Engineering Thermodynamics

The big picture

Think about what you did this morning. You boiled water for coffee, and a resistor turned electrical energy into heat. You drove or rode to work, and a hot gas pushed pistons or spun a turbine somewhere behind the scenes. Your refrigerator quietly pumped heat out of cold food and into a warm kitchen, which sounds backwards until you learn the trick. Behind every one of those small events stands a single science: thermodynamics, the study of energy, its forms, and the rules that govern its transformations. Engineers who design power plants, jet engines, air conditioners, batteries, and rockets all answer to the same two great laws you will learn in this course.

Those laws have a remarkable pedigree. They were worked out in the nineteenth century, largely by people trying to make steam engines pay, before anyone had proven that atoms exist. Sadi Carnot analyzed engines in 1824, James Joule measured the equivalence of work and heat in the 1840s, and William Thomson (Lord Kelvin) built the absolute temperature scale on their results. Physics has been overturned twice since then, by relativity and by quantum mechanics, and thermodynamics sailed through both revolutions untouched. Einstein himself regarded classical thermodynamics as the physical theory least likely ever to be overthrown. When you learn it, you are learning something built to last.

Here is the plan for today. First we define the science and stake out the single most useful habit in it: drawing a boundary around the thing you are analyzing. Then we sort the quantities that describe a system into intensive and extensive properties, pin down what a state and equilibrium mean, and name the processes and cycles that carry a system from one state to another. We close with the SI units this course will use in every calculation. None of this is glamorous, and all of it is load-bearing. Every energy balance you will ever write starts here.

The science of energy

The name thermodynamics comes from the Greek words therme, heat, and dynamis, power, and the original question was exactly that: how do you get power out of heat? A modern definition is broader. Thermodynamics is the science of energy in all its forms, of the conversion of energy from one form to another, and of the limits nature places on those conversions. The first law says energy is conserved: you can convert it but never create or destroy it. The second law says conversions have a direction and a price: heat flows from hot to cold on its own, never the reverse, and no engine can turn heat entirely into work. The whole course is an unpacking of those two sentences.

Notice what a strange and powerful kind of science this is. Thermodynamics makes almost no assumptions about what matter is made of. It did not need atoms to be discovered, and it applies equally to steam, air, liquid sodium, rubber bands, and black holes. That generality is why a course taught with steam tables and gas turbines will still serve you if you end up designing battery packs or data center cooling. The price of that generality is care with definitions, which is where we begin.

Key idea: Thermodynamics is the science of energy and its transformations, built on two laws: energy is conserved, and energy conversions have a direction and a limit.

Systems, boundaries, and surroundings

Every thermodynamic analysis begins the same way: you choose a region of the universe to study and draw a boundary around it. The region you choose is the system. Everything outside the boundary is the surroundings. The boundary itself can be real, like the steel wall of a tank, or imaginary, like a surface drawn across the inlet and outlet pipes of a pump. It can move, like the face of a piston, or stay put. Choosing the boundary is not bookkeeping; it is the decision that determines what counts as energy entering or leaving, and a poorly chosen boundary can turn an easy problem into a miserable one.

Three types of system cover everything. A closed system, also called a control mass, is a fixed quantity of matter: no mass crosses the boundary, though energy may. The air sealed inside a piston-cylinder is the classic example. An open system, also called a control volume, is a region of space through which mass flows: a water heater, a turbine, a nozzle, your own lungs. Most working machinery is analyzed as an open system. An isolated system exchanges neither mass nor energy with its surroundings. A sealed, perfectly insulated thermos would be one; strictly, the only perfect isolated system is the universe itself, but the idealization is useful whenever leakage is small over the time you care about.

Try classifying as you look around. A sealed bottle of water warming on your desk is a closed system: energy in, no mass in or out. The same bottle with the cap off becomes an open system the moment vapor escapes. A pot of soup boiling on the stove is open (steam leaves). You yourself are an open system with roughly a 100 watt energy throughput, taking in chemical energy as food and rejecting heat to the room.

Key idea: A system is whatever you draw the boundary around: closed if mass cannot cross, open if it can, isolated if neither mass nor energy crosses.

Properties, intensive and extensive

A property is any measurable characteristic of a system: pressure, temperature, volume, mass, density, energy content. Properties split into two families, and the split matters more than it first appears. Imagine an equilibrium system, say a tank of nitrogen, and mentally slice it in half. Some quantities are unchanged in each half: temperature, pressure, density. These are intensive properties; they do not depend on how much matter you have. Other quantities are halved: mass, volume, total energy. These are extensive properties; they scale with the amount of substance.

Engineers constantly convert extensive properties into intensive ones by dividing by mass, because a per-kilogram number can be tabulated once and used for any amount. Volume V divided by mass m gives the specific volume v, measured in cubic meters per kilogram, written m^3/kg (this course uses the caret to mark exponents, so m^3 means cubic meters). Specific volume is the reciprocal of density. A quick example with real numbers: suppose 2 kg of gas fills a rigid 0.5 m^3 tank. Then v = 0.5 / 2 = 0.25 m^3/kg, and the density is rho = 1 / 0.25 = 4.0 kg/m^3. Later you will meet specific internal energy, specific enthalpy, and specific entropy, all built the same way, and all listed per kilogram in property tables.

Key idea: Intensive properties like temperature and pressure are independent of system size; extensive properties like mass and volume scale with it, and dividing by mass turns extensive into intensive.

State, equilibrium, and the state postulate

The state of a system is its complete condition, the full set of its property values at one instant. But a system only has well-defined, uniform properties when nothing inside it is still trying to change, and that condition has a name: equilibrium. Thermodynamic equilibrium actually bundles several requirements. Thermal equilibrium means the temperature is the same everywhere in the system, with no internal heat flows. Mechanical equilibrium means no unbalanced pressure differences are pushing anything around. Phase equilibrium means liquid and vapor, if both are present, have stopped converting into each other on net. Chemical equilibrium means no net reactions are running. When all of these hold, the system sits still in every sense that matters, and we can speak of the state.

How many properties does it take to pin a state down? Remarkably few. For a simple compressible system, meaning one substance and no electrical, magnetic, or surface effects worth counting, the state postulate says two independent intensive properties fix the state completely. Give me the temperature and specific volume of steam, and every other property, its pressure, its energy, its entropy, is already determined; I can look them all up. This is why property tables work, and why so much of this course involves finding two properties and reading off the rest. One caution hides in the word independent: during boiling, temperature and pressure are locked together, so that pair no longer counts as two independent properties. We will handle that carefully in the lesson on pure substances.

Key idea: Equilibrium means no internal driving forces remain, and for a simple compressible substance any two independent intensive properties fix the entire state.

Processes, paths, and cycles

Change the state and you have performed a process: compression, heating, expansion, cooling. The sequence of states passed through along the way is the path. If a process is carried out slowly enough that the system stays essentially in equilibrium at every instant, we call it a quasi-equilibrium process. Real processes are never perfectly slow, but many are close, and the idealization lets us draw the path as a curve on a property diagram and compute with it. A gas compressed by a slowly advancing piston tracks quasi-equilibrium well; a gas ripped open to vacuum does not.

Certain processes are so common they carry names you should learn now. An isothermal process holds temperature constant. An isobaric process holds pressure constant, like boiling water in an open pot at one atmosphere. An isochoric (or isometric) process holds volume constant, like heating gas in a rigid sealed tank. An adiabatic process transfers no heat, like the rapid compression in a bicycle pump too quick for heat to leak out. Finally, a cycle is a sequence of processes that returns the system to its initial state. Cycles are the beating heart of engineering thermodynamics: an engine is useful precisely because its working fluid comes back to where it started, over and over, delivering net work every lap.

One more distinction will save you grief later. Properties depend only on the state, not on how the system got there; we say they are point functions. Quantities like work and heat depend on the path taken between states; they are path functions. Your bank balance is a point function; your total deposits are a path function. Keep that picture, because next lesson heat and work stop being casual words and become precise, path-dependent transfers.

Key idea: Processes carry a system between states along a path, cycles return it to the start, and properties depend only on the state while work and heat depend on the path.

Units that will not betray you

Thermodynamics punishes sloppy units more reliably than any other subject you will study, so let us fix the toolkit now. This course uses SI. The base units are the kilogram (kg) for mass, the meter (m) for length, the second (s) for time, and the kelvin (K) for temperature. Force comes in newtons: 1 N accelerates 1 kg at 1 m/s^2. Energy comes in joules: 1 J is 1 newton meter, and because a joule is small we will usually work in kilojoules, kJ, thousands of joules. Pressure comes in pascals: 1 Pa is 1 N/m^2, and we will usually use kilopascals (kPa) or megapascals (MPa). Power, the rate of energy flow, comes in watts: 1 W is 1 J/s, and 1 kW is 1 kJ/s.

Here is the discipline in action. A 1500 kg car moving at 30 m/s (about 108 km/h) carries kinetic energy KE = (1/2)(1500 kg)(30 m/s)^2 = 675,000 J = 675 kJ. Compare that with a homely thermal number: warming 2 kg of water from 20 C to 100 C takes about (2 kg)(4.18 kJ/kg K)(80 K) = 669 kJ. The car barreling down a highway and two liters of water coming to a boil hold almost the same energy. Thermal energies are enormous compared with everyday mechanical ones, which is exactly why engines that tap heat changed the world. Write units into every line of every calculation and let them multiply and cancel like algebra; when the units of your answer come out wrong, the physics is wrong, and the units will tell you first.

Key idea: Work in SI (kg, m, s, K, with J, kPa, and kW built from them), carry units through every step, and let unit cancellation audit your work.

Common misconceptions

• Thermodynamics is only about heat. It is about energy in every form: heat, work, kinetic, potential, chemical, electrical, and the rules for converting among them.
• A closed system means nothing gets in or out. Closed refers to mass only; energy crosses the boundary of a closed system freely, as heat or work. The system that blocks both is called isolated.
• Temperature and pressure always count as two independent properties. During a phase change they are locked together, so they cannot by themselves fix the state of a boiling mixture.
• Properties describe processes. Properties describe states; a process is described by the path between states and by the work and heat exchanged along it.

Recap

• Thermodynamics is the science of energy and its transformations, governed by conservation (first law) and directionality (second law).
• Every analysis starts by drawing a boundary: closed systems fix the mass, open systems let mass flow through, isolated systems exchange nothing.
• Intensive properties (T, P, density) ignore system size; extensive properties (m, V, energy) scale with it; dividing by mass gives specific properties like v = V/m.
• Equilibrium means no unbalanced thermal, mechanical, phase, or chemical driving forces, and two independent intensive properties fix the state of a simple compressible substance.
• Processes connect states, cycles return to the start, and properties are point functions while heat and work are path functions.
• SI units (kg, m, s, K, J, kPa, kW) carried through every line are your first and best error check.

Sources

1. OpenStax. (2016). Thermodynamic systems. In University physics volume 2. Rice University. openstax.org
2. OpenStax. (2016). Temperature and thermal equilibrium. In University physics volume 2. Rice University. openstax.org
3. Engineering LibreTexts. (n.d.). Thermodynamics bookshelf. LibreTexts. eng.libretexts.org
4. Khan Academy. (n.d.). Thermodynamics. khanacademy.org

The big picture

Suppose I hand you a thermometer and ask a question that sounds insultingly simple: what, exactly, does this thing measure? Not "temperature" as a word, but temperature as a physical fact about matter. And a second question like it: when a tire gauge reads 220, what is the 220 counting, and counting from where? These two everyday instruments, the thermometer and the pressure gauge, report the two properties you will use more than any others in this course. Today we make both of them precise, because every table you will read and every efficiency you will compute stands on them.

