3. Coal & Thermal Plants

Module 2 ended with a promise: every power plant is a machine for spinning magnets, and the rest of the course is a tour of ways to turn the shaft. We start with the way that still does most of the work in India.

In the financial year 2024-25, coal produced about 1,332 of India’s 1,824 terawatt-hours of electricity — roughly 73% (Central Electricity Authority figures). Whatever you feel about that number, if you want to understand how India’s lights stay on, you begin here.

The world’s most elaborate kettle

Strip away the scale and a coal plant does something you have done yourself: it boils water and uses the steam. That’s it. The rest is engineering.

Here is the loop, in seven steps:

1. Crush. Coal arrives by the trainload and is ground into a powder as fine as talcum. A lump burns only on its surface; a powder burns almost instantly.
2. Burn. The powder is blown into a furnace and ignites in a fireball at roughly 1,300–1,500 °C.
3. Boil. The furnace walls are lined with kilometres of water-filled tubing — the boiler. The water inside flashes to steam.
4. Superheat. The steam is pushed through more tubing in the hottest gases until it is far past boiling — around 540–600 °C of superheated steam, at pressures over 150 times atmospheric.
5. Spin. That steam is released against the blades of a steam turbine, which spins at 3,000 rpm — and on the same shaft sits the generator from Module 2. This is the only step that makes electricity. Everything else exists to serve it.
6. Condense. The spent steam meets the condenser: a wall of tubes carrying cool water, which turns the steam back into liquid.
7. Pump and repeat. A feed pump pushes the water back up to boiler pressure, and the loop closes. Forever.

Engineers call this the Rankine cycle. Watch it run:

‹ Step back ▶ Play Step on ›

Stage 1 of 4

1. Boiler

Burning coal boils water into high-pressure steam, then superheats it well past boiling point.

Working fluid: water → superheated steam

Two things are worth pausing on. First, the water is never used up — the same purified water goes round and round for years; what gets consumed is coal and cooling water. Second, that condenser is not a nicety. Condensing steam collapses to about one thousandth of its volume, leaving a partial vacuum that actively pulls steam through the turbine. Without a condenser, a steam plant loses a large slice of its output.

A tour of the actual building

Now the parts you can see from the road.

The tall thin chimney carries the flue gas — what’s left after burning: mostly nitrogen and carbon dioxide, plus pollutants we’ll get to. It is tall to disperse what remains high into the air.

The fat hourglass-shaped towers are cooling towers, and they answer our hook question. The condenser needs a cold side, so cooling water absorbs the waste heat and is then sprayed down through the tower. Some evaporates, carrying heat away; warm damp air rises out of the top and, meeting the cooler outside air, condenses into a visible white cloud.

That plume is a cloud. It has nothing to do with combustion. Same physics as your breath on a cold morning, or a boiling kettle. When a news photograph wants to illustrate pollution, it almost always shows cooling towers — the harmless bit — because the genuinely polluting exhaust is largely invisible.

Why the turbine comes in three sizes

Look closely at a turbine hall and you will notice the turbine is not one machine but three, in a row on a single shaft, each larger than the last. There is a reason, and it follows directly from Module 1.

Steam entering at 565 °C and enormous pressure is dense and compact. As it gives up energy to the blades it cools and expands — dramatically. By the time it reaches the condenser it occupies hundreds of times more volume than it did at the start. A single turbine sized for the entrance would be hopelessly choked at the exit; sized for the exit, it would be absurdly oversized at the entrance.

So the expansion is handled in stages:

• High-pressure (HP) turbine — small blades, taking the first, fiercest bite.
• Intermediate-pressure (IP) turbine — larger blades, steam now roomier.
• Low-pressure (LP) turbine — huge blades, over a metre long, whirling steam that is by now thin and fast-moving. These are the enormous fan-like rotors in photographs.

Between HP and IP the steam usually takes a detour called reheat: it goes back to the boiler to be heated again before continuing. This raises the average temperature at which heat enters the cycle — which, as the 🔥🔥 block will show, is exactly the lever Carnot says to pull — and it keeps the steam dry. Wet steam matters more than it sounds: water droplets striking blades at hundreds of metres per second erode them like sandblasting.

What actually comes out

Being honest about coal means naming all of it.

• Carbon dioxide. Unavoidable — it is what burning carbon produces. Roughly a tonne of CO₂ per megawatt-hour for a typical coal plant. No filter removes it; this is the climate problem in one line, and Module 13 takes it up properly.
• Sulphur dioxide (SO₂), which causes acid rain and lung disease. This one can be removed: flue-gas desulphurisation sprays limestone slurry through the exhaust, capturing sulphur as gypsum — which is then sold to make plasterboard.
• Nitrogen oxides (NOx), formed when the air’s own nitrogen burns at furnace temperatures. Controlled by clever burner design and chemical treatment.
• Fly ash — the mineral residue, and Indian coal produces a lot of it (high ash content is a known feature of domestic coal). An electrostatic precipitator charges the ash particles so they stick to collector plates, capturing well over 99% of them. Captured fly ash is genuinely useful — it goes into cement, bricks and road embankments. What isn’t used ends up in ash ponds, which have caused serious local pollution when they leak or breach.

