Heat and thermodynamics: engines, heat pumps and their limits: The first and second laws: the audit and the limit
All courses / Understanding Engineering / Module 6: Heat and thermodynamics: engines, heat pumps and their limits Part 2 of 5The first and second laws: the audit and the limit
Every heat machine is judged by two rules. The first law is a strict book-keeping rule: energy in must equal energy out plus change in store. The second law is a rule about quality, and it says that heat cannot be turned wholly into work. Together they let an engineer audit a machine on paper and know in advance the best it could ever do.
2.1 The first law is an energy audit: whatever goes in must come out or be stored
The first law of thermodynamics is a ledger. Any rise in the energy stored in a body is paid for by heat flowing in or by work done on it, and any work the body does on the outside is paid for from heat it receives or from its store. Energy changes form and place but is never made or lost. For an engineer this becomes a habit: draw a box round the machine, as the control volume of module 5 (Fluids) did for flow, and count every joule that crosses its walls. For a machine that runs in a cycle and returns to where it started, nothing is stored, so heat taken in = work given out + heat rejected. A claimed machine that gives out more than it takes in breaks the first law, and engineers reject it without opening the case.
The audit is most useful on ordinary things. Take a pressure cooker on an LPG stove. Sealed, it lets the pressure build to about one bar above the atmosphere, so the water boils near 120 °C instead of 100 °C. A rough rule says chemical reactions in cooking go about twice as fast for each 10 K rise, so 20 K hotter means roughly four times quicker, which turns a 40 minute pulse into a 10 minute one. The price is a heavier, sealed vessel with a safety valve.
The table audits one batch. It counts the heat that reaches the food and the pot, the heat carried out as vented steam, and then asks how much fuel must burn when only about half the flame's heat gets into the pan. The remainder is not lost to the universe. It goes into the kitchen air, where it does no cooking but is still counted.
gas burns; about 55 per cent of the heat reaches the pot (assumed), the rest warms the kitchen air
2 Cookerheats 2 kg of food and water by 90 K, 1 kg of aluminium pot by 90 K, and boils off about 50 g as steam
Outputs Heat in food and water754 kJ (2 kg x 4.19 x 90) Heat in the aluminium pot81 kJ (1 kg x 0.9 x 90) Heat in vented steamabout 113 kJ (0.05 kg x 2,260) Heat lost round the potabout 780 kJResults calculated. Useful total 0.95 MJ; divided by 0.55 gives 1.73 MJ of fuel. At 46 MJ/kg for LPG that is 38 g, so one 14.2 kg domestic cylinder covers about 380 such batches. If the burner takes in about 2.2 kW (some 170 g of LPG an hour, an assumed rating), the pot receives 1.2 kW and heating takes about 0.95 MJ / 1.2 kW = 790 s, some 13 minutes.
Source: Bloomfield, How Things Work (Wiley, 2015), ch. 7, 'Heat and Phase Transitions', sections 7.1 and 7.2; ch. 8, 'Thermodynamics', section 8.1 (conservation of energy)- Bloomfield, How Things Work (Wiley, 2015) · ch. 8, 'Thermodynamics', section 8.1 ('Air Conditioners': the laws of thermodynamics)
- Kosky et al., Exploring Engineering (Academic Press, 2013) · ch. 4, 'Energy Conversion', sections 4.4 and 4.5 (energy conversion; conservation of energy)
2.2 The second law says heat flows downhill, and entropy is the book-keeping that forbids a perfect engine
If you leave a hot cup of tea in a cool room, it cools. It never heats itself by taking heat from the room, though the first law would allow that. The second law records the one-way habit: heat flows by itself only from hot to cold. A refrigerator moves heat the other way, but it needs work to do so and dumps more heat into the kitchen than it removes from its inside.
Entropy is the quantity that makes this exact. In plain terms, it measures how widely a store of energy is spread among the many ways molecules can hold it. When heat Q is added to a body at absolute temperature T, it carries entropy Q/T. A joule at a low temperature carries more entropy than a joule at a high one. Work, by contrast, carries none. Now move 1,000 J from a hot body at 600 K to a cold one at 300 K. The hot body loses 1,000/600 = 1.67 J/K of entropy and the cold one gains 1,000/300 = 3.33 J/K. Net, 1.67 J/K has been created. Nothing in nature lets the total fall, and every real process, such as friction, mixing and heat flow through a temperature gap, adds some.
This fact is what limits engines. An engine takes in heat from a hot source and with it takes in entropy. Work cannot carry that entropy out. The only way to get rid of it, so that the engine can come back to its start and run again, is to reject some heat to a colder place, where each joule carries more entropy. So an engine must always discard some heat, and the colder the place it can discard to, the less it needs to give up. That reasoning gives the Carnot limit in the next section.
| Event | Working | Entropy change |
|---|---|---|
| Hot body at 600 K gives up 1,000 J | -1,000 / 600 | -1.67 J/K |
| Cold body at 300 K takes in 1,000 J | +1,000 / 300 | +3.33 J/K |
| Total | 3.33 - 1.67 | +1.67 J/K, created by the flow |
| Same 1,000 J across a smaller gap, 600 K to 500 K | -1.67 + 2.00 | +0.33 J/K, much less created |
| Impossible reverse flow, cold to hot with no work | +1.67 - 3.33 | -1.67 J/K, so it never happens by itself |
Entropy is in joules per kelvin. The smaller the temperature gap, the less entropy is created, which is why engineers aim to transfer heat across small gaps.
