What Is LCOE? The Number Behind Every Energy Cost Comparison

When you read that solar is now cheaper than coal, or that offshore wind costs a certain amount per megawatt-hour, the number being quoted is almost always the levelised cost of energy. It is the standard comparison tool in the industry, it is genuinely useful, and it leaves out things that increasingly matter.

The calculation

LCOE answers one question: what must each unit of electricity sell for, on average, to recover every cost across the plant's life?

The formula, in words: the present value of all lifetime costs divided by the present value of all lifetime electricity generated.

Costs include:

• Capital expenditure — equipment, construction, grid connection, development.
• Operations and maintenance across the asset's life.
• Fuel, where applicable.
• Financing costs.
• Decommissioning.

Energy is the total generation expected across the plant's life, after accounting for degradation and availability.

Both sides are discounted to present value, because money and electricity in year twenty are worth less today than money and electricity now.

The inputs that decide the answer

Discount rate. The most influential assumption in the entire calculation. It reflects the cost of capital and required return.

Renewable plants concentrate cost at the beginning and generate revenue for decades. That makes them acutely sensitive to the discount rate — a few percentage points move LCOE substantially, far more than any plausible improvement in module efficiency.

This is the arithmetic behind the observation in why Gulf solar is so cheap: the region's access to low-cost capital matters at least as much as its sunshine.

Capacity factor. It sits in the denominator, so a plant producing more energy from the same capital has a lower LCOE. This is why an expensive offshore wind farm can compete with cheaper onshore capacity — it produces considerably more per megawatt installed, as we explain in capacity factor explained.

Asset life. Spreading capital across 30 years rather than 20 lowers the figure — provided the plant genuinely lasts.

Degradation. Output declining over time reduces lifetime energy, raising LCOE. For solar this is a known, modelled parameter, covered in solar panel degradation and lifespan.

Fuel price. For thermal plants this is a large, volatile and unforecastable input. For renewables it is zero — which is the structural advantage that makes renewable LCOE predictable decades ahead while thermal LCOE is an estimate resting on a fuel price forecast.

What LCOE gets right

It is genuinely valuable for:

• Comparing technologies on a consistent basis.
• Tracking cost trends over time.
• Screening options early in planning.
• Understanding cost structure — which inputs dominate for which technology.

The decline in solar and wind LCOE over the past fifteen years is real, large and well measured. LCOE is the right tool for observing it.

The limitation that matters most

LCOE treats every megawatt-hour as equal. Electricity systems do not.

A megawatt-hour delivered on a summer evening when demand peaks and supply is scarce is worth far more than one delivered at midday into a solar surplus. LCOE assigns them identical value.

At low renewable penetration this barely matters. At high penetration it is decisive. A system with abundant midday solar gains little from more midday solar, however low its LCOE — and this is precisely the situation now facing Gulf and Indian grids, which is why procurement has shifted toward firm and storage-linked products, as we cover in battery storage in the Middle East.

The refinement designed to address this is value-adjusted LCOE, which weights generation by the value of electricity at the time it is produced. It is more honest and considerably harder to calculate, which is why headlines still use plain LCOE.

The costs LCOE leaves outside the fence

LCOE measures a plant. Systems need more than plants.

• Storage to shift generation to when it is needed.
• Transmission to move electricity from resource-rich areas to demand.
• Firm capacity to cover periods when renewables are unavailable.
• Grid services — inertia, frequency response, voltage support — historically provided by thermal plant.
• Curtailment, where generation cannot be used at all.

These are real costs of delivering reliable electricity, and they are not in the plant's LCOE. Comparing a solar LCOE directly against a gas LCOE therefore compares two products that are not the same product — one is variable energy, the other is dispatchable energy.

This is not an argument against renewables, whose system costs are frequently overstated by opponents. It is an argument for comparing honestly.

How to read an LCOE figure critically

• What discount rate was assumed, and does it reflect this market's cost of capital?
• What capacity factor, and is it realistic for this site?
• What asset life, and is it supported by warranties and experience?
• Does it include grid connection and development costs, or only equipment?
• For thermal: what fuel price was assumed?
• Is it value-adjusted, or does it treat all hours identically?

An LCOE quoted without these assumptions is not a finding. It is a headline.

The bottom line

LCOE is the right tool for comparing the cost of building and running generation, and the wrong tool for comparing what that generation is worth to a system. Use it for screening and trend analysis; add timing and integration costs before making a procurement decision on it.

Read the assumptions, not just the number

Energy cost comparisons shape policy and investment, and the assumptions behind them are rarely in the headline.

• Read the policy and markets desk for analysis of project economics.
• Explore energy insights for explainers on the metrics.
• Subscribe to The Energy Edit — independent reporting on Middle East and South Asian energy, free to read. Start here.

Developers, analysts and financiers: reach readers who check the discount rate. Explore partnership.