There is also a law hiding in your kitchen. When you leave a cold spoon in hot soup, the spoon warms, the soup cools imperceptibly, and eventually nothing more happens. That final condition, where objects in contact stop trading energy, is thermal equilibrium, and a deceptively modest statement about it turns out to be the logical foundation of all temperature measurement. It was recognized so late, after the first and second laws already had their names, that physicists had to number it backwards: the zeroth law.

Here is the plan for today. First the zeroth law and why it licenses the thermometer. Then the temperature scales, and the reason engineers must work in kelvins. Then pressure: its definition, its units, the difference between gauge and absolute readings, and the pressure of fluid columns, which lets you understand barometers, manometers, and blood pressure cuffs with one formula. We close with the rule this course will enforce relentlessly: absolute scales or nothing.

The zeroth law and thermal equilibrium

Put two objects in contact and, if they are free to exchange energy, their temperatures drift toward each other until the drift stops. At that point they are in thermal equilibrium: same temperature, no net energy flow. The zeroth law of thermodynamics says: if body A is in thermal equilibrium with body C, and body B is separately in thermal equilibrium with body C, then A and B are in thermal equilibrium with each other. Two things equal to a third are equal to each other, applied to hotness.

That sounds like a triviality until you notice what it buys you. Let body C be a thermometer. Touch it to a pot of water, let it equilibrate, and read 40. Touch it to a bathtub, let it equilibrate, and read 40. The zeroth law is the guarantee that the pot and the bathtub are at the same temperature as each other, even though they never touched. Without the zeroth law, a thermometer reading would be a private fact about the thermometer and one object, not a universal label you can compare across objects. The law converts temperature from a sensation into a measurable, transferable property. Ralph Fowler gave the principle its odd name in the 1930s, long after the first and second laws were established, precisely because it deserved to come before them.

Key idea: The zeroth law says thermal equilibrium is transitive, and that guarantee is what makes a thermometer a meaningful instrument.

Temperature scales, and why engineers use kelvins

A thermometer needs numbers on it, and the numbers are a human choice. The Celsius scale pins 0 near the freezing point of water and 100 near its boiling point at one standard atmosphere. The Fahrenheit scale, older and still used in the United States, puts those same two events at 32 and 212, with T(F) = 1.8 T(C) + 32. Both scales share a defect for our purposes: their zeros are arbitrary. Nothing physically special happens to matter at 0 C except to water.

Nature, it turns out, supplies a true zero. Cool a dilute gas at constant volume and its pressure falls steadily; extrapolate the line and the pressure would reach zero at -273.15 C, the same temperature for every gas. That universal floor is absolute zero, the state of minimum molecular energy, and the Kelvin scale starts there: T(K) = T(C) + 273.15. A kelvin step is the same size as a Celsius degree, and by convention we write 298 K, not 298 degrees K. Since 2019 the kelvin has been defined through the Boltzmann constant, k = 1.380649 x 10^-23 J/K, tying temperature directly to molecular energy. American practice also has an absolute scale in Fahrenheit-sized steps, the Rankine scale, T(R) = T(F) + 459.67, which you may meet in older texts.

Benchmark	Celsius	Kelvin
Absolute zero	-273.15 C	0 K
Water freezes (1 atm)	0 C	273.15 K
Room temperature	25 C	298.15 K
Human body	37 C	310.15 K
Water boils (1 atm)	100 C	373.15 K

Why does this course insist on kelvins? Because the formulas of thermodynamics use temperature ratios and products, and ratios only mean something on a scale that starts at the true zero. Is 40 C twice as hot as 20 C? In Celsius arithmetic it looks that way, but convert: 313 K versus 293 K, a ratio of only 1.07. The Celsius version lies because its zero is arbitrary. Later, when you compute a Carnot efficiency as 1 minus a temperature ratio, using Celsius will not just be imprecise, it will be nonsense. The safe habit: convert to kelvins the moment a temperature enters a calculation, and only convert back at the end if a reader needs it.

Key idea: The Kelvin scale starts at nature's true zero, and every thermodynamic formula in this course requires absolute temperature in kelvins.

Pressure: the honest definition

Gases and liquids push on every surface that contains them, and pressure is the measure of that push: normal force per unit area, P = F/A. The SI unit is the pascal, 1 Pa = 1 N/m^2, and it is tiny; the atmosphere presses on you at about 101,325 Pa. So engineers speak in kilopascals (kPa) and megapascals (MPa). You will also meet the bar, exactly 100 kPa; the standard atmosphere, 101.325 kPa; and the pound per square inch (psi), about 6.895 kPa, still standard on American tire gauges. A typical car tire runs near 32 psi, a bicycle tire near 80 psi, and the steam in a modern power plant boiler above 15 MPa, roughly 150 times atmospheric pressure.

Molecularly, gas pressure is a drumbeat of impacts: enormous numbers of molecules colliding with the walls, each transferring a little momentum. That picture explains at once why heating a rigid tank of gas raises its pressure (faster molecules hit harder and more often) and why pressure acts equally in all directions at a point in a fluid at rest.

Key idea: Pressure is normal force per unit area, produced in a gas by molecular impacts, and measured in pascals, kilopascals, and megapascals.

Absolute, gauge, and vacuum pressure

Now the tire question. Most gauges do not report pressure counted from zero; they report how much the pressure exceeds the local atmosphere, because the atmosphere pushes on the outside of everything. That reading is gauge pressure. The pressure counted from a perfect vacuum is absolute pressure, and the two are related by P(abs) = P(gauge) + P(atm). A tire gauge reading 220 kPa on a day when the barometer stands at 101 kPa means the air inside the tire is really at 321 kPa absolute. When a pressure sits below atmospheric, we often quote the shortfall as a vacuum pressure: a condenser operating at 20 kPa absolute under a 101 kPa atmosphere holds a vacuum pressure of 81 kPa.

The distinction is not pedantry. The ideal gas law, property tables, and every formula in this course require absolute pressure, for the same reason they require absolute temperature: the equations describe the state of the matter, and the matter does not know or care what the atmosphere outside the gauge happens to be. Treat gauge readings as raw data that must be converted before use.

Key idea: Gauges read pressure relative to the atmosphere, but thermodynamic calculations require absolute pressure, P(abs) = P(gauge) + P(atm).

Pressure from a column of fluid

Dive to the bottom of a pool and your ears report a fact worth a formula: pressure in a static fluid increases with depth, because each layer must carry the weight of all the fluid above it. The rule is P = rho g h, where rho is the fluid density, g = 9.81 m/s^2, and h is the depth. This one equation explains the classic instruments. A barometer is a tube of mercury inverted over a dish; the atmosphere pushes the column up until the column's weight balances it. Standard atmospheric pressure supports h = 0.760 m of mercury: check it, P = (13,590 kg/m^3)(9.81 m/s^2)(0.760 m) = 101,300 Pa, which is 101.3 kPa. Doctors still quote blood pressure in millimeters of mercury for exactly this historical reason; 120 mm Hg is about 16 kPa of gauge pressure.

Why mercury and not water? Solve for the water column that balances the atmosphere: h = P / (rho g) = 101,325 / (1000 x 9.81) = 10.3 m. A water barometer would need to be three stories tall, and this number is also why no suction pump on Earth can lift water more than about 10 meters. A manometer, the U-shaped tube of liquid used to measure tank pressures in laboratories, is the same physics read in reverse: the height difference between the two legs times rho g gives the pressure difference between the tank and the atmosphere.

Key idea: Static fluid pressure grows with depth as P = rho g h, which is the whole working principle of barometers and manometers.

Common misconceptions

• Temperature and heat are the same thing. Temperature is a property of a system; heat is energy in transit between systems at different temperatures. A bathtub at 40 C and a teacup at 40 C have equal temperatures but store vastly different thermal energy.
• Doubling the Celsius reading doubles the hotness. Ratios are meaningless on Celsius; 40 C versus 20 C is 313 K versus 293 K, a 7 percent increase in absolute temperature, not 100 percent.
• A gauge reading of zero means there is no pressure. It means the pressure equals the local atmosphere, about 101 kPa absolute at sea level; a flat tire is still full of air at atmospheric pressure.
• Absolute zero is just very cold and reachable with a good enough freezer. It is the floor of molecular energy, and the third law of thermodynamics implies it can be approached (laboratories have reached below one billionth of a kelvin) but never exactly attained.

Recap

• The zeroth law makes thermal equilibrium transitive, which is what allows a thermometer reading to be compared across objects.
• The Kelvin scale starts at absolute zero, -273.15 C, and T(K) = T(C) + 273.15; kelvins are mandatory inside thermodynamic formulas.
• Pressure is force per area; useful units are kPa, MPa, bar (100 kPa), atm (101.325 kPa), and psi (6.895 kPa).
• Absolute pressure equals gauge pressure plus atmospheric pressure, and thermodynamics runs on absolute values of both P and T.
• Fluid columns exert P = rho g h, the operating principle of the barometer (760 mm of mercury balances one atmosphere) and the manometer.

Sources

1. OpenStax. (2016). Temperature and thermal equilibrium. In University physics volume 2. Rice University. openstax.org
2. OpenStax. (2016). Thermometers and temperature scales. In University physics volume 2. Rice University. openstax.org
3. OpenStax. (2022). College physics 2e. Rice University. openstax.org
4. Khan Academy. (n.d.). Thermodynamics. khanacademy.org

The big picture

Thermodynamics is, at bottom, accounting. Not the metaphorical kind: literal accounting, with a balance sheet and transactions, except the currency is energy instead of money. Every analysis you will do in this course amounts to auditing one system's energy account. And like money, energy comes in exactly two kinds of entries. There is the balance, the energy a system stores inside itself, and there are the transactions, the energy that crosses the boundary during a process. Thermodynamics recognizes several stored forms but only two transaction types: work and heat. Everything that ever crosses the boundary of a closed system is one or the other.

This sounds tidy, and it is, but the tidiness was bought with a century of confusion. Until the 1840s, most scientists believed heat was an invisible fluid called caloric that flowed from hot bodies to cold ones and could be neither created nor destroyed. It took Benjamin Thompson watching cannon boring produce apparently limitless heat from friction, and then James Joule's meticulous paddle-wheel experiments, to establish the modern view: heat and work are not substances at all. They are both energy in transit, interchangeable at a fixed exchange rate, and once the energy arrives inside a system you cannot tell which way it came in. Today's lesson builds that view carefully, because the first law, which is next module's centerpiece, is nothing more than this accounting made exact.

Here is the plan. First we inventory the stored forms of energy and put real numbers on them. Then we make work precise, including the version that matters most in this course, the work of moving a boundary against pressure. Then heat, its definition, its mechanisms, and the specific heat that connects it to temperature change. We end with the sign convention this course will use in every equation and the difference between energy and power.

