How LCOE and Capacity Factor Decide Which Clean Energy Tech Actually Scales

Every time a solar farm, wind project, or next-generation reactor gets announced, the press release leans on one number: cost per megawatt-hour. That number is usually a levelized cost of energy (LCOE) estimate, and it has become the default scoreboard for clean energy. But LCOE is only half the conversation. The other half, capacity factor, is what separates a technology that looks cheap on paper from one that delivers when the grid needs it.

What LCOE actually measures

LCOE is an accounting tool. It takes the total lifetime cost of a generating asset, including capital expenditure, financing, fuel, operations, and maintenance, and divides it by the total energy the asset is expected to produce over its life. The result is a single price per unit of electricity, typically expressed in dollars per megawatt-hour.

The appeal is obvious. It lets you compare a gas turbine to a solar array to a small modular reactor on one axis. The problem is equally obvious once you look closely: the output denominator depends on assumptions that vary widely between projects, regions, and analysts. Discount rate, assumed lifetime, fuel price trajectory, and expected utilization all move the final figure, sometimes by a wide margin. A change in assumed cost of capital alone can shift the LCOE of a capital-heavy technology more than any improvement in the hardware itself.

This is why two credible LCOE figures for the same technology can disagree. They are not necessarily wrong; they are answering slightly different questions.

Why capacity factor changes the picture

Capacity factor is the ratio of how much energy a plant actually produces to how much it could produce if it ran at full output continuously. A plant rated at 100 megawatts that delivers an average of 40 megawatts over a year has a capacity factor of 40 percent.

This matters because LCOE already bakes in an assumed capacity factor. Lower the assumption, and the same capital cost gets spread across fewer units of energy, pushing the per-megawatt-hour price up. Solar and onshore wind have low fuel costs but variable capacity factors tied to weather and time of day. Nuclear and geothermal tend to run at high, steady capacity factors. Gas peaker plants may have very low capacity factors by design, running only when prices spike.

A technology with a low headline LCOE but an intermittent output profile is not directly comparable to a steady, dispatchable source. The market increasingly recognizes this through concepts like the value-adjusted LCOE, which attempts to weight generation by when it is actually delivered relative to demand.

Where new technologies get tricky

For mature technologies, capacity factors are well documented from years of operating data. For newer entrants, the figures are projections, and projections carry risk.

A first-of-a-kind plant rarely hits its modeled capacity factor in early years. Commissioning delays, unplanned outages, and learning-curve issues push real-world output below the spreadsheet. That gap directly inflates actual delivered cost above the advertised LCOE. Developers tend to quote nameplate-optimistic assumptions; operators discover the difference later.

This is the core reason to treat any LCOE claim for an unproven technology with caution. The number is only as good as the capacity factor behind it, and that capacity factor is a forecast until the plant has run for several years.

How to read the numbers

A few practical checks help separate signal from marketing:

• Ask what capacity factor the LCOE assumes. If it is not stated, the figure is incomplete.
• Check the assumed cost of capital. Capital-intensive technologies are extremely sensitive to it.
• Distinguish dispatchable output from intermittent output before comparing two figures side by side.
• Treat first-of-a-kind projections as ranges, not points.

LCOE and capacity factor are useful precisely because they force assumptions into the open. The trouble starts when a single LCOE number travels through headlines stripped of the assumptions that produced it. For anyone evaluating which clean energy technology scales next, the more revealing question is not what does it cost, but under what conditions, and how confident are we that those conditions hold.