What sets electricity prices? The merit order explained
The generator needed to serve the last unit of demand can determine the price paid to every dispatched generator. Here is how that works in PowerLab’s simplified market.
A different generation mix does not always mean a different electricity price. The price changes when the marginal cost of the last generator needed changes.
What is the merit order?
The merit order ranks available generators from lowest to highest marginal cost: the additional operating cost of producing one more MWh. In PowerLab’s Market Lab, each technology bids at that cost. The market dispatches the cheapest output first until it covers demand.
Read the chart horizontally as cumulative available capacity in GW and vertically as cost in €/MWh. A wider block means more capacity, not a higher price. The generator where the demand line meets the supply curve is marginal. It can be only partly dispatched.
This is a one-hour, single-zone model with divisible output and no transmission constraints. It is an educational starting point, not a complete description of every electricity market.
A worked example: why the price is €79.75/MWh
Use the default assumptions: 60 GW demand, €30/MWh gas fuel, €60 per tonne of CO₂, and 50% wind availability. Wind contributes 15 GW from 30 GW installed.
= €79.745…/MWh of electricity
Coal costs 30 + 60 × 0.90 = €84/MWh. Gas therefore comes before coal in this example.
| Technology | Available | Dispatched | Cost |
|---|---|---|---|
| Wind | 15 GW | 15 GW | €0/MWh |
| Nuclear | 20 GW | 20 GW | €12/MWh |
| Combined-cycle gas | 30 GW | 25 GW | €79.75/MWh |
| Coal | 25 GW | 0 GW | €84/MWh |
Wind and nuclear cover 35 GW; gas supplies the remaining 25 GW. The more expensive gas peaker is also unused. Combined-cycle gas sets the clearing price because it is the last technology needed.
Why more wind may leave the price unchanged
Raise wind availability from 50% to 60%, keeping the other inputs fixed. Wind output rises from 15 to 18 GW and gas dispatch falls from 25 to 22 GW. Gas is still marginal at the same cost, so the clearing price stays approximately €79.75/MWh.
Emissions fall even though the price does not. At the model’s gas emissions factor of 0.37 tonnes per MWh, replacing 3 GW of gas for one hour avoids 3,000 × 0.37 = 1,110 tonnes of operational CO₂. This is a calculation within the illustrative fleet, not an emissions forecast.
The market price is not the same as profit
Under uniform pricing, every dispatched generator receives the clearing price. Nuclear’s hourly operating margin is therefore (79.745… − 12) × 20,000 MWh, approximately €1.355 million. GW must be converted to MW before multiplying by the one-hour duration.
That operating margin excludes fixed and capital costs. It is not total profit or evidence that the plant is cheap to build. If demand exceeds available capacity, PowerLab reports unmet demand and leaves price and margins unavailable rather than inventing a scarcity price.
Retail bills are a separate question: EIA’s overview explains that fuel, power plants, networks, weather and regulation all influence electricity prices. A simulated wholesale clearing price is not a household tariff.
Put the idea to work.
Start at the default Market Lab settings, increase wind availability to 60%, and compare the price, gas dispatch and emissions. Then lower demand until gas is no longer needed.
Try the merit-order simulator →Common questions
Does the most expensive generator always set the price?
In PowerLab’s uniform-price model, the last generator needed sets the price. A more expensive plant that is not dispatched does not set it.
Does more wind always lower the price?
No. It can reduce thermal dispatch while leaving the marginal generator and its marginal cost unchanged.
Sources & model boundary
Examples and calculations: PowerLab’s illustrative model, checked against the local simulation. Background reading: EIA: factors affecting electricity prices.
This guide is educational. Grid Operator’s network, economics and construction are simplified game assumptions; Market Lab’s uniform-price model is a separate one-hour experiment.