dark spread helps analysts compare the economics of coal versus gas generation. When spark spreads are high relative to dark spreads, gas plants capture more of the dispatch stack. When dark spreads are more favorable, coal plants run more frequently. Learn more about fuel pricing in our commodity futures guide.

Clean Spreads: Carbon-Adjusted Margins

In markets with carbon pricing — such as the EU Emissions Trading System (ETS), RGGI in the northeastern US, or California’s cap-and-trade program — generators must account for the cost of emissions. The clean spark spread and clean dark spread subtract carbon costs from the gross spread:

The emission factor depends on both the fuel and the plant’s heat rate:

Approximate emission factors at typical heat rates:

• Natural gas (CCGT): ~0.4 tonnes CO₂/MWh
• Coal (baseload): ~0.9-1.0 tonnes CO₂/MWh

With carbon prices of €50-100/tonne in the EU ETS, clean spreads can be significantly lower than gross spreads — especially for coal plants. Clean spreads are increasingly the relevant metric for dispatch decisions in carbon-regulated markets.

Power Plant Dispatch Decisions

Grid operators dispatch power plants in merit order — from lowest to highest marginal cost — to meet electricity demand. The spark spread directly influences where a gas plant falls in this stack:

• Positive spark spread > variable costs — The plant is economically viable to run. It earns more from selling electricity than it spends on fuel and variable O&M.
• Positive spark spread < variable costs — The plant covers fuel but not other variable costs. It may run only during peak hours when prices spike.
• Negative spark spread — Running the plant loses money on fuel alone. The plant stays offline unless contractually obligated.

From a real options perspective, a flexible power plant can be viewed as a strip of spark spread options — the operator has the option to convert gas into electricity whenever the spread is favorable. This optionality has value, which is why peaker plants command positive valuations even when their average spark spread is low. For more on valuing capital investments, see our guide on cost of capital.

Trading Spark Spreads

Spark spreads are often traded over-the-counter (OTC) or constructed synthetically using listed gas and power contracts. Some exchanges, including ICE, offer spark spread products that combine both legs. Cleared OTC electricity contracts also facilitate spread trading with reduced counterparty risk.

Long Spark Spread (Speculative Position)

A trader expecting spark spreads to widen takes a long spark spread position:

• Buy electricity futures at a specific hub (e.g., PJM Western Hub)
• Sell natural gas futures at the relevant delivery point (e.g., Henry Hub)
• Scale the gas position by the assumed heat rate

This position profits when power prices rise relative to gas prices — betting on improved generation economics.

Generator Hedge (Locking Margins)

A power plant operator hedging forward margins takes the opposite position:

• Sell electricity forwards to lock in future revenue
• Buy natural gas forwards to lock in future fuel costs

This locks in the spark spread at today’s forward prices, eliminating price risk on both legs.

Why Trade Spark Spreads?

• Generators hedge margins: Lock in profitable generation margins rather than taking price risk on both legs
• Speculators take views: Bet on generation economics without directional exposure to gas or power alone
• Utilities manage costs: Hedge procurement costs relative to generation alternatives

Spark spreads exhibit seasonal patterns, typically widening during peak demand periods (summer cooling, winter heating) when electricity prices spike relative to gas prices.

Spark Spread vs Dark Spread vs Clean Spread

Understanding the differences between these spread types is essential for analyzing power generation economics:

Common Mistakes in Spread Calculations

Analysts and traders frequently make these errors when calculating and interpreting spark spreads:

1. Ignoring unit conversions — Mixing Btu with MMBtu or KWh with MWh produces results that are off by factors of 1,000. Always verify that heat rate units match gas price units.

2. Using Henry Hub instead of delivered gas cost — Henry Hub is a benchmark, but plants pay delivered prices that include pipeline transport, basis differentials, and local supply constraints. A plant in New England may face gas costs $2-5/MMBtu higher than Henry Hub during winter.

3. Using average heat rate instead of marginal heat rate — Dispatch decisions depend on the incremental efficiency at the current operating point, not the average across all output levels.

4. Forgetting carbon costs in regulated markets — In the EU, California, or RGGI states, ignoring emissions costs significantly overstates the effective margin.

5. Confusing positive spark spread with profitability — The spark spread is a gross margin on fuel only. A positive spark spread does not guarantee the plant is profitable after variable O&M, startup/shutdown costs, emissions allowances, and fixed costs.

6. Treating heat rate as constant — Efficiency varies with output level, ambient temperature, and equipment condition. Use operating-condition-specific heat rates for accurate dispatch analysis.

7. Ignoring locational price differences — Electricity prices vary by node due to transmission congestion. Using hub prices instead of nodal LMPs can misrepresent actual plant economics.

Limitations of Spark Spread as Dispatch Signal

• Startup and shutdown costs — Starting a plant consumes fuel and causes wear. Frequent cycling reduces net margins.
• Ramp rate constraints — Plants cannot instantly change output. Ramp limitations affect the ability to capture short-term price spikes.
• Transmission constraints — Congestion can prevent a plant from delivering power to high-price zones even when the spread is favorable.
• Ancillary services revenue — Plants may earn additional revenue from reserves, frequency regulation, or voltage support — not captured in the spark spread.
• Contractual obligations — Take-or-pay gas contracts or power purchase agreements may require operation regardless of spot economics.
• Point-in-time prices — Dispatch planning uses forward curves and expected prices, not just current spot prices.