Coal Power: The Technology Everything Else Is Measured Against
It is tempting to treat coal as a solved problem being phased out. Globally it is neither. Coal-fired generation reached record levels in the 2020s, and the average Asian coal plant is young enough that it could run for another thirty years. Any honest account of the energy system has to start by describing what is actually there.
From Rock to Rotation
A coal plant is a machine for boiling water. Coal arrives by rail or conveyor, is pulverised to a powder finer than flour so it burns almost like a gas, and is blown into a furnace where it burns at around 1,500 degrees. The furnace walls are lined with tubes carrying water, which absorbs the heat and becomes steam.
That steam, at up to 600 degrees and 250 or more atmospheres, drives a turbine through several stages of decreasing pressure. The turbine turns a generator, and the electricity leaves through a transformer. The exhausted steam is condensed back to water in a condenser cooled by a river, the sea or a cooling tower, and the cycle repeats. This is the Rankine cycle, and it is the same architecture used by nuclear fission, biomass and concentrating solar plants - only the heat source differs.
The energy density involved is what made coal the foundation of industrial power. A kilogram of hard coal holds 24 to 30 megajoules, enough to keep a household kettle running for hours. It is dense, storable, transportable by ship and rail, and it sits in a pile in the yard without leaking, evaporating or needing to be kept cold.
That last property is underrated in comparisons. A coal plant carries weeks of fuel on site, so it is indifferent to weather, time of day and season - the capacity factor issue that defines wind and solar does not arise. What it trades for that is everything that comes out of the chimney.
Why Efficiency Stops Around 45 Percent
The theoretical ceiling on any heat engine is the Carnot efficiency, set by the temperatures between which it operates. A plant taking steam at 600 degrees and rejecting heat at 30 has a Carnot limit around 65 percent. Real plants reach 45 percent at best, because real turbines, pumps and pipes have friction and real heat exchangers need temperature differences to work.
Efficiency therefore rises with steam temperature, and the whole history of coal plant development is the history of pushing that temperature up. Subcritical plants operate below the critical point of water, around 540 degrees, and reach 33 to 37 percent. Supercritical plants exceed 374 degrees and 221 bar, where water ceases to distinguish between liquid and vapour, and reach 40 to 42. Ultra-supercritical designs push to 600 degrees and beyond for about 45.
The obstacle above that is metallurgy, not thermodynamics. Boiler tubes and turbine blades at 700 degrees under 300 atmospheres for thirty years require nickel-based superalloys whose cost has so far exceeded the value of the efficiency gained. Advanced ultra-supercritical programmes have been pursued in Europe, Japan and China for two decades without reaching commercial deployment.
The gap between the best and the average matters more than the ceiling. Global fleet efficiency averages around 37 percent, so a substantial share of world coal generation runs at technology that is decades old. Raising the average toward the best available would cut emissions by roughly a fifth for the same electricity, which is one of the larger and least discussed levers available.
What Comes Out
Carbon dioxide is unavoidable. Burning carbon produces it, and no combustion process can avoid the chemistry. A coal plant emits roughly 820 grams per kilowatt-hour on a life-cycle basis - against about 490 for natural gas, 12 for wind, 11 for nuclear and 41 for solar photovoltaics. Reducing it requires either burning less coal or capturing the product, which is what carbon capture attempts.
Everything else in the flue gas can be controlled, and in wealthy countries largely has been. Sulphur dioxide, which causes acid rain, is removed by flue gas desulphurisation, typically by reacting it with limestone to produce gypsum. Nitrogen oxides are reduced by staged combustion and selective catalytic reduction. Particulates are captured by electrostatic precipitators and fabric filters at efficiencies above 99 percent.
Mercury is the awkward one. Coal contains trace mercury which volatilises in the furnace and is difficult to capture, and coal combustion is among the largest human sources of atmospheric mercury worldwide.
The solid residue is ash - roughly 10 percent of the coal by mass, containing the trace elements that were in the rock, including small amounts of uranium and thorium. Much of it is used in cement and concrete, which is genuinely useful, and the rest is stored in ponds or landfills where containment is the ongoing issue.
Why It Persists
Coal generation has fallen sharply in Europe and North America and risen in Asia, and the global total reached record levels in the 2020s. The explanation is not that anyone regards coal as good. It is that the plants exist, they are paid for, and many of them are young.
The average coal plant in China and India is around 15 years old against a design life of 40 or more. Closing one is not a matter of ceasing to build; it is writing off an asset with decades of value remaining, in economies where electricity demand is still rising and where the plant is often owned by a state or a bank that lent against it. That is a financial and political problem, and it is why coal retirement schedules are negotiated rather than engineered.
The second reason is what coal provides beyond energy. It runs regardless of weather, it carries its own fuel store, and its spinning turbines supply the grid inertia that stabilises frequency. A grid replacing coal needs to replace those services too, not merely the kilowatt-hours - which is why the discussion turns quickly to storage, transmission and sources that run continuously.
The third is employment and region. Coal mining and generation are concentrated geographically, so closure falls on specific towns rather than being spread thinly. Every serious transition programme has had to treat that as a central problem rather than an afterthought, and the ones that did not have generally failed politically.
Frequently asked questions
How efficient is a coal power plant?
A modern ultra-supercritical plant converts about 45 percent of the fuel's energy into electricity. Older subcritical plants reach 33 to 37 percent, and the global fleet averages around 37. The limit comes from the Carnot relationship between steam temperature and condenser temperature, and from what boiler alloys tolerate.
Why can't coal plants be made much more efficient?
Efficiency rises with steam temperature, but tubes and turbine blades running at 700 degrees and 300 atmospheres for decades require nickel superalloys that cost more than the efficiency gain is worth. Advanced ultra-supercritical programmes have run for twenty years without commercial deployment.
How much carbon dioxide does coal produce?
Roughly 820 grams per kilowatt-hour on a life-cycle basis, compared with about 490 for natural gas, 41 for solar photovoltaics, 12 for wind and 11 for nuclear. Coal supplies about a third of global electricity and around 40 percent of energy-related carbon dioxide emissions.
Can coal be burned cleanly?
Sulphur dioxide, nitrogen oxides and particulates can be removed from flue gas at high efficiency, and in wealthy countries largely are. Carbon dioxide cannot be removed by combustion chemistry - it is the product. Capturing it requires carbon capture equipment, which adds cost and consumes part of the plant's output.
Why is coal use still rising globally?
Because the Asian fleet is young - around 15 years old on average against a 40-year design life - and demand is still growing. Closing a plant means writing off decades of remaining value, usually against state or bank balance sheets. The constraint is financial and political rather than technical.
