What does electricity really cost?
‘Too expensive’ is the most common counter-argument. This page answers it with numbers rather than attitude, and above all shows that ‘cost’ is not one thing: there are four levels that are constantly mixed up.
‘Wind power is cheap’ and ‘electricity prices are high’ can be true at the same time, because they are two different cost levels. Once you grasp that, you read every energy debate differently.
Level 1
Levelised cost of electricity (LCOE)
LCOE state what it costs to generate a kilowatt-hour in a newly built plant, over its whole lifetime. Ground-mounted PV and onshore wind are not only the cheapest renewables here, but the cheapest power plants of any kind.
What LCOE does not include
LCOE measure only generation at the plant. Not included are grid expansion, reserve and backup capacity, storage, and the cost of system integration. And they apply to new plants, a written-off existing plant pays off differently. LCOE are therefore not an electricity price and not a ‘total cost’.
The ranges come from full-load hours (site/weather), investment costs and the CO₂ price, which pushes coal and gas towards the upper end in particular; for hydrogen plants, hydrogen procurement adds to it.
Ranges: Fraunhofer ISE, ‘Levelised Cost of Electricity, Renewable Energies’, July 2024 (reference year 2024). Figures cited, not an open dataset.
Level 2
System costs: and the controversy
An electricity system needs more than generation: grid expansion, reserve, storage and flexibility. These system costs are missing from LCOE, and that is exactly what is disputed: how high they are and to whom they are attributed.
The critique of the ISE study
Critics consider the system costs of renewables understated: the comparison does not attribute the cost of grid, reserve and storage sufficiently (a charge raised in early 2026, among others), and some assumptions, such as full-load hours and capital costs, are chosen in favour of renewables. The reproducibility of individual results is also questioned.
The counter-position
Against this: system costs arise for conventional plants too (fuel logistics, reserve, CO₂ infrastructure), and a rising CO₂ price pushes fossil LCOE further up anyway. The study discloses its assumptions openly and works with ranges, not point values.
The separation matters: the method and definition of system costs are fact; their economic assessment is an open debate, neither side gets the preferred answer here.
Level 3
What consumers pay
The price on the electricity bill is something other than the generation cost. It is made up of several parts:
- Wholesale price, what electricity costs on the exchange (fluctuates hourly; cheap renewables push it down on average).
- Grid fees, for transport and distribution, varying by region.
- Taxes, levies, surcharges, electricity tax and VAT plus various levies.
So cheap generation does not automatically mean a cheap bill: the wholesale price is often only a quarter to a third of the retail price, the rest is grids, taxes and levies.
Level 4
External costs: what society bears
Some costs appear on no bill: environmental and health damage from emissions. Germany's Environment Agency quantifies them in its methodology convention with cost rates, around €200/t CO₂ for climate damage, plus rates for air pollutants. Fossil generation carries by far the highest external costs here.
These are the same pollutants measured in the air-quality module (NO₂, particulates): there as a local health burden, here as economic costs, cumulative and statistical, without single-event attribution.
Four levels, never mixed: LCOE, system costs, consumer prices and external costs are methodically not additive. There is deliberately no single ‘total cost’ figure here. It would obscure more than it explains.
Sources: Fraunhofer ISE, Levelised Cost of Electricity, Renewable Energies (July 2024, reference year 2024); German Environment Agency, methodology convention for environmental costs (cost rates); price components via Energy-Charts (Fraunhofer ISE). System-cost level shown qualitatively. Figures cited with attribution, no self-estimated value (ADR 0114).