The stored forms: kinetic, potential, internal

A system can carry energy in three ways we will track. It can move as a whole: kinetic energy, KE = (1/2) m V^2, where V is velocity. It can sit high in a gravity field: potential energy, PE = m g z, where z is elevation and g = 9.81 m/s^2. And it can store energy invisibly, in the microscopic world: the random translation, rotation, and vibration of its molecules, the stretching of bonds, the forces between neighboring molecules. That microscopic hoard is the internal energy, symbol U, specific form u = U/m. The total energy is E = U + KE + PE, and for most stationary engineering equipment the kinetic and potential pieces are negligible, which leaves internal energy carrying the story.

Let us calibrate with numbers. A 1200 kg car at 25 m/s (90 km/h) carries KE = (1/2)(1200)(25)^2 = 375,000 J = 375 kJ. A kilogram of water at the top of a 50 m dam holds PE = (1)(9.81)(50) = 490 J, about 0.49 kJ. Now the internal-energy side: merely boiling away 1 kg of water absorbs 2257 kJ. That single kilogram of steam took in six times the kinetic energy of the speeding car, and as much energy as lifting the same kilogram 230 kilometers straight up. Molecular energies dwarf everyday mechanical ones. This is why heat engines changed civilization and why a course about them is worth your time.

Key idea: Systems store energy as kinetic, potential, and above all internal energy, and the microscopic internal hoard is enormous compared with everyday mechanical energies.

Work: energy crossing by force

In mechanics, work is force acting through a distance, W = F d, measured in joules. Thermodynamics keeps that definition and adds a broader test: an energy transfer is work if its sole effect on the surroundings could have been the raising or lowering of a weight. Electricity flowing into a motor is work by this test; so is a rotating shaft, where W = 2 pi n T for n revolutions against torque T; so is a piston pushing a crankshaft. The test sounds fussy now, but it is what will let us keep work and heat cleanly separate when the second law starts charging them different prices.

The star of this course is boundary work, the work done when a system's boundary moves against pressure, as in every engine cylinder ever built. Picture gas in a piston-cylinder at pressure P pushing the piston face of area A outward a small distance ds. The force is P A, the work is P A ds, and A ds is the small volume change dV. So each sliver of work is P dV, and for a quasi-equilibrium process the total is the integral of P dV from the initial to the final volume: geometrically, the area under the process curve on a P-V diagram. That picture explains immediately why work is a path function: different curves between the same two endpoints enclose different areas.

Two special cases cover most homework and much of industry. For a constant-pressure process, the integral collapses to W = P (V2 - V1). Example: gas at 200 kPa expands from 0.1 m^3 to 0.3 m^3; the work is W = 200 x 0.2 = 40 kJ. Watch the units earn their keep: kPa times m^3 is kN/m^2 times m^3, which is kN m, which is kJ. For an isothermal ideal gas process, pressure falls as volume grows with P V constant, and the integral gives W = P1 V1 ln(V2/V1). Example: air at 100 kPa filling 0.5 m^3 doubles its volume at constant temperature: W = 100 x 0.5 x ln(2) = 50 x 0.693 = 34.7 kJ, noticeably less than the 50 kJ a constant-pressure expansion of the same volume change would deliver, because the pressure sagged along the way.

Key idea: Boundary work is the integral of P dV, the area under the path on a P-V diagram: W = P (V2 - V1) at constant pressure, W = P1 V1 ln(V2/V1) for an isothermal ideal gas.

Heat: energy crossing by temperature difference

Now the other transaction. Heat is energy transferred between a system and its surroundings solely because of a temperature difference. No temperature difference, no heat: that is the entire definition, and it is worth memorizing verbatim, because everyday English uses the word heat for at least five different physical ideas and the exam will not. A process during which no heat crosses the boundary is adiabatic, from the Greek for not passing through, achieved in practice with insulation or with speed, since a fast process leaves no time for thermal leakage. Adiabatic does not mean the temperature is constant: compress a gas adiabatically and its temperature rises, from the work input alone.

Heat finds its way across boundaries by three mechanisms you should be able to name. Conduction passes molecular kinetic energy from neighbor to neighbor through a stationary material, fast in metals, slow in air, which is why cookware is steel and oven mitts are quilted. Convection carries energy with a moving fluid, as when wind strips heat from your skin. Radiation travels as electromagnetic waves and needs no medium at all: it is how the Sun delivers about 1360 W to each square meter facing it above our atmosphere, and how a campfire warms your face from meters away. A full course in heat transfer studies the rates; thermodynamics mostly cares about the amounts.

How much heat does a given temperature change require? For a substance staying in one phase, Q = m c (T2 - T1), where c is the specific heat, the energy to raise one kilogram by one kelvin. Water's is famously large, about 4.18 kJ/kg K; iron's is about 0.45; air's, near room conditions and constant pressure, about 1.005. Example: warming 2 kg of water for tea from 20 C to 70 C takes Q = 2 x 4.18 x 50 = 418 kJ. Notice the temperature difference of 50 needs no Kelvin conversion: a difference is the same size on both scales. Only ratios and absolute values demand kelvins.

Key idea: Heat is energy transfer driven by temperature difference alone, moving by conduction, convection, and radiation, and single-phase heating obeys Q = m c (T2 - T1).

Signs, paths, and the rate of flow

Bookkeeping needs a sign convention, and this course uses the classic engine builder's version: heat into the system is positive, and work done by the system is positive. Both choices flatter the machines we love, which drink heat and deliver work. So a gas that absorbs 100 kJ of heat while doing 40 kJ of work on its piston has Q = +100 kJ and W = +40 kJ; a gas being compressed while losing heat has negative work and negative heat. Some physics texts flip the work sign; the physics is identical, but within this course we hold one convention so the first law can be written once, next module, as a fixed formula.

Remember also what kind of quantities these are. Work and heat are path functions: they describe the journey, not the destination, which is why we never say a system contains 50 kJ of heat or of work. A system contains energy. Heat and work exist only while crossing the boundary, like a payment that exists only in transit between accounts. Once Joule's paddle wheel stirred his water, the water held more internal energy, and no measurement on the water could reveal whether it had been warmed by friction or by flame.

Finally, rate matters as much as amount. Power is energy per unit time: 1 W = 1 J/s, 1 kW = 1 kJ/s. A 2 kW kettle raising 1.5 kg of water from 20 C to 100 C must deliver Q = 1.5 x 4.18 x 80 = 502 kJ, which at 2 kJ/s takes about 251 seconds, a little over four minutes. Your electricity bill charges by the kilowatt-hour, the energy of one kilowatt flowing for one hour: 1 kWh = 3600 kJ. The table below collects the energy units you will meet in engineering practice.

Unit	Equals	Where you meet it
1 kJ	1000 J	The working unit of this course
1 kWh	3600 kJ	Electricity billing
1 kcal (food Calorie)	about 4.19 kJ	Nutrition labels
1 Btu	about 1.055 kJ	American HVAC and heating equipment

Key idea: Heat in and work out are positive, both are path functions that exist only while crossing a boundary, and power is the rate of energy flow, with 1 kWh = 3600 kJ.

Common misconceptions

• A hot object contains a lot of heat. It contains internal energy. Heat is the transfer, not the store; once energy is inside, its origin as heat or work is untraceable.
• Adiabatic means constant temperature. Adiabatic means no heat transfer. An adiabatically compressed gas gets hotter purely from work input, as a hand pump demonstrates.
• Work in thermodynamics is only mechanical pushing. Electrical energy crossing a boundary, a rotating shaft, and a moving piston face are all work; the common test is that each could, in principle, have raised a weight.
• Temperature differences need converting to kelvins. A difference of 50 Celsius degrees is exactly 50 kelvins. Only absolute temperatures and their ratios require conversion.

Recap

• Systems store energy as kinetic ((1/2) m V^2), potential (m g z), and internal (U); transfers across the boundary are only work or heat.
• Boundary work is the area under the P-V path: P (V2 - V1) at constant pressure, P1 V1 ln(V2/V1) for isothermal ideal gas processes.
• Heat flows only where temperature differs, by conduction, convection, and radiation; adiabatic processes exchange none.
• Single-phase heating follows Q = m c (T2 - T1), with water's c about 4.18 kJ/kg K.
• Sign convention: heat into the system positive, work done by the system positive; both are path functions.
• Power is energy per time (kW = kJ/s), and 1 kWh = 3600 kJ.

Sources

1. OpenStax. (2016). Work, heat, and internal energy. In University physics volume 2. Rice University. openstax.org
2. OpenStax. (2016). Heat transfer, specific heat, and calorimetry. In University physics volume 2. Rice University. openstax.org
3. Britannica. (n.d.). Heat. In Encyclopaedia Britannica. britannica.com
4. Khan Academy. (n.d.). Thermodynamics. khanacademy.org

The big picture

Most of the machines in this course breathe. Engines inhale air, compressors squeeze it, gas turbines blast it out the back of airliners. To analyze any of them you need to answer questions like: how much air is in this cylinder? If I squeeze it to a tenth of its volume, what happens to its pressure and temperature? Remarkably, one short equation answers all of these for any gas that is far from condensing, and it does not care whether the gas is air, helium, nitrogen, or carbon dioxide. That equation is the ideal gas law, and today you learn to use it the way a practicing engineer does: quantitatively, in SI units, with the discipline of absolute temperature and absolute pressure you built in the last lessons.

The law has an unusual origin story: it was assembled like a jigsaw over 170 years. Robert Boyle found in 1662 that halving a gas's volume doubles its pressure at fixed temperature. Jacques Charles and Joseph Gay-Lussac, both balloonists in the great French ballooning craze, found around 1800 that gases expand in proportion to absolute temperature. Amedeo Avogadro proposed in 1811 that equal volumes of any gas at the same conditions hold equal numbers of molecules. Emile Clapeyron finally fused the pieces into one law in 1834. The kinetic theory of gases later explained why so many different substances obey one equation: at low density, molecules are so far apart that their individual sizes and mutual attractions stop mattering, and only their count, their speed, and the container remain in the story.

The plan: first the law and its constants, then three worked examples of increasing richness, then density and why hot air rises, and finally the honest boundaries of the model, because knowing when an equation fails is as much a part of engineering as using it.

The law and its constants

In molar form the law reads P V = n Ru T, where n is the number of kilomoles of gas and Ru = 8.314 kJ/kmol K is the universal gas constant, the same for every gas. Engineers usually prefer to count mass rather than moles, and dividing by the molar mass M converts one to the other. The result is the working form of this course: P V = m R T, where R = Ru/M is the specific gas constant, different for each gas because each gas has a different molar mass. Dividing through by mass gives the intensive version P v = R T, with v the specific volume. In every form, P is absolute pressure in kPa, T is absolute temperature in kelvins, and the units of R, kJ/kg K, knit them together.

Gas	Molar mass M (kg/kmol)	R = Ru/M (kJ/kg K)
Air	28.97	0.287
Nitrogen (N2)	28.01	0.297
Oxygen (O2)	32.00	0.260
Carbon dioxide (CO2)	44.01	0.189
Helium	4.003	2.077
Hydrogen (H2)	2.016	4.124

Air's value, R = 0.287 kJ/kg K, is worth memorizing: you will use it constantly. Notice the pattern in the table: light gases have large gas constants. A kilogram of helium contains seven times as many molecules as a kilogram of air, and since pressure comes from molecular impacts, that kilogram pushes correspondingly harder at the same temperature and volume.

Key idea: P V = m R T with R = Ru/M links absolute pressure, volume, mass, and absolute temperature for any gas far from condensation, with R = 0.287 kJ/kg K for air.

