What Is LCOE? Levelized Cost of Energy Explained

LCOE stands for Levelized Cost of Energy. It is the total cost of building and operating a power plant, divided by the total electricity it produces over its lifetime. The result is a single number, expressed in dollars per megawatt-hour ($/MWh), that lets you compare any electricity source against any other on a level playing field.

Comparing the cost of a solar farm to a coal plant is not straightforward. A solar farm costs a lot upfront but nothing to fuel. A coal plant costs less to build but requires a constant supply of fuel and generates emissions that carry financial risk. LCOE solves this by collapsing all costs, capital, fuel, operations, maintenance, decommissioning, and financing, into a single per-unit figure across the entire operating life.

The metric was developed by energy economists to enable apples-to-apples comparison across technologies with fundamentally different cost structures. It is published annually by IRENA, Lazard, the IEA, and BloombergNEF. When energy analysts, investors, and policymakers discuss which technology "wins on cost," they are almost always referencing LCOE data.

The formula is conceptually simple: take all costs over a plant's lifetime and divide by all the electricity it produces, adjusted for the time value of money.

Each variable in the equation has a different weight depending on the technology:

The critical insight: for solar and wind, fuel cost is zero. Once the plant is built, the marginal cost of each additional kilowatt-hour is near zero. For gas, fuel accounts for roughly two-thirds of the lifetime cost. That means gas LCOE is permanently hostage to fuel price volatility. Solar and wind LCOE is locked in at construction.

This cost structure difference is why renewable LCOE keeps declining while fossil LCOE stays volatile. Manufacturing learning curves drive capital costs down. There is no learning curve for the price of gas.

The numbers are no longer close. As of 2023 data (the most recent full-year figures from IRENA and Lazard), solar PV and onshore wind are the cheapest sources of new electricity in most of the world. Not in ideal conditions. Not with subsidies. On a pure unsubsidized cost basis.

Globally, the picture is even more decisive. IRENA's 2024 data shows that 91% of new renewable capacity added worldwide was cheaper than the cheapest fossil fuel alternative available locally. Renewables avoided an estimated $467 billion in fuel costs in 2024 alone.

In emerging markets with high solar irradiance, auction prices have fallen below $30/MWh. India, Brazil, and the Middle East routinely contract utility-scale solar at prices that would have been considered impossible a decade ago. The driver at these levels is no longer panel cost. It is financing cost. Cheaper capital means cheaper solar.

BloombergNEF's 2H 2023 benchmark recorded a record-low global solar LCOE of $41/MWh, a 4% year-on-year improvement. Offshore wind, which struggled with supply-chain inflation in 2022, has returned to cost parity with coal at $81/MWh and is on a downward trajectory again.

Solar PV LCOE has declined 90% since 2010. From $0.445/kWh to $0.044/kWh. That is not a subsidy effect. It is a manufacturing learning curve at global scale.

IRENA's decomposition of solar cost decline identifies the drivers. Module cost reductions account for roughly 45% of the total decline. Engineering, procurement, and construction (EPC) cost reductions contribute another 28%. Inverter improvements add 9%, and racking/mounting systems contribute 9% more. The remainder comes from lower operations and maintenance costs, cheaper financing, and higher capacity factors from better technology.

Onshore wind LCOE declined 69% over the same period, from $0.107/kWh to $0.033/kWh. The mechanism is similar: larger rotors capture more energy per tower, manufacturing at scale drives down unit costs, and better siting and operations squeeze out inefficiency.

The critical point is that these drivers are structural. They are not subsidies that can be withdrawn or commodity prices that can spike. Manufacturing learning curves do not reverse. A factory that has learned to produce solar cells at $0.10/watt does not forget how to do it. Every cost reduction is permanent. Every new gigawatt of deployment makes the next gigawatt cheaper.

Fossil fuels have no equivalent mechanism. The cost of extracting gas or coal depends on geology and commodity markets. There is no learning curve that makes fuel extraction permanently cheaper over time. This asymmetry is why the gap between renewable LCOE and fossil LCOE is widening, not narrowing.

The standard critique of renewable LCOE is fair: solar and wind are intermittent. The sun does not shine at night. Comparing solar LCOE to gas LCOE without accounting for storage is incomplete because gas provides dispatchable power on demand.

Lazard addresses this with LCOE+, which combines solar or wind with 4-hour lithium-ion battery storage. The result: solar plus storage is already competitive with new gas combined-cycle plants and dramatically cheaper than gas peaker plants.

Battery costs are falling on their own learning curve. Lithium-ion battery pack prices dropped from over $1,200/kWh in 2010 to below $140/kWh by 2023, and crossed below $100/kWh in 2024 for the first time. LFP (lithium iron phosphate) chemistry, which uses no cobalt or nickel, is accelerating this trajectory. LFP cells offer 3,000-5,000+ charge cycles versus 1,000-2,000 for NMC, making them ideal for stationary grid storage.

As battery costs continue to decline, the LCOE+ for solar-plus-storage will continue to fall while gas peaker costs stay flat or rise with fuel prices. The crossover already happened. The gap is widening.

LCOE is the most cited metric in energy economics for good reason, but it has genuine limitations. Understanding them is essential to using the metric correctly.

Timing. LCOE does not distinguish between electricity generated at 2 PM (peak demand, high value) and 2 AM (low demand, low value). A solar farm produces most of its output midday. A gas plant can produce on demand. LCOE treats every kilowatt-hour as equal. The IEA's Value-Adjusted LCOE (VALCOE) metric partially addresses this by weighting output by market value at the time of generation.

System costs. LCOE measures the cost at the plant level. It does not include grid upgrades, transmission buildout, or the backup capacity needed to cover intermittent generation. At low renewable penetration (below 30% of the grid), these costs are small. At higher penetration, they become significant. System-level analyses from organizations like the IEA and MIT Energy Initiative show that even with system costs included, renewables remain the cheapest option in most scenarios.

Externalities. Standard LCOE does not include the cost of carbon emissions, air pollution, water consumption, or health impacts. When Lazard models carbon pricing at $20-40 per tonne of CO2, coal LCOE rises from $117/MWh to approximately $147/MWh. Gas CCGT rises by $10-20/MWh. Renewables are unaffected. If externalities were fully priced, the fossil fuel cost disadvantage would be even larger than LCOE shows.

Capacity factor assumptions. LCOE calculations depend on assumed capacity factors, the percentage of time a plant generates at its maximum rated output. Solar capacity factors range from 9% to 35% depending on geography and tracking systems. Wind ranges from 25% to 50%. Using global averages can mask large regional variations. Always check which capacity factor assumptions underpin any LCOE comparison.

These limitations are real. None of them change the fundamental direction of the data. They refine the comparison. They do not reverse it.

LCOE is the metric that settled the economic argument for renewable energy. Not politics, not ideology, not environmental sentiment. Cost per kilowatt-hour.

In 2010, building a solar farm cost ten times what it costs today. The idea that solar would be the cheapest source of electricity in human history was considered optimistic at best, delusional at worst. Every projection from the IEA, EIA, and major energy consultancies underestimated the speed of cost decline. The forecasters kept drawing lines. The costs kept falling below them.

Today, 91% of new renewable capacity added worldwide is cheaper than the cheapest local fossil fuel alternative. Solar plus storage already undercuts gas peakers. The gap is widening. The cost curves that drove this transition are structural, driven by manufacturing learning rates that do not reverse.

This is what LCOE shows. Not a projection. Not a hope. A measurement of what already happened and is continuing to happen. The energy transition stopped being a question of political will and became a question of capital deployment speed. The economics are settled. The deployment is underway. LCOE is the receipt.