What is the levelized cost of energy (LCOE)?

The short answer

LCOE answers a simple question: what is the average cost of one megawatt-hour of electricity, accounting for all upfront and ongoing costs across the asset’s life? The numerator includes capital expenditure, fuel, operations and maintenance, and taxes. The denominator is the total MWh delivered. The result is expressed in dollars per MWh.

Lazard’s annual analysis (now in version 18) is the most-cited reference. Their 2024-2025 estimates show unsubsidized utility-scale solar in the US ranging $30-72/MWh, onshore wind $27-73/MWh, gas combined-cycle $45-108/MWh, and coal $69-168/MWh. The IEA reports a 2024 global average solar LCOE around $39/MWh.

The crucial limitation: LCOE does not include the system-level costs of integrating intermittent generation — and those costs change the comparison materially.

→ New to investing concepts? Financial education hub

What the data shows

LCOE estimates from Lazard, IEA, and IRENA converge on similar ranges with technology-specific variation. According to Lazard LCOE+ v17-18 (2024-2025) and IEA data:

• Utility-scale solar US: $30-72/MWh unsubsidized; ~$39/MWh global average 2024 (IEA)
• Onshore wind US: $27-73/MWh, $50/MWh midpoint
• Offshore wind: rising to $91/MWh in 2027 dollars (Dominion CVOW project)
• Gas combined-cycle: $45-108/MWh, depending on capacity factor and gas price
• Vogtle nuclear (units 3-4): ~$31.5 Bn capital cost for 2.2 GW capacity, capacity factor 97%, operating life 60-80 years
• Lazard fuel assumptions: gas $3.45/MMBTU, coal $1.47/MMBTU, nuclear $0.85/MMBTU
• Battery storage 4-hour: ~$120-200/MWh standalone

The exception worth noting: when Lazard calculates total system LCOE including firming costs to provide reliable supply, onshore wind plus firming can rise toward $80-90/MWh in CAISO and similar markets. Solar plus firming can be higher still in winter-peaking grids. The LCOE of the technology itself is not the LCOE of the delivered, reliable kilowatt-hour.

→ Dataset: WTI crude oil price history

Why it happens — the macro mechanism

LCOE is sensitive to three families of inputs that shape its interpretation.

Channel 1 — Cost of capital sensitivity. Renewables are capital-intensive upfront with near-zero marginal cost. The Lazard methodology assumes 60% debt at 8% interest and 40% equity at 12% cost (after-tax WACC ~7.7%). Higher rates push LCOEs up disproportionately for capital-intensive technologies. Lazard’s sensitivity analysis shows that an after-tax IRR/WACC moving from 7.7% to 10% can raise solar LCOE by 25-30%, while raising gas combined-cycle LCOE by less than 10%. Real yields drive the relative cost structure of generation choices more than headline technology cost.

Channel 2 — Capacity factor and firming. A 100 MW solar farm at 23% capacity factor (US average per EIA 2024) delivers the same annual MWh as a 23 MW gas plant at 100% capacity factor. The LCOE divides by total MWh, so it accounts for capacity factor differences. But it does NOT account for the cost of providing the same dispatchable, reliable supply that the gas plant delivers. This is the most underappreciated dimension: comparing renewable LCOEs to thermal LCOEs without firming costs is comparing different products. Lazard introduced “LCOE plus firming” precisely because this comparison was being misused in policy debates.

A short note on subsidy treatment. Lazard reports both subsidized and unsubsidized LCOEs. The Investment Tax Credit (30% for solar) and Production Tax Credit ($27.50/MWh for wind) materially lower subsidized LCOEs and have driven much of the deployment growth in the US since 2022.

