CompanyGraph
Capacity factor measures the structural gap between what a power source could theoretically produce and what it actually delivers — a gap that reflects physics and economics, not quality or efficiency.
The Structural Question
When countries, companies, or commentators discuss energy capacity, they frequently cite nameplate capacity — the maximum rated output of a generator in megawatts or gigawatts. A headline announcing “500 MW of new solar capacity” sounds equivalent to one announcing “500 MW of new nuclear capacity.” Both produce 500 megawatts at their rated maximum.
But nameplate capacity describes potential, not performance. What matters for an energy system is not how much power a source can theoretically produce at its peak, but how much energy it actually delivers over time. Capacity factor is the metric that bridges this gap. It answers a specific question: what fraction of its theoretical maximum did this source actually produce?
The answer varies enormously by technology, and the variation is not random. It reflects the physical operating characteristics of each generation type — whether it runs continuously, intermittently, or on demand. Understanding capacity factor is essential for interpreting any comparison of energy sources, costs, or system adequacy that involves nameplate capacity figures.
The Arithmetic: What Capacity Factor Measures
Capacity factor is calculated by dividing actual energy output over a period by the theoretical maximum output over that same period. The theoretical maximum assumes the source runs at full rated output for every hour of the period.
For a 100 MW source over one year (8,760 hours), the theoretical maximum output is 876,000 MWh, or 876 GWh. If the source actually produced 175 GWh, its capacity factor is 175 divided by 876, or approximately twenty percent. If it produced 806 GWh, its capacity factor is approximately ninety-two percent.
The simplicity of this calculation masks the structural information it encodes. A twenty percent capacity factor and a ninety-two percent capacity factor describe fundamentally different relationships between a source and the grid it serves. The first source delivers energy for roughly one-fifth of the hours in a year (or at partial output for more hours). The second delivers energy for nearly every hour. The implications for grid planning, backup requirements, and cost per unit of energy are correspondingly different.
Capacity Factors by Technology: What the Numbers Reflect
Nuclear power plants typically achieve capacity factors of ninety to ninety-three percent in well-operated fleets. The remaining seven to ten percent represents planned refueling outages (typically every eighteen to twenty-four months, lasting several weeks) and occasional unplanned maintenance. Nuclear’s high capacity factor reflects its design purpose: continuous operation at constant output. The technology is optimized for maximum utilization because its economics depend on spreading enormous capital costs across as many megawatt-hours as possible.
Geothermal plants achieve similar capacity factors — often exceeding ninety percent — because the underground heat source flows continuously and predictably. Geothermal does not depend on weather, seasons, or fuel deliveries. Its capacity factor reflects the reliability of the geological heat source and the mechanical availability of the surface equipment.
Coal-fired plants historically operated at capacity factors of seventy to eighty-five percent in their baseload role. As coal has been displaced from baseload by cheaper gas and renewables in many markets, coal plant capacity factors have declined to forty to sixty percent in some regions, reflecting their shift to mid-merit operation where they run during peak hours but are shut down when cheaper sources are available. The declining capacity factor of coal plants is not a technical deterioration — it is an economic signal that their position in the merit order has changed.
Combined cycle gas turbines typically operate at forty to sixty percent capacity factor, reflecting their mid-merit role: they run during most daytime hours but may be shut down overnight when demand drops and cheaper sources suffice. Their capacity factor reflects their economic position in the dispatch stack, not a physical limitation — a combined cycle plant could run continuously if the economics justified it.
Onshore wind turbines achieve capacity factors of twenty-five to forty-five percent, depending on the wind resource at their location. The variation is determined by geography: turbines in consistently windy locations (coastal areas, plains, mountain passes) achieve higher capacity factors than those in less windy terrain. The capacity factor reflects how often and how strongly the wind blows, a physical parameter of the site rather than a property of the technology.
Offshore wind turbines achieve higher capacity factors than onshore — typically thirty-five to fifty-five percent — because offshore wind resources are generally stronger and more consistent. The additional capacity factor partially compensates for the substantially higher capital and maintenance costs of offshore installations.
Solar photovoltaic systems achieve capacity factors of fifteen to twenty-five percent in most locations, reflecting the daily cycle (no generation at night) and seasonal variation (less generation in winter). Desert locations with intense, consistent sunlight achieve the higher end of this range. Northern latitudes with short winter days achieve the lower end. Cloud cover introduces additional variation within these geographic constraints.
