Rankine Cycle (Steam Power)

How the Rankine cycle works

Four components in a closed loop — boiler, turbine, condenser, pump — repeat one process per leg.

1. Pump (1 → 2). A feed pump compresses liquid water from condenser pressure (≈ 5 kPa) to boiler pressure (often 16–24 MPa). Because liquid water is nearly incompressible, the pump consumes a tiny fraction of cycle work — about 1–2%.
2. Boiler (2 → 3). High-pressure water enters the boiler, absorbs heat at constant pressure, and exits as superheated steam at 540–600°C. This is where fuel — coal, gas, uranium, sun, or geothermal heat — does its work.
3. Turbine (3 → 4). Steam expands across multiple turbine stages, dropping in pressure and temperature, spinning a shaft that drives the generator. Expansion is approximately adiabatic. About 70% of plant gross power leaves through this shaft; the other 30% is rejected as waste heat.
4. Condenser (4 → 1). Low-pressure exhaust steam contacts cooled tubes carrying river, sea, or cooling-tower water. Heat exits to the cold reservoir and steam condenses to liquid at about 30°C. The cycle closes.

     ┌──────────┐  high-P, high-T steam
     │  BOILER  │═══════════════╗
     │   (3)    │               ▼
     └────▲─────┘          ┌────────┐
          ║  high-P liquid │TURBINE │ ── shaft work →
          ║       (2)      │  (3→4) │
       ┌──╨──┐             └────╥───┘
       │PUMP │                  ║  low-P wet steam
       │(1→2)│                  ▼
       └──▲──┘            ┌──────────┐
          ║   low-P liquid│CONDENSER │ ── reject Q_c →
          ╚════════(1)════│   (4→1)  │
                          └──────────┘

Worked example: a 540°C / 240 bar plant

Modern subcritical coal plants commonly run at 16–24 MPa and 540–565°C live-steam conditions. Take the round number: 240 bar (24 MPa) and 540°C entering the high-pressure turbine, condensing at 5 kPa (about 33°C).

From steam tables, enthalpy at boiler exit h₃ ≈ 3325 kJ/kg. After ideal expansion to 5 kPa, h₄s ≈ 1985 kJ/kg, but real turbines have isentropic efficiency ≈ 88%, giving h₄ ≈ 2150 kJ/kg. Pump work is negligible: h₁ ≈ 138 kJ/kg, h₂ ≈ 162 kJ/kg.

   q_in   = h₃ − h₂ = 3325 − 162 = 3163 kJ/kg
   w_turb = h₃ − h₄ = 3325 − 2150 = 1175 kJ/kg
   w_pump = h₂ − h₁ = 162  − 138  =   24 kJ/kg
   w_net  = w_turb − w_pump        = 1151 kJ/kg

   η_thermal = w_net / q_in = 1151 / 3163 ≈ 0.36  →  36%

The simple cycle gets to about 36%. Real plants stack improvements — reheating the steam between turbine stages, regenerative feedwater heating with extracted steam, supercritical operation — to climb past 42%. Carnot for the same 540°C top and 33°C bottom would be 1 − 306/813 ≈ 62%, so a well-designed plant captures roughly 60–70% of its Carnot ceiling.

Variants that boost efficiency

• Reheat Rankine. Steam exits the high-pressure turbine partway through the cycle, returns to the boiler for reheating to 540°C, and re-enters the intermediate-pressure turbine. Adds 4–5 percentage points of efficiency and keeps low-pressure stages dry.
• Regenerative Rankine. Extract small amounts of steam from intermediate turbine stages and feed it into open or closed feedwater heaters that warm the pump-return water. A typical large plant has 6–8 feedwater heaters and gains 5–8 percentage points overall.
• Supercritical Rankine. Boiler pressure pushed above water's critical point (22.1 MPa). The phase change disappears — water transitions smoothly to a supercritical fluid. Efficiency 42–45%, common in modern coal and increasingly in nuclear concept designs.
• Ultra-supercritical (USC). 28–32 MPa, 600–620°C steam, requiring nickel-based superalloy boiler tubing. Efficiency 46–48%. Demonstration plants in Japan and Germany approach 50%.
• Combined cycle. A gas-turbine (Brayton) topping cycle exhausts into a heat-recovery steam generator that runs a Rankine bottoming cycle. Combined-cycle gas plants reach 60–64% — the highest thermal efficiency of any heat engine in commercial service.
• Organic Rankine cycle (ORC). Replaces water with low-boiling-point fluids (pentane, toluene, refrigerants) for low-temperature heat sources — geothermal at 100–200°C, industrial waste heat, biomass. Efficiencies 8–20%, but recovers energy that water can't.