Source: Bloomfield, How Things Work (Wiley, 2015), ch. 8, 'Thermodynamics', section 8.1 ('Air Conditioners': entropy); table values calculated- Bloomfield, How Things Work (Wiley, 2015) · ch. 8, 'Thermodynamics', section 8.1 ('Air Conditioners': entropy)
- Feynman, Leighton and Sands, The Feynman Lectures on Physics, vol. I (Caltech online edition) · further reading: ch. 44, 'The Laws of Thermodynamics', https://www.feynmanlectures.caltech.edu/I_44.html
- Feynman, Leighton and Sands, The Feynman Lectures on Physics, vol. I (Caltech online edition) · further reading: ch. 46, 'Ratchet and Pawl', https://www.feynmanlectures.caltech.edu/I_46.html
2.3 The Carnot limit: the best efficiency of any heat engine depends only on two temperatures
Carry the entropy argument one step on. An ideal engine takes heat Qh at hot temperature Th and rejects Qc at cold temperature Tc, creating no entropy. Entropy in equals entropy out: Qh/Th = Qc/Tc. So Qc/Qh = Tc/Th, and the work is Qh - Qc. Efficiency, work divided by heat in, is therefore one minus Tc/Th, with both temperatures in kelvin. This is the Carnot limit. It does not depend on the fluid, the layout or the cleverness of the designer, which is why it is a law and not a rule of thumb.
Two lessons follow. To raise efficiency, make the hot end hotter or the cold end colder. And because the cold end is usually the surroundings, at about 300 K, the work comes mostly from the height of Th. The limit is large only when Th is far above ambient: a low-temperature source cannot give a good engine, which is why ocean thermal plants and low-grade waste-heat engines never exceed single-digit or low double-digit efficiency.
Real engines fall below the limit for three reasons. Heat is added over a range of temperatures, not at the top one only. Heat transfer needs a temperature gap, and each gap creates entropy. And there is friction, leakage and stray loss. The table shows the ideal limit beside real machines. A coal plant with steam at 540 °C and a condenser at 40 °C has a limit above 60 per cent but delivers 33 to 40. Look at the Gas turbine and Steam turbine nodes: the first runs hotter, the second colder, and the best plants put them in series.
The trade-off is always material strength and cost against efficiency. Hotter metal creeps and oxidises (module 4, Structures and materials), so each rise of a few tens of kelvin needs a better alloy.
| Machine | Th and Tc | Carnot limit 1 - Tc/Th | Typical real value |
|---|---|---|---|
| Ocean thermal plant | 301 K (28 °C), 278 K (5 °C) | 7.6% | about 3% net, after pumping |
| Waste-heat engine on 150 °C flue gas | 423 K, 303 K (30 °C) | 28% | about 8 to 12% |
| Steam plant, subcritical, 540 °C steam | 813 K, 313 K (40 °C) | 61% | 33 to 40% |
| Simple gas turbine, 1,400 °C inlet | 1,673 K, 303 K (30 °C) | 82% | 35 to 44% |
| Combined gas and steam cycle | same two ends | 82% | about 60 to 63% in service; record 64.18% under ISO test conditions (2024) |
Carnot values calculated from the formula; temperatures rounded. Real values are typical figures for machines of the 2020s from makers' data and plant reports. The combined-cycle record is the corrected gross efficiency verified by Guinness World Records for Keadby 2 (Siemens Energy) in May 2024.
Source: Bloomfield, How Things Work (Wiley, 2015), ch. 8, 'Thermodynamics', section 8.2 ('Automobiles': heat engines and their efficiency)- Bloomfield, How Things Work (Wiley, 2015) · ch. 8, 'Thermodynamics', sections 8.1 and 8.2 ('Air Conditioners'; 'Automobiles')
- Feynman, Leighton and Sands, The Feynman Lectures on Physics, vol. I (Caltech online edition) · further reading: ch. 44, 'The Laws of Thermodynamics', https://www.feynmanlectures.caltech.edu/I_44.html
- Guinness World Records, 'Most efficient combined cycle power plant' (verified 21 May 2024; read November 2026) · 64.18 per cent corrected gross efficiency under ISO conditions, Keadby 2, Siemens Energy; https://www.guinnessworldrecords.com/world-records/431420-most-efficient-combined-cycle-power-plant
See it in the table
H₂O WaterTea TeaEl ElectricityCheck your understanding
A closed room contains only a running ceiling fan of 50 W. What happens to the room's temperature?
The first law: the 50 W of electricity must go somewhere, and in a closed room it ends as heat. The fan feels cool to you only because moving air helps sweat evaporate.In the entropy tally, why can 1,000 J not flow by itself from a 300 K body to a 600 K body?
The cold body would lose 3.33 J/K and the hot body gain only 1.67 J/K, so total entropy would decrease. Energy would be conserved, but the second law forbids it without added work.A heat engine takes in heat at 600 K and rejects it at 300 K. What is the most efficiency it could have?
The Carnot limit is 1 - 300/600 = 0.5. No design can beat this, and real engines fall below it because of finite temperature gaps and friction.