Worked example 1: weighing the air in a room

Question: what is the mass of the air in a room 5 m long, 4 m wide, and 3 m high, at 100 kPa and 25 C? Setup first, always: the system is the air, an ideal gas; the state is fixed by P and T; the volume is V = 5 x 4 x 3 = 60 m^3. Convert temperature to kelvins: T = 25 + 273 = 298 K. Now solve the law for mass:

m = P V / (R T) = (100 x 60) / (0.287 x 298) = 6000 / 85.5 = 70.2 kg.

Check the units: kPa times m^3 gives kJ; dividing by kJ/kg K times K leaves kilograms. The answer deserves a pause: the air in an ordinary bedroom outweighs most adults. We wade through a substantial ocean of gas and never feel it, because its pressure pushes equally from all sides. When a homework answer surprises you, do what we just did: audit the units, then ask whether the physics makes sense. Both checks passed here.

Worked example 2: the heated tire

A tire is, to good approximation, a rigid container, so V is constant, and no air leaks, so m is constant. Then the law says P/T = m R / V = constant: pressure tracks absolute temperature. Suppose a tire holds air at 320 kPa absolute on a 27 C morning (300 K), and highway driving warms it to 57 C (330 K). The new pressure is

P2 = P1 x (T2/T1) = 320 x (330/300) = 352 kPa.

A 10 percent rise in absolute temperature produced a 10 percent rise in absolute pressure, 32 kPa, which is why tire manuals tell you to check pressure cold. Notice how the Celsius habit would have wrecked this: 57/27 is not 330/300. The ratio only means something in kelvins.

Worked example 3: the general fixed-mass process

When P, V, and T all change for a fixed mass, write the law at both states and divide one by the other. The mass and R cancel, leaving the combined gas law: P1 V1 / T1 = P2 V2 / T2. Example: a gas at 100 kPa and 300 K occupying 0.4 m^3 is compressed to 0.1 m^3 while its temperature rises to 450 K. The final pressure is

P2 = P1 x (V1/V2) x (T2/T1) = 100 x (0.4/0.1) x (450/300) = 100 x 4 x 1.5 = 600 kPa.

Squeezing to a quarter of the volume quadrupled the pressure, and the temperature rise multiplied it by another half. This two-ratio structure, one factor for volume, one for temperature, is how you should read every fixed-mass gas problem, and it prefigures the compression stroke calculations we will do inside engine cylinders in the cycles module.

Key idea: For a fixed mass of ideal gas, P1 V1 / T1 = P2 V2 / T2, so rigid heating scales pressure with absolute temperature and compression scales it with the volume ratio.

Density, and why hot air rises

Rearranging P v = R T gives the gas density directly: rho = 1/v = P / (R T). Air at 100 kPa and 300 K has rho = 100 / (0.287 x 300) = 1.16 kg/m^3. Heat that air to 400 K at the same pressure and the density falls to 100 / (0.287 x 400) = 0.87 kg/m^3. There, in two lines, is the hot air balloon: every cubic meter of heated air inside the envelope weighs about 0.29 kg less than the cool air it displaced, and by Archimedes' principle the balloon feels that difference as lift. A 2800 m^3 balloon at those conditions buoys up roughly 2800 x 0.29 = 800 kg of envelope, basket, burner, and passengers. The same physics stratifies the air in your house, warm at the ceiling and cool at the floor, and drives the convection currents that make weather.

Key idea: Ideal gas density is rho = P / (R T), so at fixed pressure hotter gas is lighter, which is the entire secret of balloons and buoyant convection.

Where the ideal gas model ends

The kinetic theory behind the law assumes molecules are vanishingly small and do not attract one another. Both assumptions fail when molecules crowd together: at high pressures, and near the conditions where the gas is about to condense into liquid. So the honest rule is: the ideal gas law is excellent at low pressure and high temperature relative to the substance's condensation conditions, and it degrades as you approach the liquid. Room air is essentially perfect ideal gas territory; the errors are a fraction of a percent. Superheated steam at low pressure behaves nearly ideally too. But steam near a power plant boiler at 15 MPa, or refrigerant vapor about to condense in an air conditioner, can miss the law by tens of percent, and for those we use property tables, the subject of the next lesson.

Engineers quantify the miss with the compressibility factor, Z = P v / (R T). For a perfect ideal gas Z = 1 exactly; real gases run above or below. Z within about 1 percent of unity means the ideal model is fine; Z = 0.85 means a 15 percent error waiting to happen. You will not compute Z from first principles in this course, but you should carry the concept as a warning label: before applying P V = m R T to any gas, ask whether it is anywhere near becoming a liquid. If yes, reach for the tables.

Key idea: The ideal gas law holds far from condensation, its accuracy is measured by how close Z = P v / (R T) sits to 1, and near saturation you must use property tables instead.

Common misconceptions

• R is a universal constant. Ru = 8.314 kJ/kmol K is universal; the R in P V = m R T is Ru divided by the molar mass and differs gas by gas: 0.287 for air, 2.077 for helium.
• Gauge pressure and Celsius temperature work fine in the gas law. The law relates absolute quantities; a tire at 220 kPa gauge is at about 321 kPa absolute, and 27 C is 300 K. Using gauge or Celsius values produces garbage.
• The ideal gas law applies to any gas, always. It applies far from condensation; steam near a boiler or refrigerant near a condenser can deviate wildly, which is why steam tables exist.
• Heavier gases exert more pressure per kilogram. The opposite: a kilogram of light gas holds more molecules, so at fixed V and T it exerts more pressure, which is the meaning of its larger R.

Recap

• The working forms are P V = m R T and P v = R T, with P absolute (kPa), T absolute (K), and R = Ru/M in kJ/kg K.
• Ru = 8.314 kJ/kmol K for every gas; R = 0.287 kJ/kg K for air.
• Fixed-mass processes obey P1 V1 / T1 = P2 V2 / T2: rigid heating scales P with T, isothermal compression scales P with the inverse volume ratio.
• Density rho = P / (R T) falls as gas warms at constant pressure, producing buoyancy, balloons, and convection.
• The model is trustworthy far from condensation (Z near 1) and must give way to property tables near saturation, high pressure, or the critical point.

Sources

1. OpenStax. (2016). Molecular model of an ideal gas. In University physics volume 2. Rice University. openstax.org
2. Britannica. (n.d.). Ideal gas law. In Encyclopaedia Britannica. britannica.com
3. National Institute of Standards and Technology. (n.d.). Thermophysical properties of fluid systems. NIST Chemistry WebBook. webbook.nist.gov
4. Khan Academy. (n.d.). Thermodynamics. khanacademy.org

The big picture

The last lesson ended with a warning: the ideal gas law abandons you exactly where power engineering lives. A steam power plant boils water on purpose, an air conditioner condenses refrigerant on purpose, and both machines spend their working lives shuttling between liquid and vapor. Why court such complicated behavior? Because phase change is nature's energy bargain. Boiling a kilogram of water absorbs 2257 kJ while heating that same kilogram from ice-cold to boiling absorbs only 419 kJ. The plateau is where the energy is, so that is where the machines go, and an engineer who cannot compute in the two-phase region cannot analyze a single real power cycle. Today you learn the map of phase change and the tables that replace the ideal gas law when a substance is anywhere near its liquid.

Our subject is the pure substance: fixed chemical composition throughout, like water, nitrogen, or refrigerant R-134a. A mixture of liquid water and its own steam still counts, because the composition is H2O everywhere. Air counts as effectively pure while it stays gaseous, since its composition does not change. We will speak mostly of water, partly because steam built the industrial world and partly because its tables are the ones every engineer learns first, but every pure substance plays the same game with different numbers.

The plan: walk one kilogram of water through a constant-pressure heating, naming each regime; assemble those walks into the T-v diagram with its dome and critical point; define quality, the variable that locates you inside the dome; then work a full table problem end to end.

One kilogram of water, heated at constant pressure

Put 1 kg of liquid water at 25 C in a piston-cylinder held at 101.325 kPa, one standard atmosphere, and add heat slowly. At first the water just warms: 30 C, 60 C, 90 C, swelling only slightly. In this regime it is a compressed liquid, also called a subcooled liquid: liquid that is not about to boil. At 100 C something changes. The water has reached the boiling threshold for this pressure, and we call it a saturated liquid: still entirely liquid, but any further heat begins converting it to vapor.

Keep heating and the temperature does something students never quite believe until they watch a thermometer do it: it stops rising. The added energy goes entirely into tearing molecules out of the liquid and into vapor, not into speeding them up, and the mixture sits locked at 100 C while it turns, gram by gram, into steam. Through this plateau the system is a saturated liquid-vapor mixture, two phases coexisting in equilibrium. The energy absorbed across the whole plateau, liquid to vapor, is the enthalpy of vaporization hfg, the latent heat: 2257 kJ/kg for water at one atmosphere. When the last droplet vanishes we have saturated vapor: pure steam, but steam on the edge, since removing any heat starts condensation. Only with further heating does the temperature move again, to 120 C, 200 C, 300 C, producing superheated vapor, steam safely away from condensing. There, and only there, does steam begin to act roughly like an ideal gas.

Key idea: At fixed pressure a pure substance heats as compressed liquid, halts at the saturation temperature while boiling absorbs the latent heat hfg, and only resumes warming as superheated vapor.

Saturation: temperature and pressure are married

The plateau sat at 100 C because the pressure was one atmosphere. Change the pressure and the boiling point moves: at a given pressure, boiling happens at one and only one temperature, the saturation temperature Tsat, and equivalently at a given temperature there is one saturation pressure Psat. Water at 200 kPa boils at 120.2 C, which is the entire trick of a pressure cooker: trapping steam raises the pressure, which raises the boiling point, and food cooks much faster at 120 C than at 100 C. Near the summit of Mount Everest the atmosphere is only about 31 kPa and water boils near 70 C, too cool to brew decent tea, as generations of climbers have complained. Inside a power plant boiler at 8 MPa, water does not boil until 295 C.

This marriage of T and P during phase change is why the state postulate carried a warning back in lesson one. Inside the dome, temperature and pressure are not independent: quoting both tells you only that you are somewhere on the plateau, not where. To locate the state you need one more number, and that number is quality, defined below.

Key idea: Each pressure has exactly one boiling temperature, Tsat(P): raise the pressure and boiling shifts hotter, lower it and water boils cool, so T and P are not independent inside the two-phase region.

The T-v diagram, the dome, and the critical point

Now run that constant-pressure experiment at many different pressures and plot every state on a graph of temperature versus specific volume. Each run traces a rising line, a flat plateau, and a rising tail. Connect all the left ends of the plateaus (saturated liquid points) and all the right ends (saturated vapor points) and you get a dome-shaped curve, the vapor dome. Left of the dome lives compressed liquid; under the dome, two-phase mixtures; right of it, superheated vapor. This picture, the T-v diagram, is the single most useful mental image in the course. Sketch it now, label the three regions, and sketch it again on every problem you solve this module.

As pressure climbs, the plateaus shrink: the saturated liquid grows less dense and the saturated vapor more dense, the two ends creeping toward each other. At water's critical point, 22.06 MPa and 373.95 C, the plateau vanishes entirely: liquid and vapor become indistinguishable, and above this pressure water passes from liquid-like to gas-like continuously, with no boiling at all. Modern supercritical power plants run their boilers above this point partly to dodge the mechanics of boiling. At the other extreme sits the triple point, 0.01 C and 0.6117 kPa for water, the lone condition where solid, liquid, and vapor coexist; below its pressure, ice sublimates directly to vapor, which is how freeze-drying works and why frost disappears on cold dry days without ever melting.