Channel 3 — Fuel price assumptions. Gas-fired generation LCOE is highly sensitive to gas price assumptions. Lazard’s $3.45/MMBTU baseline ranges $2.59-4.31/MMBTU for sensitivity. At $7-8/MMBTU (European 2022 levels), gas combined-cycle LCOE roughly doubles. Coal LCOE is similarly fuel-sensitive but with smaller swings. Renewable LCOE has near-zero fuel cost and therefore offers volatility insurance — a feature LCOE does not directly capture.

Synthesis by regime: in the 2010-2020 cost-decline regime, renewable LCOE fell from over $100/MWh to under $50/MWh, driving deployment globally; in the 2022-2024 high-rate regime, capital cost increases offset falling technology costs and slowed deployment in some markets, especially offshore wind; in scenarios with significant carbon pricing, gas combined-cycle LCOE rises by EUR 15-25/MWh per EUR 50/tCO2 of carbon price, materially closing the gap with solar plus firming.

The LCOE of a technology is not the LCOE of a reliable kilowatt-hour delivered to a grid in real time.

→ Framework: Physical constraints on growth

What it means for different economic actors

Savers. LCOE shapes wholesale electricity prices and ultimately retail rates. Falling renewable LCOE has compressed wholesale power prices in many markets during midday solar peaks, with implications for utility revenues and consumer bills.

Investors. LCOE drives expected returns for project developers and financiers. Equity returns on solar/wind projects depend critically on the cost of capital, the contracted PPA price, and the curtailment risk. As the LCOE of building new capacity falls below the marginal cost of operating existing capacity, retirement decisions accelerate for older thermal plants.

Industrial firms. Large industrial energy users increasingly sign virtual PPAs at long-dated LCOE-based prices, locking in input costs for 10-20 years. This contractual structure has become a major channel of corporate decarbonization independent of explicit carbon pricing.

A common error is to use LCOE as a single metric to declare a “winner” technology. LCOE is one input into resource adequacy planning, alongside capacity factor, dispatch profile, transmission availability, and firming requirements. Treating it as the bottom line oversimplifies the planning problem.

Practical observation

Go deeper

Related questions

• Why does intermittency matter for renewable electricity?
• How does battery storage change energy economics?
• What is the nuclear renaissance and will it happen?
• Why are utility companies critical to the energy transition?

Frequently asked questions

Why is LCOE different from the wholesale electricity price?

LCOE is the average cost over an asset’s lifetime, while wholesale prices reflect the marginal cost of meeting demand at each moment. In markets with significant renewable penetration, wholesale prices fall during high-output periods (sometimes negative during midday solar peaks) and rise during low-output periods (evening, no wind). A solar farm with $40/MWh LCOE may sell at much lower or higher wholesale prices depending on dispatch timing, which is why merchant developers increasingly seek long-dated contracts (PPAs) rather than spot exposure.

How does LCOE include firming costs in newer Lazard versions?

Since LCOE+ v15, Lazard publishes a “total LCOE including firming” that adds the cost of capacity needed to ensure resource adequacy. The methodology uses Effective Load Carrying Capability (ELCC) values from grid operators to calculate the additional firm-capacity equivalent needed. Result: onshore wind LCOE of $50/MWh midpoint may rise to $80-90/MWh in CAISO when firming is included, narrowing or eliminating the apparent advantage over gas combined-cycle. This adjustment is technically correct and increasingly used in policy analysis.

Is LCOE useful for comparing nuclear with other technologies?

Nuclear LCOE is dominated by capital cost, with very long operating lives (60-80 years) and high capacity factors (~97% for Vogtle). The Lazard estimate based on Vogtle ($31.5 Bn for 2.2 GW) yields LCOEs significantly higher than gas or renewables — but that calculation is sensitive to overruns specific to first-of-a-kind builds. Earlier nuclear builds in South Korea or France, and emerging SMR designs, suggest substantially lower LCOEs are achievable. The high-leverage role of cost of capital makes nuclear LCOE particularly sensitive to financing structures, which is why government loans (US DOE LPO) and capacity contracts can dramatically alter the headline figure.

Last updated — 21 July 2026