Why Low Capacity Factor Does Not Mean Low Quality
The instinct to interpret low capacity factor as poor performance misreads what the metric measures. Capacity factor describes the relationship between a source and time — how much of the time it produces, and at what fraction of its maximum. It does not describe how well the source performs when it does produce.
A modern solar panel converts approximately twenty to twenty-two percent of incident sunlight to electricity — its conversion efficiency. This efficiency is a measure of the technology’s quality. But the solar panel produces zero electricity at night, regardless of its conversion efficiency, because there is no sunlight to convert. Its capacity factor of fifteen to twenty-five percent reflects the availability of its input (sunlight), not the quality of its conversion.
Similarly, a wind turbine may have an aerodynamic efficiency near the theoretical maximum (the Betz limit of fifty-nine percent). It captures wind energy as effectively as physics permits. But when the wind does not blow, the turbine produces nothing. Its capacity factor reflects wind availability, not turbine quality.
A gas peaking plant may have a capacity factor of ten percent — it runs only during the highest-demand hours. This low capacity factor does not indicate poor performance. It indicates that the plant’s economic role is to provide power only during peak periods. It is available ninety-five percent of the time but dispatched only ten percent. The gap between availability and utilization reflects its position in the merit order, not a limitation of the technology.
Curtailment: When the System Reduces Capacity Factor
A source’s capacity factor can be reduced not by any limitation of the source itself but by constraints in the system it connects to. This phenomenon — curtailment — occurs when a generator is ordered to reduce output because the grid cannot absorb all the power being produced.
Curtailment is increasingly common for wind and solar generators. When midday solar floods the grid with more power than demand requires, grid operators instruct solar farms to reduce output or disconnect entirely. The solar panels are still producing — photons are still hitting cells — but the electricity has nowhere to go. The curtailed energy is wasted, and the generator’s capacity factor drops.
Curtailment can also result from transmission bottlenecks. A wind farm in a remote area may be able to produce its full output, but the transmission line connecting it to the demand center may lack the capacity to carry all of it. The excess must be curtailed. The capacity factor drops not because of the wind resource or the turbine technology but because of infrastructure constraints between the source and the consumer.
In some markets, curtailment rates for renewables have reached significant levels. China has experienced wind curtailment rates exceeding fifteen percent in some provinces. California has curtailed billions of kilowatt-hours of solar annually. Each curtailed megawatt-hour reduces the capacity factor of the affected generators and worsens their economics — the capital cost is the same, but fewer megawatt-hours of revenue are produced to cover it.
Curtailment-driven capacity factor reduction is a system failure, not a source failure. It indicates that the grid’s transmission, storage, or demand flexibility is insufficient to absorb the generation that available sources could produce. Treating curtailment-reduced capacity factor as a property of the generation technology rather than the system misattributes the constraint.
The Cost Implication: Capital Cost per MWh vs Capital Cost per MW
Capacity factor directly determines the relationship between the capital cost of building a generator (measured per MW of capacity) and the capital cost of the energy it produces (measured per MWh of output). This relationship is where the structural significance of capacity factor becomes most visible.
Consider two sources, each costing one billion dollars per GW of nameplate capacity. Source A operates at ninety percent capacity factor, producing 7,884 GWh per year. Source B operates at twenty percent capacity factor, producing 1,752 GWh per year. Per megawatt of capacity, they cost the same. Per megawatt-hour of energy, Source A costs one-quarter what Source B costs in capital terms, because it produces 4.5 times as much energy from the same capacity.
This arithmetic is why comparisons of energy costs that cite only capital cost per MW of capacity are structurally misleading. Solar may be cheaper per MW of capacity than nuclear, but the comparison reverses or narrows when measured per MWh of energy produced, because the capacity factors differ by a factor of four to five. Neither comparison alone tells the full story — solar’s lower capital cost per MW may outweigh its lower capacity factor, or it may not, depending on the specific costs — but any comparison that uses only one metric without adjusting for capacity factor is incomplete.
The levelized cost of energy (LCOE) attempts to resolve this by dividing total lifetime costs (capital, fuel, maintenance, financing) by total lifetime energy output. Capacity factor enters this calculation directly: lower capacity factor means fewer megawatt-hours over which to spread costs, increasing the LCOE. But LCOE itself has limitations — it does not account for when energy is produced, how dispatchable it is, or what system costs (storage, backup, transmission) are needed to integrate it. Capacity factor is a necessary input to cost comparison but not a sufficient one.
