Coal has traditionally been one of the most reliable fuel sources for electricity generation, and in certain respects it still is. Coal plants can run around the clock regardless of weather, they store months of fuel on-site, and their spinning turbines help keep the electrical grid stable. But reliability isn’t just about availability. Rising maintenance costs, an aging fleet, and accelerating plant retirements are changing the picture.
How Often Coal Plants Actually Run
Reliability in the power industry is often measured by “capacity factor,” which is the percentage of time a plant generates electricity compared to its theoretical maximum. Coal plants in the U.S. have historically operated at capacity factors between 40% and 70%, depending on the plant’s age, efficiency, and regional electricity demand. That range has been sliding downward over the past decade as cheaper natural gas and renewables have pushed coal lower in the dispatch order, meaning grid operators call on coal plants less frequently than they used to.
For comparison, wind turbines operated at roughly 34% capacity in 2024, and solar panels came in around the same range. Natural gas plants averaged about 35% for non-fossil-fuel-paired units, though combined-cycle gas plants dedicated to baseload power often run much higher. The key distinction is why each source isn’t running at full capacity. Coal and gas plants can generate power whenever operators need them but may sit idle because the electricity isn’t needed or because cheaper options are available. Wind and solar, by contrast, can only generate when the wind blows or the sun shines, regardless of demand.
Fuel Stockpiles Give Coal a Unique Advantage
One of coal’s strongest reliability arguments is fuel storage. U.S. coal plants held an estimated 124 million short tons of coal on-site as of mid-2024, enough to keep burning for roughly 93 days at typical consumption rates. The EIA projects that stockpiles will range between 90 and 120 days of supply through the end of 2026, about a month more than plants kept on hand between 2019 and 2022.
This matters during extreme weather or supply disruptions. Natural gas plants depend on a continuous pipeline network. When pipelines freeze or demand spikes during a brutal cold snap (as happened during Winter Storm Uri in Texas in 2021), gas-fired plants can lose their fuel supply entirely. Coal plants sitting on a three-month pile of fuel don’t face that particular vulnerability. Solar and wind, of course, carry no fuel at all, which makes them immune to fuel supply disruptions but dependent on weather conditions instead.
Grid Stability and Spinning Turbines
Coal plants contribute something to the grid that solar panels and most wind turbines physically cannot: inertia. The massive steam turbines inside a coal plant spin in sync with the electrical grid’s frequency, typically 60 cycles per second in the U.S. When demand suddenly shifts or a generator trips offline, that spinning mass acts as a buffer, releasing or absorbing kinetic energy to keep the grid’s frequency from changing too quickly.
Solar panels and most wind turbines connect to the grid through electronic inverters, which means they’re electrically decoupled from the grid’s frequency. They can be programmed to mimic some inertia response, but they don’t provide it naturally the way a spinning turbine does. As coal plants retire and more inverter-based generation comes online, grid operators face the challenge of maintaining frequency stability with fewer sources of natural inertia. Some regions are exploring alternatives like synchronous condensers (essentially giant spinning machines that provide inertia without generating power), but this remains an active engineering challenge.
Rising Costs Are Undermining Coal’s Reliability
A power source is only as reliable as its economics allow it to be. If a plant becomes too expensive to maintain, it shuts down, and that’s exactly what’s been happening across the U.S. coal fleet. Operating and maintenance costs for coal plants have ranged from about $20 per megawatt-hour at the cheapest facilities to $40 per megawatt-hour at the most expensive. The plants at the high end of that range have been retiring at a steep pace: 66% of units in the highest cost group are expected to close between 2019 and 2030, compared with 36% of the lowest-cost units.
This creates a reliability paradox. Coal’s physical characteristics (fuel storage, inertia, all-weather operation) make it dependable in theory, but the shrinking fleet means fewer coal plants are available to call on during peak demand or emergencies. Many of the plants still operating are older, which means they require more maintenance and are more prone to unplanned outages. Between 2008 and 2017, the number of operating coal plants in the highest-cost category dropped from 75% of the fleet to 47%, not because they got cheaper but because the expensive ones closed.
How Coal Compares Overall
No single energy source is reliable in every dimension. Coal excels at on-site fuel storage and grid inertia, two qualities that become especially valuable during emergencies. It can generate power on demand, day or night, in any weather. These are real advantages that grid planners still factor into their reliability models.
But coal’s weaknesses are growing. The fleet is aging and shrinking. Maintenance costs are pushing more plants into retirement every year. And while coal can technically run whenever it’s needed, it’s increasingly being outcompeted on price by natural gas and renewables, which means it runs less often and generates less revenue to cover those rising maintenance bills. The result is a feedback loop: higher costs lead to lower utilization, which leads to higher per-unit costs, which leads to more retirements.
For grid reliability as a whole, the concern isn’t really whether coal itself is reliable. It’s whether the grid can replace what coal provided (dispatchable power, fuel security, inertia) as coal plants continue to close. That replacement is happening through a combination of natural gas, battery storage, demand response programs, and grid-forming inverters, but the transition introduces its own reliability risks during the in-between period.