4. Gas, Diesel & Combined Cycle

Module 3’s coal plant keeps its fire outside the working fluid: flames heat a sealed water loop, and the two never meet. This module takes the opposite approach — put the fire inside the engine — and then does something so neat that it produces the most efficient fuel-burning machine humanity has ever built.

It also asks an uncomfortable Indian question: why does a country short of clean, flexible power leave roughly 25 gigawatts of exactly that sitting idle?

A jet engine, bolted to the ground

Picture an aircraft engine. Now bolt it to a concrete pad, point the shaft at a generator instead of a fan, and you have a gas turbine. This is not an analogy — the industrial machines are direct descendants of aviation engines, and some are literally aircraft engines adapted for the job.

Three steps, in one continuous flow:

1. Compress. A compressor — dozens of rows of blades on the shaft — squeezes incoming air to perhaps 20–30 times atmospheric pressure. Squeezing heats it, to several hundred degrees, before any fuel is added.
2. Burn. Natural gas is injected into that compressed hot air and ignites continuously, reaching roughly 1,400 °C — far hotter than a coal plant’s steam.
3. Expand. The hot, high-pressure gas blasts through turbine blades and spins the shaft. Part of that rotation drives the compressor at the front; the rest turns the generator.

Engineers call this the Brayton cycle, and the contrast with Module 3’s Rankine cycle is worth holding onto. Rankine boils and condenses the same water forever in a closed loop. Brayton takes in fresh air, burns fuel in it and throws the products out of the back — an open loop.

The practical consequence is speed. A coal plant is a mountain of hot metal; heating it up gently enough to avoid cracking anything takes hours, and a cold start can take most of a day. A gas turbine has no such thermal inertia and can be at full power in minutes. Remember that; it is the whole reason gas has a future on a grid full of solar.

The trick: use the exhaust twice

A gas turbine on its own converts roughly 35–40% of its fuel to electricity — respectable, not remarkable. Its exhaust leaves at around 600 °C.

Look at that number and think like Module 3. Six hundred degrees is hotter than the steam in a coal plant. Throwing it away is throwing away a whole power station.

So don’t. Duct the exhaust into a boiler — called a heat recovery steam generator — raise steam with it, and run a second, entirely conventional steam turbine with its own generator. One lot of fuel, two engines in series, two generators.

That is combined cycle, and it reaches about 60% efficiency, with the best modern machines around 64%. Nothing else that burns fuel comes close: the finest ultra-supercritical coal plant tops out near 45%.

Why does it win so decisively? It is the Carnot limit from Module 3 — the ceiling that said even a flawless coal plant could not get past about 63%. That ceiling depends on just two temperatures, the hot end and the cold end:

ceiling = 1 − T_cold / T_hot (both temperatures in kelvin, where K = °C + 273)

The kelvin part matters far more than it looks. Kelvin counts from absolute zero — the coldest anything can possibly be — so it measures how much heat is genuinely in something. Celsius starts at the freezing point of water, which is an arbitrary place to begin counting. Put Celsius numbers into that formula and the answer is meaningless.

So do it properly. Coal’s steam enters at ~565 °C, which is 838 K, and the condenser sits near 40 °C, or 313 K: the ceiling is 1 − 313/838 ≈ 63%. A gas turbine’s gas enters at ~1,400 °C — 1,673 K — because the hot part is a short blade path rather than kilometres of pressurised tubing. Same cold end, far hotter start: 1 − 313/1673 ≈ 81%.

That is the whole advantage in one line. Starting hotter lifts the ceiling from about 63% to about 81%, and combined cycle is the machine built to actually collect across that taller drop — the gas turbine takes the scorching top of it, and the steam plant bolted to its exhaust harvests the leftovers all the way down. It is a staircase built across a far taller fall.

Plants that exist to wait

Not every gas plant is built for efficiency. Some are built for speed, and run deliberately little.

Demand is not flat. In most of India it climbs through the evening and spikes hard around 7 pm, when workplaces are still drawing power, homes switch on lights, fans, TVs and cooking, and — crucially — the sun has just set on all that solar capacity. Something must cover that spike, as it happens.

A peaker plant is that something: usually an open-cycle gas turbine, with no steam bottom half, chosen because it can go from standstill to full output faster than almost anything else. It may run two hours a day, or a few weeks a year.

That makes its electricity expensive — the whole capital cost is spread over very few units — but the grid pays for a capability, not a commodity. The measure that matters is its ramp rate: megawatts per minute. Hold onto this idea. As solar grows, the evening ramp gets steeper (you will watch it happen with a slider in Module 7), and fast plants become more valuable, not less.

India’s idle fleet

Now the hook question. India had roughly 24.5 GW of gas capacity on the books in FY 2024-25 — and gas produced just 31 TWh, about 2% of the country’s electricity. In the year to March 2025, the fleet’s plant load factor averaged about 14%. Some southern plants ran below 4%. Around 5 GW has been retired outright, having sat unused so long that machinery was sold off or simply rusted beyond use.