Utilization vs Availability: The Peaking Plant Paradox
A gas peaking plant illustrates why capacity factor alone can mislead. Such a plant may be mechanically available ninety-five percent of the time — ready to generate on a few minutes’ notice. But it may actually run only five to fifteen percent of the time, during peak demand hours when its high marginal cost is justified by high market prices.
Its low capacity factor does not indicate unreliability. The plant is highly reliable — it starts and produces when called upon. Its low utilization reflects the economic structure of the market: it earns enough revenue during peak hours to cover its costs even though it sits idle most of the time. The gap between availability and utilization is the source of its value — it provides capacity precisely when the system most needs it.
This distinction matters for grid planning. A system that needs ten GW of reliable capacity during peak hours cannot substitute ten GW of solar (at twenty percent capacity factor) and expect the same reliability, because the solar may not produce during peak hours. But it also cannot substitute ten GW of baseload nuclear, because nuclear cannot ramp quickly to cover sudden demand surges. The peaking plant’s value lies in its combination of high availability and dispatchability, not in its capacity factor.
Where This Appears Across Energy Systems
Capacity factor differences create structural patterns across energy systems that become visible when nameplate capacity is translated to actual generation.
Germany has installed over 150 GW of combined wind and solar nameplate capacity, but these sources generate roughly thirty to forty percent of the country’s electricity in a typical year. The gap between installed capacity and generation share reflects average capacity factors of roughly twenty percent for solar and twenty-five percent for onshore wind. The remaining sixty to seventy percent of generation comes from sources with higher capacity factors — coal, gas, nuclear (now phased out), and biomass.
The United States has approximately 100 GW of nuclear nameplate capacity that generates roughly twenty percent of the country’s electricity — a share disproportionate to its capacity because of nuclear’s high capacity factor. By contrast, solar capacity has grown rapidly but contributes a smaller share of generation than its capacity additions might suggest, because each GW of solar produces roughly one-fifth the annual energy of a GW of nuclear.
China’s massive renewable buildout illustrates the capacity factor dynamic at scale. China has installed more wind and solar capacity than any other country, yet coal still generates roughly sixty percent of its electricity. The high capacity factor of coal plants (running continuously as baseload and mid-merit) means that each GW of coal capacity contributes several times the annual energy of each GW of solar or wind.
Offshore wind’s higher capacity factor compared to onshore explains some of the interest in offshore development despite its substantially higher capital cost. A 1 GW offshore wind farm at forty-five percent capacity factor produces roughly 3,940 GWh per year, compared to roughly 2,630 GWh from a 1 GW onshore farm at thirty percent. The additional capital cost of building offshore must be weighed against this roughly fifty percent increase in energy output per MW installed.
Diagnostic Boundaries
This article describes capacity factor as a structural metric that reveals the gap between installed capacity and delivered energy. It does not resolve several questions that require analysis beyond this observation.
Capacity factor alone cannot determine which generation technology is superior or more cost-effective. That determination requires evaluating capital costs, fuel costs, maintenance costs, system integration costs, environmental costs, and the value of the specific services each technology provides (continuous generation, dispatchability, flexibility, emissions profile). Capacity factor is one input among many.
The article does not assess whether capacity factors for specific technologies will improve over time. Solar panel efficiency improvements, taller wind turbines accessing stronger winds, advanced nuclear designs with higher thermal efficiency, and enhanced geothermal systems all have trajectories that may shift capacity factors. The current ranges described are observed values, not permanent physical limits.
The article cannot determine what capacity factor is “good enough” for a given technology to be economically viable. That threshold depends on the capital cost per MW, the value of the energy produced (which depends on when it is produced), the availability and cost of complementary technologies (storage, backup), and the regulatory and market framework. A twenty percent capacity factor may be fully economic if capital costs are low enough and system costs are manageable, or it may be insufficient if those conditions do not hold.
Capacity factor describes a structural relationship between potential and actual output. It makes visible the arithmetic that nameplate capacity comparisons obscure. How that arithmetic translates into economic viability, system adequacy, or technology preference depends on variables that capacity factor alone does not capture.