P (kPa)	Tsat (C)	vg (m^3/kg)	hf (kJ/kg)	hfg (kJ/kg)	hg (kJ/kg)
10	45.8	14.67	191.8	2392.8	2584.7
101.325	100.0	1.673	419.0	2257.0	2676.1
200	120.2	0.8857	504.7	2201.9	2706.7
1000	179.9	0.1944	762.8	2015.3	2778.1
8000	295.1	0.0235	1316.6	1441.4	2758.0

Read the trends in this excerpt from the saturation table: hotter boiling at higher pressure, vapor volumes collapsing by a factor of 600 across the range, and the latent heat hfg withering from 2393 toward zero as the critical point approaches. Numbers like these come from decades of measurement; the authoritative modern source is the NIST WebBook, which will generate any of them for you on demand.

Key idea: The vapor dome on the T-v diagram separates compressed liquid, two-phase mixture, and superheated vapor, with the plateaus shrinking to nothing at the critical point.

Quality: locating yourself inside the dome

Inside the dome the state is part liquid, part vapor, and the locating variable is the quality: x = (mass of vapor) / (total mass). Quality runs from 0 at saturated liquid to 1 at saturated vapor and has no meaning outside the dome. Every specific property of the mixture is then a weighted average of the saturated liquid value (subscript f) and saturated vapor value (subscript g). For specific volume: v = vf + x vfg, where vfg = vg - vf. For enthalpy: h = hf + x hfg. The same pattern serves internal energy and, later, entropy. Given any one mixture property, you can solve backwards for x and then compute all the others: that is the whole art of two-phase table work.

Worked example, end to end. A rigid 0.5 m^3 vessel holds 2 kg of water at 200 kPa. Find the temperature, the quality, and the enthalpy. First, the specific volume is fixed by geometry: v = V/m = 0.5/2 = 0.25 m^3/kg. Second, look at the 200 kPa saturation row: vf = 0.001061, vg = 0.8857. Our v lies between them, so the state is under the dome, and the temperature must be Tsat = 120.2 C; no other answer is possible at this pressure. Third, solve for quality: x = (v - vf) / (vg - vf) = (0.25 - 0.001061) / (0.8857 - 0.001061) = 0.2489 / 0.8846 = 0.281. About 28 percent of the mass, 0.56 kg, is vapor, and that vapor fills nearly all the volume. Fourth, the enthalpy: h = hf + x hfg = 504.7 + 0.281 x 2201.9 = 504.7 + 618.7 = 1123.4 kJ/kg. One geometric fact plus one table row produced the entire state.

Two closing table skills. If v (or h) lands above vg, the state is superheated: turn to the superheat table, which lists properties on a grid of pressure and temperature, and interpolate as needed. If it lands below vf, the state is compressed liquid, and there is a gracious shortcut: liquid properties barely feel pressure, so use the saturated liquid values at the same temperature, v approximately vf(T) and h approximately hf(T). The approximation is usually good to a percent or better.

Key idea: Quality x = mvapor/mtotal locates a state inside the dome, and mixture properties follow v = vf + x vfg and h = hf + x hfg.

Common misconceptions

• Adding heat always raises temperature. During phase change at constant pressure it does not; the plateau absorbs the latent heat at constant temperature, which is precisely what makes boilers and condensers such effective energy movers.
• Water boils at 100 C, period. Water boils at Tsat for its pressure: 70 C near Everest's summit, 120 C in a pressure cooker, 295 C in an 8 MPa boiler.
• Quality is the fraction of the volume that is vapor. Quality is a mass fraction. In our worked example the vapor was 28 percent of the mass but occupied more than 99 percent of the volume, because vapor is a thousand times less dense.
• Steam is visible white mist. True steam, water vapor, is invisible; the white plume above a kettle is condensed liquid droplets. Superheated steam in a pipe cannot be seen at all, which is one reason industrial steam leaks are treated with such respect.

Recap

• A pure substance heated at constant pressure passes through compressed liquid, saturated liquid, two-phase mixture (constant T), saturated vapor, and superheated vapor.
• Saturation temperature and pressure are locked pairs: 100 C at 101.325 kPa, 120.2 C at 200 kPa, 295 C at 8 MPa for water.
• The T-v diagram's dome bounds the two-phase region, topped by the critical point (22.06 MPa, 373.95 C for water), where liquid and vapor merge.
• Latent heat is large: vaporizing water at 1 atm takes 2257 kJ/kg, over five times the energy of heating it from 0 C to 100 C.
• Inside the dome, x = mvapor/mtotal, v = vf + x vfg, h = hf + x hfg; above the dome use superheat tables; compressed liquid approximates saturated liquid at the same temperature.

Sources

1. National Institute of Standards and Technology. (n.d.). Thermophysical properties of fluid systems. NIST Chemistry WebBook. webbook.nist.gov
2. OpenStax. (2016). Phase changes. In University physics volume 2. Rice University. openstax.org
3. Wikipedia. (n.d.). Latent heat. Wikimedia Foundation. en.wikipedia.org
4. Engineering LibreTexts. (n.d.). Thermodynamics bookshelf. LibreTexts. eng.libretexts.org

The big picture

In 1845 James Joule stirred water with a paddle wheel driven by falling weights, measured the temperature creep upward with thermometers he could read to a few thousandths of a degree, and established one of the most consequential exchange rates ever published: a fixed quantity of mechanical work always produces the same rise in a body's stored energy as a fixed quantity of heat. Work and heat, the two transactions of lesson three, deposit into the same account. That single fact, generalized, is the first law of thermodynamics: energy can change form and change hands, but the books always balance. Nothing is created; nothing vanishes.

You have been using this idea informally since childhood, every time you assumed a hot drink cools because its energy leaks to the room rather than simply disappearing. Today we make it an equation you can compute with, and the computing form is almost embarrassingly short: for a closed system, Q - W = delta U. The art is never in the equation. It is in drawing the boundary, deciding what crosses it, catching the signs, and knowing which terms are zero for the process at hand. That art is today's entire agenda: we will run the balance on rigid tanks, stirred tanks, and moving pistons, meet the property enthalpy that makes constant-pressure work bookkeeping automatic, and close with what the first law forbids.

The energy balance

Write the first law the way you would write any conservation statement: energy in, minus energy out, equals the change in energy stored. For a closed system, only heat and work cross the boundary, and with our sign convention (heat into the system positive, work by the system positive) the balance reads:

Q - W = delta E = delta U + delta KE + delta PE.

Most closed systems on an engineer's desk are not flying or falling, so the kinetic and potential terms drop and the working form is Q - W = delta U, often written per unit mass as q - w = delta u. Read it aloud in words: the heat added, less the work extracted, is what remains behind as stored internal energy. Every term is in kilojoules; every analysis starts by writing this line and crossing out the terms the process kills.

First numbers. A gas in a piston-cylinder absorbs 100 kJ of heat while doing 40 kJ of work pushing its piston. The internal energy change is delta U = Q - W = 100 - 40 = +60 kJ: sixty kilojoules deposited. Now reverse the machine: a gas being compressed receives 80 kJ of work (so W = -80 kJ, work done by the system is negative) while losing 30 kJ of heat to cooling water (Q = -30 kJ). Then delta U = -30 - (-80) = +50 kJ. The gas warmed even while losing heat, because the compression paid in more than the cooling took out. If the signs feel slippery, do not memorize outcomes; re-derive them from in minus out every time, the way accountants do not memorize whether checks are good news.

Key idea: For a closed system, Q - W = delta U: heat in, less work out, equals the change in stored internal energy, with signs handled by the convention rather than by intuition.

Constant volume: the rigid tank

Seal gas in a rigid tank and heat it. The boundary cannot move, so boundary work is zero, and if no shaft or wires enter, W = 0 entirely. The first law collapses to Q = delta U: every joule of heat lands in internal energy, which is why rigid-vessel heating raises temperature and pressure so briskly. This is the configuration of an aerosol can in a fire, and the reason the label warns you.

Now the stirred version, Joule's own. Take a rigid, perfectly insulated tank of water and spin a paddle wheel inside it with 500 kJ of shaft work from outside. Insulation makes Q = 0. The work is done on the system, so W = -500 kJ, and the balance gives delta U = 0 - (-500) = +500 kJ. The water warms exactly as if 500 kJ of heat had entered, and once the stirring stops, no measurement on the water can reveal which it was. That interchangeability is the deep content of Joule's experiment: internal energy is a property with no memory of how it arrived. It is also your first meeting with an idea the second law will sharpen later: work converted entirely into internal energy, but no one has ever unstirred the tank and gotten the 500 kJ of work back out.

Key idea: With the boundary fixed and no shafts or wires, Q = delta U; work done on a system counts as negative W and deposits energy exactly as heat would.

Constant pressure, and the invention of enthalpy

Now let the boundary move. A piston-cylinder holds air at a constant 150 kPa, kept there by the piston's weight and the atmosphere above it. We heat the air and it expands from 0.2 m^3 to 0.5 m^3. Two things happen at once: the air's internal energy rises, and the air does boundary work W = P (V2 - V1) = 150 x 0.3 = 45 kJ shoving the piston. Suppose property tables tell us the internal energy rose by delta U = 80 kJ. The heat required is Q = delta U + W = 80 + 45 = 125 kJ. Notice the extra 45: at constant pressure you must pay for the expansion work on top of the energy stored. Heating things in open or piston-loaded containers always carries this surcharge.

Engineers run this exact calculation so often that they invented a property to automate it. Define the enthalpy: H = U + PV, or per kilogram, h = u + P v. For a constant-pressure process, delta H = delta U + P delta V, which is precisely delta U plus the boundary work. So the first law for constant-pressure heating becomes simply Q = delta H. One property change captures both the stored energy and the expansion surcharge. Check it on our example: delta H = 80 + 45 = 125 kJ = Q. This is why enthalpy, not internal energy, is the headline column in steam tables, and why hfg, the enthalpy of vaporization, was the right measure of boiling's cost in the last lesson: boiling in an open boiler is a constant-pressure process. Try it: boiling off 0.5 kg of water in a pot at one atmosphere requires Q = m hfg = 0.5 x 2257 = 1128.5 kJ, enthalpy doing the whole job in one line. Enthalpy will graduate to an even bigger role when mass starts flowing through open systems in two lessons; the groundwork is laid here.

Key idea: Enthalpy h = u + P v bundles internal energy with expansion work so that constant-pressure heating obeys Q = delta H, the reason tables tabulate h so prominently.

Two instructive extremes: free expansion and the cycle

Here is a process that looks paradoxical until the balance dissolves it. An insulated rigid vessel is divided by a membrane: gas on one side, vacuum on the other. Puncture the membrane. The gas rushes to fill everything. What happened to its energy? Insulation gives Q = 0. And though the gas expanded, it pushed against nothing: a vacuum offers no resisting pressure, so no work left the system, W = 0. Therefore delta U = 0. For an ideal gas, whose internal energy depends on temperature alone (next lesson makes this precise), the temperature ends where it began. Joule tried this experiment too, and the near-zero temperature change he found is part of why we model gas internal energy as u(T). The lesson generalizes: expansion only costs energy when it pushes something.

