Lesson 7 — Cross-Commodity Spreads

📖 Glossary · ← Learning Hub Lesson 7 Knowledge + Skills Power + Gas + Emissions
Spark, dark, and clean spreads — how cross-commodity relationships drive dispatch decisions, hedging, and trading strategies.

Energy commodities don't exist in isolation. Gas becomes power through a turbine. Coal becomes power through a boiler. Both emit CO₂ requiring allowances. These physical transformation relationships create tradeable spreads — the margin between input cost and output revenue. Spreads are the economic signal that determines which plants run, which are mothballed, and how portfolios are hedged.

The Fundamental Spreads

Spark Spread (Gas → Power)

Spark Spread = P_power − P_gas / η_gas

The gross margin of a gas-fired power plant. P_gas is divided by thermal efficiency η (typically 0.45-0.60 for CCGT) to convert from fuel cost per MWh_thermal to fuel cost per MWh_electrical. Positive spark = running is profitable (before carbon, O&M).

Dark Spread (Coal → Power)

Dark Spread = P_power − P_coal / η_coal

The gross margin of a coal-fired plant. Coal efficiency η typically 0.33-0.42. Coal priced in $/t (API2 benchmark for NW Europe delivery) converted to €/MWh_th using calorific value (~6,978 kcal/kg for API2).

Clean Spark Spread

Clean Spark = P_power − P_gas / η_gas − P_EUA × EF_gas / η_gas

Spark spread minus carbon cost. EF_gas ≈ 0.202 tCO₂/MWh_th (emission factor for natural gas). This is the true dispatch margin — what a gas plant earns after paying for fuel AND carbon.

Clean Dark Spread

Clean Dark = P_power − P_coal / η_coal − P_EUA × EF_coal / η_coal

Dark spread minus carbon cost. EF_coal ≈ 0.341 tCO₂/MWh_th. Coal emits ~70% more CO₂ per MWh_th than gas → more sensitive to EUA prices. The "fuel switch price" is where clean spark = clean dark.

Fuel Switch Price

P_EUA,switch = (P_coal/η_coal − P_gas/η_gas) / (EF_gas/η_gas − EF_coal/η_coal)

The EUA price at which gas becomes cheaper to run than coal. Below this price, coal dispatches first (lower fuel cost despite higher emissions). Above it, gas dispatches first. This formula drives EU climate policy's impact on the merit order.

Why Spreads Matter: The Merit Order

The merit order stacks generators by marginal cost. Cross-commodity spreads determine each unit's position in the stack:

Marginal cost (€/MWh) ───────────────────────────────────────────────→ Nuclear Wind/Solar Lignite Hard Coal CCGT OCGT Oil ~5-10 ~0 (must-run) ~25-35 ~45-65 ~55-80 ~90-130 ~150+ │ │ │ │ │ │ │ └── baseload ──────────────────────────┘ └── mid-merit / peaker ──┘ ← These spreads determine position: ─────────────────────────────────────────────→ Clean Dark Spread determines coal's position: MC_coal = P_coal/η + P_EUA × EF_coal/η + VOM Clean Spark Spread determines gas's position: MC_gas = P_gas/η + P_EUA × EF_gas/η + VOM When P_EUA rises → coal MC rises faster than gas MC → fuel switching (coal has higher EF, so carbon cost hits harder)

The fuel switch as policy lever: The EU ETS was designed to make coal-to-gas switching happen at EUA prices around 30-50 €/tCO₂ (depending on gas/coal price ratio). When EUA prices were below 10€ (2013-2018), coal ran ahead of gas despite higher emissions. The Market Stability Reserve (MSR) tightened supply, pushing EUAs above 80€ by 2022 — firmly placing efficient gas ahead of coal in the merit order.

Spread Trading Strategies

Hedging a Generation Asset

Strategy	Position	View Expressed	Risk
Long clean spark	Buy power, sell gas, sell EUAs	Gas plants will be dispatched (power rises relative to input costs)	Gas plants fall out of merit; power drops
Short clean dark	Sell power, buy coal, buy EUAs	Coal plants will NOT be dispatched (input costs exceed power revenue)	Coal comes back in-merit (cold snap, gas shortage)
Fuel switch	Long gas generation, short coal generation (via spreads)	EUA prices will rise making gas cheaper than coal to run	EUAs drop; coal becomes competitive again
Heat rate play	Long high-efficiency, short low-efficiency (via spread differential)	Efficient CCGTs run more hours than old OCGTs	Scarcity pricing favors all gas regardless of efficiency

A CCGT operator can lock in margins by trading the spread components separately on forward markets:

The result is a locked clean spark spread for the hedged period. The remaining risk: the actual plant may not run in those hours (if spot spread is negative), efficiency may differ from assumed η, and O&M costs are unhedged.