The honest summary: the local pollutants can be engineered down a long way, at a cost, and India’s FGD retrofit programme has been repeatedly extended as plants struggle to comply. The global pollutant, CO₂, cannot be filtered at all.

India’s coal machine

Coal is not an abstraction here; it is an enormous physical logistics operation.

Vindhyachal, NTPC’s station in Singrauli district, Madhya Pradesh, is India’s largest power station at 4,783 MW across 13 coal units. Its first unit started in 1987 and its last in 2015 — a plant built across three decades.

Mundra, in Kutch district, Gujarat, tells the other half of the story. Adani Power’s station there runs at 4,620 MW (four 330 MW units and five 660 MW units) largely on imported coal, much of it from Indonesia — which is why it was built on the coast, with its own port. That single design decision ties its economics to the international coal price and the rupee-dollar exchange rate, and when global coal prices spiked, imported-coal plants found themselves generating electricity that cost more than their contracts allowed them to charge. Where a plant sits, and whose coal it burns, turns out to matter as much as how efficiently it runs.

To feed stations like it, coal moves by rail on a scale that shapes the whole country. Coal is Indian Railways’ single biggest freight commodity, at roughly 45% of freight revenue — which is why the railway’s finances and the power sector’s are quietly joined at the hip, and why decarbonisation is a railway problem as much as an energy one.

Coal also builds towns. Singrauli, Korba, Talcher, Dhanbad — places where the mine, the plant, the railway siding and everyone’s livelihood are the same story. When Module 13 discusses a “just transition”, these towns are what the phrase actually means.

And the ground is shifting. In FY 2025-26, coal-fired generation in India fell by about 4% — its first decline in five years, and the first driven by clean-energy growth rather than by a collapse in demand — as solar and wind additions surged. Coal has already dropped below half of installed capacity. It remains, for now, comfortably the largest source of actual electricity.

🔥Why a plant that burns hotter earns more: subcritical to ultra-supercriticalAdvanced · optional

Coal plants are sorted into three grades, and the difference is simply how hot and how hard they push the steam.

Water has a critical point at about 22.1 megapascals (roughly 218 atmospheres) and 374 °C. Below it, water and steam are distinct: you can point at a boiling surface. Above it, the distinction vanishes — you get a single fluid that is neither quite liquid nor quite gas.

• Subcritical plants stay below that pressure. Most of India’s older fleet. Efficiency roughly 33–38%.
• Supercritical plants run above it, typically around 24 MPa and 565 °C. No boiling drum needed, since there is no boiling to separate. Efficiency around 40–42%.
• Ultra-supercritical pushes further still — around 600 °C and beyond — reaching roughly 45%.

The prize is worth chasing: raising efficiency from 35% to 45% means about 20% less coal, and 20% less CO₂, for exactly the same electricity. India has mandated supercritical or better for new large units for years now.

So why not go hotter still? Metallurgy. At 700 °C the alloys strong enough to hold pressurised steam for 30 years become exotic and eye-wateringly expensive. Every increment is a materials-science fight, and that is the real frontier of coal engineering.

🔥🔥The Carnot ceiling — and why 'waste' heat is not wasteDeep · optional

In 1824 a young French engineer, Sadi Carnot, proved something extraordinary about heat engines: their maximum possible efficiency depends on nothing except the two temperatures they work between. Not the fuel, not the working fluid, not the cleverness of the design.

η_max = 1 − T_cold / T_hot (temperatures in kelvin)

Put in a modern supercritical plant: steam at 565 °C = 838 K; condenser at 40 °C = 313 K.

η_max = 1 − 313/838 ≈ 0.626

Even a perfect, frictionless, infinitely well-engineered plant could convert only about 63% of the heat into work. Real plants achieve 40–45% because of unavoidable irreversibilities: friction, turbulence, finite-rate heat transfer, energy spent driving the plant’s own pumps and mills.

The deep reason is the second law of thermodynamics. A heat engine does not consume heat; it lets heat fall from hot to cold and takes a cut on the way, exactly as a water turbine takes a cut from falling water. To keep the flow going you must dump heat at the cold end. On a temperature–entropy (T–s) diagram, the cycle’s enclosed area is the work you get; the rectangle beneath it, running out to the condenser temperature, is the heat you are obliged to throw away. It is a licence fee for extracting work at all.

This reframes two things:

1. The condenser’s waste heat is not sloppiness — it is the price of the second law. That is why a coal plant discards roughly half of its fuel’s energy as warm water and moist air, and why no invention will ever eliminate it.
2. It explains the whole design logic of Part B. Want more efficiency? Raise T_hot (supercritical boilers, and the gas turbines of Module 4, which enter at ~1,400 °C) or lower T_cold (cold seawater, cooler climates). There is no third option, because Carnot says there isn’t.

One escape hatch exists: stop treating the low-grade heat as electricity’s failed by-product and use it — for district heating or industrial processes (cogeneration). You can’t beat Carnot, but you can decline to waste what Carnot forces you to reject.

What comes next

You now know the machine that made the modern world, in the form that still dominates India’s grid — and the exact reason it can never be much better than 45% efficient. Module 4 keeps the flame but changes the engine: gas turbines, which are jet engines bolted to the ground, and the ingenious trick of bolting a steam plant onto their exhaust to beat coal’s ceiling outright.