The other extreme is the cycle, the configuration every engine lives in. Run a closed system around any loop of processes back to its starting state. Internal energy is a property, so it returns to its starting value: delta U = 0 for the loop, no exceptions. The first law then says Q(net) = W(net): over a cycle, the net heat taken in equals the net work delivered. An engine cannot output more work per lap than the net heat it absorbs. A machine claimed to do so, to create energy from nothing, is a perpetual motion machine of the first kind, and the United States Patent and Trademark Office has for over a century treated such applications as inoperable precisely because they contradict this law. When an advertisement promises a generator that outputs more energy than it consumes, you now possess the one-line audit: delta U = 0 around a cycle, so net out cannot exceed net in. No exceptions have ever been observed.

Key idea: Free expansion against vacuum changes no energy (Q = 0, W = 0, so delta U = 0), and around any cycle delta U = 0 forces net work to equal net heat, forbidding first-kind perpetual motion.

A worked audit, start to finish

Let us finish with the full discipline on one problem. A rigid tank holds 2 kg of air at 300 K. A paddle wheel does 120 kJ of work on the air while 40 kJ of heat leaks out through the walls. Find the internal energy change and comment. Step one, system: the air, closed, rigid. Step two, balance: Q - W = delta U. Step three, signs: heat leaks out, Q = -40 kJ; work is done on the system, W = -120 kJ. Step four, arithmetic: delta U = -40 - (-120) = +80 kJ. Step five, sanity: the air gained energy because the paddle deposited more than the leak removed; per kilogram that is 40 kJ/kg, and next lesson's specific heats will let us convert it to a temperature rise (about 56 K, it will turn out). The five steps never change. Boundary, balance, signs, arithmetic, sanity: run them in order and the first law will never betray you, on an exam or on a design review.

Common misconceptions

• The first law says energy in equals energy out. Only for steady or cyclic operation. In general the difference between in and out is exactly what accumulates inside: Q - W = delta U.
• If a system loses heat, its internal energy must fall. Not if work is being done on it faster than heat leaves; a compressed, cooled gas can warm, as our 80 kJ example showed.
• Enthalpy is just a fancy word for heat content. Enthalpy is a property, u + P v, defined at every state whether or not any heating is happening; it merely equals the heat transfer for the special case of a constant-pressure process.
• A gas expanding always does work. Work requires pushing against resistance. Expansion into vacuum moves no piston and transfers no energy, which is why free expansion leaves internal energy unchanged.

Recap

• The first law for closed systems: Q - W = delta U (plus kinetic and potential terms when they matter), with Q positive in and W positive out.
• Rigid, unstirred tanks give Q = delta U; work done on a system enters as negative W and is indistinguishable, once inside, from heat.
• Constant-pressure processes obey Q = delta H, where enthalpy H = U + PV absorbs the boundary-work surcharge automatically.
• Free expansion against vacuum: Q = 0, W = 0, delta U = 0.
• Around any cycle delta U = 0, so Q(net) = W(net), and machines that output energy from nothing (perpetual motion of the first kind) are impossible.
• The audit discipline: boundary, balance, signs, arithmetic, sanity, in that order, every time.

Sources

1. OpenStax. (2016). First law of thermodynamics. In University physics volume 2. Rice University. openstax.org
2. OpenStax. (2016). Thermodynamic processes. In University physics volume 2. Rice University. openstax.org
3. Britannica. (n.d.). Thermodynamics. In Encyclopaedia Britannica. britannica.com
4. Khan Academy. (n.d.). Thermodynamics. khanacademy.org

The big picture

The first law hands you energy changes in kilojoules; thermometers hand you temperatures in kelvins. Engineering lives in the exchange rate between them. Pour 100 kJ into a tank of air: how hot does it get? Compress air in a diesel cylinder to one eighteenth of its volume: does it get hot enough to ignite fuel with no spark plug? (Yes, and by the end of this lesson you will compute exactly how hot.) The exchange rate is the specific heat, and the wrinkle that makes this lesson necessary is that a gas does not have one specific heat. It has two, they differ by exactly the gas constant R, and their ratio k runs the adiabatic formulas that govern every engine's compression stroke. This is the most formula-dense lesson of the course, and also one of the most used: the numbers you meet here, 0.718, 1.005, and 1.4 for air, will appear in nearly every cycle calculation from now to the final exam.

The plan: define the two specific heats and see why an ideal gas's internal energy cares only about temperature; establish the relations cp = cv + R and k = cp/cv; work matched heating examples that show the two heats in action; derive what adiabatic compression does to temperature; and finish with the simpler world of solids and liquids, where one c suffices and mixing problems settle into a single balance.

Two specific heats, and Joule's simplification

Lesson three defined specific heat loosely as energy per kilogram per kelvin. For gases we must be precise about the conditions. Heat a gas in a rigid tank and every joule lands in internal energy; the temperature rise per unit energy defines the specific heat at constant volume, cv, and delta u = cv delta T. Heat the same gas under a floating piston and part of each joule leaks out as expansion work; more heat is needed per degree, and that larger rate defines the specific heat at constant pressure, cp, with delta h = cp delta T. For air near room temperature, cv = 0.718 kJ/kg K and cp = 1.005 kJ/kg K.

Now the simplification that makes gas thermodynamics tractable. Joule's free expansion experiment (last lesson) showed that changing an ideal gas's volume and pressure without changing its temperature leaves its internal energy untouched. Conclusion: for an ideal gas, u depends on temperature alone. And since h = u + P v = u + R T, enthalpy also depends on temperature alone. This has a consequence students persistently underuse: the formulas delta u = cv delta T and delta h = cp delta T hold for any ideal gas process whatsoever, not merely constant-volume or constant-pressure ones. The names describe how the two heats are measured, not where the formulas apply. In an adiabatic compression, delta u is still cv delta T. In a wild, wandering process from state 1 to state 2, delta h is still cp delta T. Temperature change alone sets the energy change; the path sets only how that change was paid for, in heat versus work.

Key idea: Ideal gas internal energy and enthalpy depend on temperature only, so delta u = cv delta T and delta h = cp delta T on every path, whatever the pressure and volume are doing.

The family relations: cp = cv + R and k

The two heats are not independent. Take the definition h = u + R T and difference it: delta h = delta u + R delta T, so cp delta T = cv delta T + R delta T, and dividing by delta T gives cp = cv + R. The gap between the two specific heats is exactly the gas constant, and it is no coincidence: R delta T is precisely the boundary work a kilogram of ideal gas does while expanding at constant pressure through delta T. The constant-pressure heat is the constant-volume heat plus the work surcharge. Check air: 0.718 + 0.287 = 1.005. The books balance to the third decimal.

The ratio k = cp/cv, called the specific heat ratio, will run the adiabatic machinery shortly. For air, k = 1.005/0.718 = 1.400. Kinetic theory explains the values via equipartition: each way a molecule can store energy contributes (1/2)R to cv. A monatomic atom like helium can only translate in three directions: cv = (3/2)R and k = 5/3, about 1.667. A diatomic molecule like N2 or O2 adds two rotations: cv = (5/2)R and k = 7/5 = 1.4, which is why air, essentially a diatomic mixture, lands there. Bulkier molecules like CO2 store energy more ways still, pushing cv up and k down. At high temperatures vibration awakens and the constants drift, so precise work uses temperature-dependent tables; this course, following universal practice for a first pass, uses the room-temperature constants below.

Gas	cp (kJ/kg K)	cv (kJ/kg K)	R (kJ/kg K)	k = cp/cv
Air	1.005	0.718	0.287	1.400
Nitrogen (N2)	1.039	0.743	0.297	1.400
Oxygen (O2)	0.918	0.658	0.260	1.395
Helium	5.193	3.116	2.077	1.667
Carbon dioxide (CO2)	0.846	0.657	0.189	1.289

Key idea: cp = cv + R, because constant-pressure heating pays an R delta T work surcharge per kilogram, and the ratio k = cp/cv is about 1.4 for air and other diatomic gases.

Matched examples: the same gas, two containers

Take 2 kg of air from 300 K to 400 K, twice. First in a rigid tank: no work, so Q = delta U = m cv delta T = 2 x 0.718 x 100 = 143.6 kJ. Second under a constant-pressure piston: now Q = delta H = m cp delta T = 2 x 1.005 x 100 = 201.0 kJ. Same gas, same temperature rise, but the piston case cost 57.4 kJ more. Where did the difference go? Into boundary work: W = m R delta T = 2 x 0.287 x 100 = 57.4 kJ, to the decimal. Internal energy rose by the same 143.6 kJ in both cases, as it must, since delta U = m cv delta T on any path. The three numbers, 143.6 stored, 57.4 worked, 201.0 paid, audit perfectly, and running such audits is how you should check every gas problem you solve.

This also settles a promise from last lesson. There, a paddle wheel and a heat leak left 2 kg of air with delta U = +80 kJ, and I claimed a 56 K temperature rise. Now you can verify: delta T = delta U / (m cv) = 80 / (2 x 0.718) = 55.7 K. Claim honored.

Adiabatic compression: why diesels need no spark plug

Now the marquee application. Compress a gas quickly enough that no significant heat escapes: Q = 0, an adiabatic process. The first law says the work you push in goes entirely into internal energy, so the temperature must climb. For a quasi-equilibrium adiabatic (the idealization is called isentropic, a name that will make sense in the entropy lesson), combining the first law with the ideal gas law yields three equivalent path equations worth writing on your formula sheet: P v^k is constant, T v^(k-1) is constant, and T2/T1 = (P2/P1)^((k-1)/k). For air, the exponents are k - 1 = 0.4 and (k-1)/k = 0.2857.

Apply the temperature-volume form to a diesel engine, which squeezes its air to about one eighteenth of the intake volume. Starting from 300 K: T2 = T1 (V1/V2)^(k-1) = 300 x 18^0.4 = 300 x 3.18 = 953 K, about 680 C, comfortably beyond diesel fuel's autoignition temperature. Inject fuel into that furnace and it lights itself: the entire reason diesel engines carry no spark plugs is the number you just computed. A gentler example with the pressure form: air compressed from 100 kPa and 300 K to 800 kPa reaches T2 = 300 x 8^0.2857 = 300 x 1.811 = 543 K, or 270 C, which is why the outlet pipe of a workshop compressor is painfully hot and why compressed air systems need aftercoolers. The same physics runs in reverse: adiabatic expansion cools, which is how a turbine extracts energy from hot gas and, incidentally, why escaping CO2 can frost a fire extinguisher's horn.

Key idea: Adiabatic quasi-equilibrium ideal gas processes obey P v^k = constant and T2/T1 = (P2/P1)^((k-1)/k), so compression heats sharply (300 K becomes 953 K at ratio 18) and expansion cools.

Solids and liquids: one c, and calorimetry

Solids and liquids are nearly incompressible: heating them produces negligible expansion work, so the constant-volume and constant-pressure heats collapse into a single specific heat c, with delta u approximately c delta T. Water's is 4.18 kJ/kg K, ice's 2.11, aluminum's 0.897, iron's 0.45. Water's enormous value, among the highest of common substances, is why coastal climates are mild, why it takes so long to boil a pot, and why water is the coolant of choice in everything from car engines to power plant condensers.