This is not a technical failure. It is arithmetic:

• India’s own gas fields do not produce nearly enough for its power sector, and the cheap domestic gas that exists is allocated to higher-priority uses like fertiliser and city gas.
• The shortfall must be imported as LNG, at globally set prices that spiked sharply after 2021.
• Electricity from expensive imported gas costs more than distribution companies — themselves financially stressed, as Module 12 explains — are willing to pay when coal and solar are cheaper.

So the plants wait. It is a genuinely painful outcome: gas emits roughly half the CO₂ of coal per unit of electricity and can start fast, making it exactly the partner a solar-heavy grid wants. India owns the machines and cannot afford the fuel.

The generator in the basement

There is one more Indian gas-and-diesel story, and it is the one you have actually heard: the deep growl from a shop’s back room or an apartment block’s basement when the power cuts.

A diesel genset is a diesel engine turning an ordinary generator. Hospitals, data centres, malls, tower blocks, telecom masts, wedding halls — an entire parallel power system, sized in kilowatts, that exists because grid supply has historically not been reliable enough to bet on.

It is a terrible way to make electricity: a small engine converts perhaps a third of its fuel to power, and diesel is far costlier per unit of energy than coal. Gensets are also noisy and locally polluting, which is why several Indian cities restrict them during severe air-quality episodes.

But nobody buys a genset for cheapness. They buy it because an operating theatre cannot pause, and because a shop with no fan on a June afternoon has no customers. The size of India’s backup-power industry is, in effect, a measurement of how much people distrust the grid — which makes it one of the most honest indicators of grid reliability there is. As supply improves, that industry should shrink. Watching whether it does is a real test of the grid story in Module 10.

🔥Why gas is the awkward middle of the climate argumentAdvanced · optional

Gas occupies genuinely contested ground, and both sides have a point.

The case for. Burning methane produces roughly half the CO₂ per unit of electricity that coal does — partly chemistry (more of the energy comes from hydrogen atoms becoming water rather than carbon becoming CO₂) and partly that superior ~60% efficiency. It emits almost no sulphur dioxide or ash. And its speed complements solar: when a cloud bank crosses a solar park or the sun sets on the evening peak, gas can cover the gap in minutes, where coal cannot.

The case against. Two problems.

Methane leaks. Natural gas is mostly methane, and methane leaking un-burnt into the atmosphere is a far more potent greenhouse gas than CO₂ over the short term. Leak rates across the supply chain — wells, pipelines, LNG terminals — are hard to measure and have often been underestimated. Leak enough and the advantage over coal narrows sharply.

Lock-in. A gas plant built today may run for 30 years. Building the “bridge” can extend the journey, if it delays cheaper clean alternatives and creates an industry with an interest in continued use.

Where India actually sits is a third position that neither slogan captures: it has already built the bridge and cannot afford to walk on it. The debate over whether gas is a transition fuel is, here, partly academic — the binding constraint is the price of imported LNG, not climate philosophy.

🔥🔥The Brayton cycle, and where a gas turbine's work actually goesDeep · optional

The ideal Brayton cycle has four steps: isentropic compression, constant-pressure heat addition (combustion), isentropic expansion, constant-pressure heat rejection. For an ideal gas, its efficiency depends only on the pressure ratio r_p = p₂/p₁:

η = 1 − r_p^(−(γ−1)/γ)

with γ ≈ 1.4 for air. At r_p = 20: η = 1 − 20^(−0.2857) ≈ 1 − 0.425 ≈ 0.575.

Two things fall out of that formula. First, efficiency rises with pressure ratio — hence the relentless engineering push to compress harder. Second, unlike Carnot, it does not mention the combustion temperature at all. Firing temperature matters for a different reason: specific work, the useful work per kilogram of air.

That distinction explains a fact that surprises people. In a gas turbine, the compressor consumes a large fraction — often half or more — of the gross work the turbine produces. Only the remainder reaches the generator. This is why a gas turbine’s output falls on a hot day: intake air is less dense, so the same compressor moves less mass, and the compressor’s parasitic share grows. It is also why gas turbines are so much more compact than steam plants — the working fluid does its job in metres, not kilometres of tubing.

Stacking the cycles. Combined-cycle efficiency is not the sum of the two but a cascade:

η_CC = η_gas + η_steam × (1 − η_gas)

The steam cycle only ever works on what the gas turbine rejected. With η_gas = 0.40 and η_steam = 0.33: η_CC = 0.40 + 0.33 × 0.60 = 0.60. Notice that improving the gas turbine has diminishing returns, since it leaves less heat for the bottoming cycle — which is exactly why designers optimise the pair together rather than each alone.

What comes next

Two ways to make heat, two engines, one shared ceiling — and gas’s speed quietly waiting for a job the grid is about to invent. Module 5 abandons fire entirely. No fuel, no combustion, no chimney: just water that has been lifted uphill by the sun and is trying to come back down.