Rankine variants compared

Variant	Steam conditions	Working fluid	Typical efficiency	Where used	Capital cost
Subcritical Rankine	540°C / 16–18 MPa	Water/steam	33–38%	Older coal, biomass	Low
Supercritical Rankine	565°C / 24–25 MPa	Water (supercritical)	42–45%	Modern coal, lignite	Medium
Ultra-supercritical	600–620°C / 28–32 MPa	Water (supercritical)	46–48%	New-build coal in Japan, Germany	High (Ni-superalloys)
Combined cycle	Bottoming cycle, varies	Steam after gas-turbine exhaust	60–64% (gas + steam)	Modern gas-fired plants	Medium-high
Nuclear (PWR secondary)	285°C / 7 MPa saturated	Water/steam	33–35%	Most large reactors worldwide	Very high
Geothermal binary (ORC)	120–180°C source	Pentane, isobutane, R-245fa	8–15%	Low-T geothermal fields	Medium
Concentrated solar	390°C molten salt or oil	Steam	30–37%	CSP towers and trough plants	High

Real-world specs

• Drax Unit 1, UK. 660 MW unit, 568°C / 16.5 MPa, biomass-fired, 38% net thermal efficiency.
• J-Power Isogo Unit 2, Japan. 600 MW USC plant, 600°C / 25 MPa main steam, 620°C reheat — net efficiency 45.5%, one of the highest coal-fired figures recorded.
• Westinghouse AP1000 PWR. Steam-generator outlet 273°C, 5.7 MPa saturated. Net plant efficiency about 32%. Limited by zircaloy fuel cladding, not by Rankine theory.
• The Geysers, California. World's largest geothermal field, 1.5 GW capacity. Direct steam at 175–200°C drives turbines at about 17% net efficiency — modest, but the steam is free.
• GE 9HA combined cycle. Gas-turbine exhaust at 640°C feeds a triple-pressure reheat Rankine bottoming cycle. Block efficiency: 64.0% LHV, set in 2019 at Bouchain, France.

Common failure modes

• Feed pump cavitation. If condensate temperature climbs too close to the boiling point at suction pressure, vapour bubbles form and collapse on impeller blades. Cavitation pits the metal, reduces flow, and can destroy a multi-stage feed pump in days. Cure: maintain net positive suction head (NPSH) margin and pre-heat condensate adequately.
• Turbine blade erosion. Wet steam in the last LP stages carries droplets that hit blade leading edges at hundreds of m/s. Long-term erosion thins blades and shifts vibration modes. Mitigated by superheat, reheat, water-trap rings, and stellite-clad leading edges.
• Boiler tube failure from overheating. Steam-side fouling or under-feedwater conditions starve a tube of cooling. Wall temperature creeps up; over months, creep deformation thins the tube until it ruptures. Most utility outages trace to boiler-tube failures.
• Condenser tube fouling. Biofilms and scale on cooling-water-side tubes raise the temperature head needed for the same heat rejection, raising condenser pressure and cutting cycle efficiency. Ball-cleaning systems run continuously to scrub tubes.
• Steam-water cycle chemistry upsets. Even ppb-level chloride or oxygen ingress causes stress-corrosion cracking in superheater tubes and turbine rotors. Plants run continuous chemistry monitoring and cation-conductivity alarms.

Historical context

William John Macquorn Rankine, a Scottish engineer and physicist, formalised the cycle in 1859 in his Manual of the Steam Engine. By then steam had powered the Industrial Revolution for a century — Watt's separate-condenser engines (1769) had made steam economically viable, and Trevithick's high-pressure engines (1800s) had multiplied power density. Rankine's contribution was the analytical bookkeeping: showing the cycle on a thermodynamic diagram, comparing it to the Carnot ideal, and giving engineers a way to score new designs.

Steam pressures climbed steadily — 1 MPa in 1900, 10 MPa by the 1930s, 16 MPa standard by the 1960s, supercritical 25 MPa by the 1990s, ultra-supercritical 30 MPa today. Each jump required new alloys: carbon steel gave way to chrome-moly, then to advanced 9–12% Cr martensitic steels, and now to nickel-based superalloys for USC tubing. A 0.5% efficiency gain on a 1 GW plant saves about 100,000 tonnes of coal per year, which is why the engineering community keeps pushing.