Volumetric risk: If you sell power forward (flat profile) but your plant doesn't run every hour (because spot spread is negative in some hours), you're short power in those hours at potentially high prices. This is why sophisticated hedging uses the expected generation profile (from dispatch simulation against the curve) rather than flat hedging. The residual shape is "intrinsic optimization value" — left unhedged to capture via spot trading.

Conversion Factors & Practical Calculations

Worked Example

Parameter	Typical Value	Source
Gas emission factor (EF_gas)	0.202 tCO₂/MWh_th	IPCC default
Coal emission factor (EF_coal)	0.341 tCO₂/MWh_th	IPCC default (hard coal)
CCGT efficiency (η_gas)	0.50–0.62 (HHV)	Plant-specific; F-class ~55%, H-class ~62%
Coal plant efficiency (η_coal)	0.33–0.42 (HHV)	Plant-specific; old ~35%, USC ~42%
API2 coal calorific value	6,000 kcal/kg (NAR)	Contract specification
Coal → €/MWh_th	$/t ÷ 7.0 × (USD/EUR rate)	6000 kcal/kg ≈ 6.978 MWh_th/t

Example: Q1-2026 Clean Spark Calculation

Given: DE Base Q1-26 = 78 €/MWh | TTF Q1-26 = 32 €/MWh | EUA Dec-25 = 72 €/tCO₂ | η = 0.56

Fuel cost = 32 / 0.56 = 57.14 €/MWh_el

Carbon cost = 72 × 0.202 / 0.56 = 25.97 €/MWh_el

Clean Spark = 78 − 57.14 − 25.97 = −5.11 €/MWh

Negative clean spark → this CCGT would NOT dispatch at these forward prices. Either the plant stays offline (if baseload), or runs only during peak hours where the peak price exceeds the marginal cost (peak clean spark may still be positive).

Spread Dynamics and Market Regimes

Regime	Characteristic	Spread Behavior
Gas-on-margin	Gas CCGT is the marginal price-setting unit	Power ≈ gas/η + carbon. Clean spark ~ VOM. Changes in gas/EUA pass through 1:1 to power.
Coal-on-margin	Coal is marginal (low gas, low carbon periods)	Power ≈ coal/η + carbon. Clean dark ~ VOM. Gas plants enjoy wide clean spark (infra-marginal rent).
RES-surplus	Renewables exceed demand (midday solar, high wind)	Power drops to 0 or negative. All thermal spreads deeply negative. Storage/flex captures value.
Scarcity	Demand exceeds all available generation	Power spikes to VOLL/cap. All thermal spreads extremely positive. Even inefficient peakers are in-merit.

Your tangible win: You can now compute any cross-commodity spread from first principles, understand how spreads drive merit-order dispatch, identify the fuel switch price, and construct hedging strategies for generation assets. You see how gas, coal, emissions, and power are coupled through physical transformation — not merely correlated.

Primary Source

For the theoretical link between spread dynamics and investment/dispatch decisions:
Green, R. & Vasilakos, N. (2010). "Market behaviour with large amounts of intermittent generation." Energy Policy, 38(7), 3211–3220.

For practical spread trading mechanics:
ICE Education, European Power & Gas Spreads — freely available webinars and explainers at ice.com/insights

Retrieval Practice

Why is the carbon cost per MWh electrical higher for coal than for gas, proportionally more than just their emission factors would suggest?

Coal plants must surrender more allowances per year due to their larger typical installed capacity
Coal has both a higher emission factor AND lower thermal efficiency, compounding the carbon cost per MWh of output
EU ETS applies a carbon multiplier to coal emissions but not to gas under the revised directive
Coal procurement contracts include embedded carbon surcharges that gas supply agreements do not carry

Retrieval Practice

An operator has locked a clean spark spread via forward trades. What residual risk remains?

The risk that EUA prices move between the hedge date and the date of futures settlement
The volumetric risk that the plant won't dispatch in all hedged hours if spot spreads turn negative
The risk that the exchange changes contract specifications for power futures during the delivery period
The counterparty credit risk on the central clearing party that guarantees the futures contracts

Retrieval Practice

In a "gas-on-margin" pricing regime, what happens to the power price when TTF increases by 5 €/MWh?

Power price increases by exactly 5 €/MWh since gas and power prices are perfectly correlated
Power price increases by approximately 5/η €/MWh (e.g., ~9 €/MWh at 56% efficiency) as the marginal plant's cost rises
Power price is unchanged because forward power is already hedged against gas price movements
Power price increases by 5 × η €/MWh since only the fraction converted to electricity is relevant

💡 Want to explore quantitative spread models, the fuel switch sensitivity analysis, how to build a spark-spread option valuation, or how RES-surplus hours are collapsing traditional spread logic? Just ask.