The classic use is calorimetry: drop a hot object into cool liquid inside an insulated container and find the shared final temperature. The balance is pure first law: with no heat escaping, the energy the hot object loses equals the energy the cool one gains. Example: a 0.5 kg aluminum block at 150 C is quenched in 2 kg of water at 20 C. Set the losses equal to the gains with T as the final temperature: 0.5 x 0.897 x (150 - T) = 2 x 4.18 x (T - 20). The left side is 0.4485 (150 - T) and the right is 8.36 (T - 20); solving, 67.28 + 167.2 = 8.809 T, so T = 26.6 C. The water barely warmed while the block plunged 123 degrees, because the water side carries nearly twenty times the heat capacity (mass times c). Blacksmiths, coffee machines, and nuclear plant designers all live by this arithmetic.

Key idea: Incompressible solids and liquids need only one specific heat, and insulated mixing problems reduce to setting one side's m c delta T loss equal to the other's gain.

Common misconceptions

• delta u = cv delta T only works for constant-volume processes. For ideal gases it works for every process, because u depends on temperature alone; constant volume is merely how cv is measured.
• cp and cv are two independent material quirks. They differ by exactly R, the per-degree expansion work of a kilogram of gas at constant pressure: cp = cv + R, always, for ideal gases.
• Compressing a gas quickly cannot change its temperature much because there is no time for heating. Backwards: speed prevents heat from escaping, so all the compression work lands in internal energy and the temperature rise is at its largest. That is the diesel's ignition source.
• All substances need two specific heats. Only compressible substances do; solids and liquids expand so little that one c covers engineering work.

Recap

• cv governs internal energy (delta u = cv delta T) and cp governs enthalpy (delta h = cp delta T), on any ideal gas path; for air, cv = 0.718 and cp = 1.005 kJ/kg K.
• cp = cv + R, and k = cp/cv is 1.4 for air, 1.667 for monatomic gases, lower for complex molecules.
• Matched heating of 2 kg air through 100 K: 143.6 kJ rigid versus 201.0 kJ constant-pressure, the 57.4 kJ difference being boundary work m R delta T.
• Adiabatic ideal gas processes follow P v^k = constant, T v^(k-1) = constant, T2/T1 = (P2/P1)^((k-1)/k); compression ratio 18 lifts 300 K air to 953 K, igniting diesel fuel unaided.
• Solids and liquids use one c (water 4.18 kJ/kg K), and insulated mixing balances m c delta T lost against m c delta T gained.

Sources

1. OpenStax. (2016). Heat capacities of an ideal gas. In University physics volume 2. Rice University. openstax.org
2. OpenStax. (2016). Adiabatic processes for an ideal gas. In University physics volume 2. Rice University. openstax.org
3. OpenStax. (2016). Heat capacity and equipartition of energy. In University physics volume 2. Rice University. openstax.org
4. Khan Academy. (n.d.). Thermodynamics. khanacademy.org

The big picture

Nearly everything interesting in energy engineering flows. Steam pours through a turbine at hundreds of kilograms per second; air streams through a jet engine; refrigerant circulates endlessly through your air conditioner. The closed-system first law cannot follow a particular kilogram through such a machine without heroic bookkeeping. So engineers flip the viewpoint: nail the boundary in space, let the matter stream through it, and audit the region itself. This is the control volume analysis promised back in lesson one, and it comes with a gift: once a machine reaches steady state, running without change hour after hour, nothing inside the boundary accumulates. Mass in equals mass out, energy in equals energy out, and the analysis collapses to one algebraic line per device.

That line, the steady-flow energy equation, is the single most used equation in applied thermodynamics. Today we build it, meet the small surprise hiding in it (enthalpy takes over from internal energy, for a reason you can now understand), and then tour the five devices from which almost every energy system is assembled: nozzles, turbines, compressors, throttling valves, and heat exchangers. Master these five and next module's power plants and refrigerators will assemble before your eyes like kit furniture.

Counting the flow: mass balance

Start with mass. The mass flow rate through a port of cross-sectional area A, where fluid of density rho moves at velocity V, is mdot = rho A V, in kilograms per second; equivalently mdot = A V / v using specific volume. Example: water at 1000 kg/m^3 flowing at 2 m/s through a pipe of 0.005 m^2 area carries mdot = 1000 x 0.005 x 2 = 10 kg/s. At steady state, whatever enters must leave: mdot(in) = mdot(out). This innocent statement, the continuity equation, explains why a nozzle's narrowing throat forces fluid to accelerate and why a river quickens through a gorge: same mass per second, smaller area, higher speed.

Key idea: mdot = rho A V, and in steady state the mass flow into a control volume equals the mass flow out.

Flow work, and enthalpy's promotion

Now energy. Each kilogram crossing the boundary carries its stored energy: internal energy u, kinetic energy V^2/2, potential energy g z. But there is one more entry, easy to miss and essential. To push a kilogram of fluid into the control volume against the local pressure, the fluid behind it must do work on it: force times distance works out to exactly P v per kilogram. This flow work is the price of admission at the inlet, and the control volume pays it back at the exit, shoving each departing kilogram out against the exit pressure. So the energy delivered per kilogram crossing a port is u + P v plus the kinetic and potential terms. And u + P v has a name you already know: enthalpy. This is enthalpy's real job. It was convenient shorthand for closed systems at constant pressure; for open systems it is the natural currency, the stored energy plus the admission fee, and this is why property tables headline h.

Balancing energy in against energy out for a steady control volume with heat transfer rate Qdot in and shaft power Wdot out gives the steady-flow energy equation for a single stream:

Qdot - Wdot = mdot [ (h2 - h1) + (V2^2 - V1^2)/2 + g (z2 - z1) ]

with one unit caution: h is in kJ/kg but V^2/2 and g z emerge in J/kg, so divide the kinetic and potential terms by 1000 before adding. Calibrate the sizes: a brisk 45 m/s pipe flow carries V^2/2 = 1012 J/kg, a single kilojoule per kilogram, and a 100 m elevation drop carries about 1 kJ/kg, while enthalpy changes through turbines and boilers run to hundreds or thousands of kJ/kg. Hence the standing engineering habit: neglect kinetic and potential terms unless the device's whole purpose is velocity (nozzles) or height (pumped hydro). Every device analysis below is this one equation with different terms crossed out.

Key idea: Flowing streams carry enthalpy, not just internal energy, because each kilogram pays P v of flow work at the boundary; the steady-flow energy equation balances h, kinetic, and potential terms against heat and shaft work.

The device gallery

Nozzles and diffusers. A nozzle is a shaped duct that trades enthalpy for speed; a diffuser runs the trade backwards, slowing flow to raise its pressure. No shaft, negligible heat, so the balance is h1 + V1^2/2 = h2 + V2^2/2. Starting from rest, V2 = sqrt(2000 (h1 - h2)) with h in kJ/kg. Steam entering slowly at 3000 kJ/kg and leaving at 2900 kJ/kg exits at V2 = sqrt(2000 x 100) = 447 m/s, faster than an airliner, from a 100 kJ/kg enthalpy drop that a boiler would call small change. This is how rocket engines convert combustion heat into thrust and how the stationary nozzle ring in a steam turbine creates the jets that strike the blades. A diffuser sits at a jet engine's inlet doing the reverse: taming 250 m/s flight air into slow, pressurized feed for the compressor.

Turbines. A turbine extracts shaft work from a high-enthalpy stream: blades redirect the flow, the flow pushes back, the rotor turns. Modeled adiabatic (large machines lose under a percent of throughput as heat) with negligible velocity change, the balance gives simply w = h1 - h2 per kilogram, and power Wdot = mdot (h1 - h2). Steam entering at 3350 kJ/kg and leaving at 2550 kJ/kg while flowing at 25 kg/s delivers Wdot = 25 x 800 = 20,000 kW = 20 MW, the electricity of a small city from a machine the size of a bus. Read the equation's message: everything about a turbine's output is in how far you can drop the enthalpy, which is why the cycles module will obsess over inlet superheat and exhaust vacuum.

Compressors, fans, and pumps. The same machine mirrored: shaft work in, enthalpy up, w(in) = h2 - h1. An adiabatic air compressor taking air from 300 K to 500 K demands w = cp delta T = 1.005 x 200 = 201 kJ/kg; at 2 kg/s that is 402 kW, a large industrial motor. Pumps do the same job on liquids and get a spectacular discount: because liquid specific volume is tiny, the ideal pump work is w = v delta P, and pressurizing water by a full 5990 kPa costs only about 0.00101 x 5990 = 6 kJ/kg, roughly 3 percent of what compressing gas through a similar span would take. Tuck that 6 kJ/kg away; it returns as the Rankine cycle's best structural feature.

Throttling valves. Force flow through a restriction, a partly closed valve, a porous plug, a capillary tube, and pressure plummets with no work extracted and no time for heat: the balance collapses to h2 = h1. Constant enthalpy sounds like nothing happened, but for a liquid refrigerant crossing into the two-phase dome, constant h at falling pressure forces partial flash evaporation, and the latent heat for that flash is stolen from the stream itself, chilling it drastically. This dumb, motionless device is what makes the cold in nearly every refrigerator on Earth, as the refrigeration lesson will compute in detail.

Heat exchangers. Two streams pass in thermal contact without mixing: car radiators, condensers, boiler economizers. No shaft work, negligible outside losses, so the balance is simply that the hot stream's enthalpy loss feeds the cold stream's gain: mdot(hot) x delta h(hot) = mdot(cold) x delta h(cold). Example with real plant flavor: exhaust steam condenses at 10 kg/s from h = 2400 kJ/kg to saturated liquid at 192 kJ/kg, shedding 10 x 2208 = 22,080 kW. Cooling water allowed to warm 10 K must flow at mdot = 22,080 / (4.18 x 10) = 528 kg/s, a small river, which is why power stations sit beside seas and rivers and why cooling towers loom the way they do.

Key idea: One balance, pruned five ways: nozzles convert delta h to velocity, turbines to shaft work out, compressors reverse it, throttles hold h constant, and heat exchangers match one stream's enthalpy loss to the other's gain.

Common misconceptions

• The energy of a flowing stream is its internal energy. Each kilogram also carries P v of flow work paid at the boundary; the honest total is enthalpy plus kinetic and potential energy, which is why open-system equations run on h, not u.
• Throttling wastes pressure, so it must waste energy. Enthalpy is conserved across a throttle; no energy is lost. What is lost is the opportunity to extract work, a distinction the second law will make precise.
• Kinetic energy always matters in flow devices. At 45 m/s it amounts to 1 kJ/kg against enthalpy changes of hundreds; it matters only where velocity is the point, as in nozzles and diffusers.
• Steady state means nothing is happening. Torrents of mass and megawatts of energy pour through; steady means the rates and the properties at each location hold constant, so nothing accumulates inside the boundary.

Recap

• Control volume analysis fixes the boundary in space; at steady state, mass and energy flows balance with no accumulation, and mdot = rho A V.
• Flow work P v rides along with every kilogram crossing a boundary, promoting enthalpy h = u + P v to the natural energy currency of open systems.
• Steady-flow energy equation: Qdot - Wdot = mdot [(h2 - h1) + (V2^2 - V1^2)/2 + g(z2 - z1)], with the velocity and height terms in J/kg and usually negligible.
• Nozzle: V2 = sqrt(2000 (h1 - h2)). Turbine: Wdot = mdot (h1 - h2). Compressor: w = h2 - h1. Pump: w = v delta P, tiny for liquids. Throttle: h2 = h1. Heat exchanger: enthalpy lost = enthalpy gained.
• Worked anchors: 100 kJ/kg through a nozzle gives 447 m/s; 800 kJ/kg at 25 kg/s gives 20 MW; condensing 10 kg/s of steam demands about 528 kg/s of cooling water for a 10 K rise.

Sources

1. Wikipedia. (n.d.). Control volume. Wikimedia Foundation. en.wikipedia.org
2. U.S. Department of Energy. (n.d.). How gas turbine power plants work. Office of Fossil Energy and Carbon Management. energy.gov
3. Engineering LibreTexts. (n.d.). Thermodynamics bookshelf. LibreTexts. eng.libretexts.org
4. MIT OpenCourseWare. (n.d.). Free online course materials. Massachusetts Institute of Technology. ocw.mit.edu

The big picture

The first law would permit some wonderful machines. A cup of coffee could grow hotter by borrowing energy from the cooler room, books balanced. An ocean liner could suck heat out of seawater and turn it entirely into propulsion, leaving a trail of slightly chilled ocean, energy rigorously conserved. No such machine has ever run. Nature enforces a second rule, independent of the first: processes have a direction, and energy conversions charge a toll. Coffee cools; it never uncoools. Work degrades into heat completely and gladly; heat converts back into work only partially and grudgingly. Today's lesson states that rule precisely, twice, and then extracts from it the most famous inequality in engineering: the ceiling, discovered by a 28-year-old French engineer in 1824, on the efficiency of every engine that will ever be built.

Sadi Carnot published his little book on the motive power of fire when steam engines were transforming Europe and nobody could say what limited them. His answer, reasoned about idealized engines decades before energy conservation itself was established, survives untouched: what limits an engine is not friction, not materials, not cleverness, but the temperatures it works between. That result governs the design of every power plant, jet engine, and refrigerator on Earth, and it will occupy the second half of this lesson. First we need the machines and the vocabulary.

Heat engines and thermal efficiency

A heat engine is any device that runs in a cycle, drawing heat from a hot source and delivering net work. Idealize the source as a thermal reservoir at TH, a body so large its temperature never budges (a furnace, effectively; combustion gases in practice), and idealize the environment as a cold reservoir at TL (a river, the atmosphere). Per cycle the engine takes in QH from the hot side, delivers net work W(net), and, crucially, rejects waste heat QL to the cold side. The first law across the cycle demands W(net) = QH - QL. The figure of merit is the thermal efficiency: eta = W(net)/QH = 1 - QL/QH, the fraction of purchased heat that becomes work. An engine absorbing 500 kJ and rejecting 300 kJ delivers 200 kJ of work at eta = 1 - 300/500 = 0.40, forty percent.

Could a cleverer engine reject nothing and hit 100 percent? The Kelvin-Planck statement of the second law says no: no device operating in a cycle can receive heat from a single reservoir and produce a net amount of work. Some heat must always be rejected to a colder body. The rejection is not an engineering blemish to be polished away; it is the toll nature charges for converting disorganized thermal energy into organized work. A hypothetical machine violating this statement, drinking from one reservoir and working forever, is a perpetual motion machine of the second kind: it would break no energy books, and it is impossible anyway.

Key idea: A heat engine converts QH into W(net) = QH - QL with eta = 1 - QL/QH, and Kelvin-Planck forbids QL = 0: every engine must reject heat to a colder reservoir.

Refrigerators, heat pumps, and the Clausius statement

Run the arrows backwards and you get the other great machine family. A refrigerator absorbs QL from a cold space, accepts work W from a compressor, and rejects QH = QL + W to the warm surroundings. A heat pump is the identical machine with the opposite sales pitch: its product is the QH delivered to a warm space, pumped up from the cold outdoors. Efficiency here is not bounded by one, so we use the honest name coefficient of performance: COP(R) = QL/W for refrigerators, COP(HP) = QH/W for heat pumps. Since QH = QL + W, the two are related by COP(HP) = COP(R) + 1, always. A refrigerator that removes 600 kJ from its cold box using 200 kJ of electricity has COP(R) = 3.0 and dumps 800 kJ into your kitchen. A COP above 1 is normal and violates nothing: the machine is not creating energy, merely moving it, paying the work toll to force heat uphill.

Uphill is the operative word. The Clausius statement of the second law: no device operating in a cycle can transfer heat from a cooler body to a hotter body with no other effect. Heat flows downhill unaided; moving it uphill demands paid work, which is why your refrigerator has a power cord and your freezer stops without one. The Kelvin-Planck and Clausius statements sound unrelated, but each implies the other: if you could convert heat wholly into work (violating Kelvin-Planck), you could feed that work to a heat pump and the combination would push heat cold-to-hot with no net work drawn, violating Clausius. Two phrasings, one law.

Key idea: Refrigerators and heat pumps move heat uphill at COP(R) = QL/W and COP(HP) = QL/W + 1, and Clausius forbids any cyclic device from doing it for free.

Reversibility: the idealization worth naming

To ask what the best possible machine looks like, we need the idea of a reversible process: one that can be undone, returning both system and surroundings exactly to their initial states, leaving no trace on the universe. Real processes never qualify, and it pays to memorize the reasons, the irreversibilities: friction, which scrubs work into heat you cannot fully recover; heat transfer across a finite temperature difference, which can never spontaneously run backwards; unresisted (free) expansion; and the mixing of different substances. A process free of all these, quasi-equilibrium, frictionless, with heat exchanged only across vanishing temperature differences, is reversible, and exists only as a limit approached, the way a real surface approaches frictionless. The limit is not idle philosophy: it defines the ceiling of performance, because every irreversibility present in a real machine subtracts from what the reversible version would deliver.

Carnot's theorem and the temperature ceiling

Carnot imagined the fully reversible engine cycle, two isothermal legs exchanging heat with the reservoirs across vanishing temperature differences and two adiabatic legs bridging them, and proved two propositions that survive as the Carnot principles: no engine between two given reservoirs can beat a reversible engine, and all reversible engines between the same two reservoirs share the same efficiency, regardless of construction or working fluid. That second principle is startling: it means the ceiling depends on nothing but the reservoir temperatures. Kelvin seized on it to define temperature itself: for reversible engines, the heats exchanged are proportional to absolute temperatures, QH/QL = TH/TL. Substituting into eta = 1 - QL/QH gives the crown jewel:

eta(Carnot) = 1 - TL/TH, with both temperatures in kelvins, no exceptions.

Work it: a plant receiving heat at 500 C (773 K) and rejecting to a river at 25 C (298 K) has eta(Carnot) = 1 - 298/773 = 0.614. Not 100 percent: 61.4 percent, and that is the unbeatable ideal. Real steam plants between roughly those temperatures deliver 35 to 45 percent, the gap paid to friction, finite temperature differences in boilers and condensers, and other irreversibilities. The formula also says where progress lives: raise TH (hence the metallurgy race toward hotter turbine blades) or lower TL (hence condensers running near vacuum). The reversed cycle gives the matching ceilings for heat movers: COP(R, Carnot) = TL/(TH - TL) and COP(HP, Carnot) = TH/(TH - TL). A freezer holding 268 K in a 303 K kitchen can at very best reach COP = 268/35 = 7.7; real machines manage 2 to 4. And notice the structure: the smaller the temperature lift, the higher the possible COP, which is why heat pumps shine in mild climates and strain in arctic ones.

These formulas make you a fraud auditor. An inventor claims an engine producing work at 80 percent efficiency between 600 K and 300 K. Ceiling: 1 - 300/600 = 50 percent. Verdict: impossible, no hearing required, the same verdict patent offices have delivered for a century. Any claimed engine efficiency above 1 - TL/TH, or claimed refrigerator COP above TL/(TH - TL), is a second-law violation in a press release.

Key idea: Between reservoirs at TH and TL, no engine can beat eta = 1 - TL/TH and no refrigerator can beat COP = TL/(TH - TL); the ceilings depend on temperatures alone and expose impossible claims instantly.

Why we do not simply build Carnot engines

If the reversible engine is best, why does anyone build anything else? Because reversibility is purchased with vanishing temperature differences and quasi-equilibrium gentleness, which means vanishing rates: a truly reversible engine delivers its work infinitely slowly, and a power plant that produces zero kilowatts perfectly efficiently sells no electricity. Real design is the art of spending irreversibility wisely: accepting finite temperature differences to get finite heat flows, tolerating friction to get compact machines, and clawing back losses where the price is right. The great cycles of the next module, Rankine, vapor-compression, Otto, Diesel, Brayton, are exactly such negotiated settlements between Carnot's ceiling and the market's demand for power now. Keep both numbers in view from here on: the Carnot limit tells you what the temperatures allow, and the cycle analysis tells you what your machine actually collects.

Common misconceptions

• Rejected heat means engineers were sloppy. Kelvin-Planck makes rejection mandatory for any cyclic engine; even the perfect reversible engine between 773 K and 298 K rejects 38.6 percent of its intake.
• A COP greater than 1 violates energy conservation. A heat pump moves energy rather than creating it: 1 kJ of work can usher several kJ of heat from outdoors to indoors, books balanced, second law satisfied.
• Efficiency limits are technological and will fall to better engineering. The Carnot bound is set by reservoir temperatures alone; better engineering approaches it but can never cross it. Progress means hotter TH or colder TL, not repealing the law.
• The Kelvin-Planck and Clausius statements are two different laws. Each implies the other; violating one lets you build a machine violating the second. They are one law in two dialects.

Recap

• Heat engines: eta = W(net)/QH = 1 - QL/QH; Kelvin-Planck forbids converting heat entirely to work in a cycle.
• Refrigerators and heat pumps: COP(R) = QL/W, COP(HP) = COP(R) + 1; Clausius forbids moving heat cold-to-hot with no work.
• Irreversibilities: friction, heat transfer across finite temperature differences, free expansion, mixing; reversible processes are the traceless ideal limit.
• Carnot: all reversible engines between TH and TL share eta = 1 - TL/TH; reversed, COP(R) = TL/(TH - TL). Kelvins mandatory.
• Real machines run well below the ceilings (steam plants 35-45 percent against a 61 percent Carnot; freezers COP 2-4 against 7.7), and claims above the ceiling are frauds by inspection.

Sources

1. OpenStax. (2016). Statements of the second law of thermodynamics. In University physics volume 2. Rice University. openstax.org
2. OpenStax. (2016). Heat engines. In University physics volume 2. Rice University. openstax.org
3. OpenStax. (2016). The Carnot cycle. In University physics volume 2. Rice University. openstax.org
4. Britannica. (n.d.). Second law of thermodynamics. In Encyclopaedia Britannica. britannica.com

The big picture

The second law, as you met it last lesson, is a pair of prohibitions: no perfect engines, no free uphill heat. Prohibitions are useful, but engineers want arithmetic, a quantity to compute that tells you whether a proposed process can happen and how badly a real one falls short of ideal. In 1865 Rudolf Clausius produced exactly that. Buried in the mathematics of reversible cycles he found a new property, as real as pressure or energy, that tracks the one-way character of nature, and he named it entropy, from the Greek for transformation, deliberately built to rhyme with energy. His summary of all thermodynamics fits in two lines: the energy of the universe is constant; the entropy of the universe tends toward a maximum.

Entropy has acquired a fearsome reputation it does not deserve. In this course it is a working number: you will look it up in the same steam tables as enthalpy, compute its changes with three short formulas, and use it for two eminently practical purposes. First, as a compass: a process whose total entropy change would be negative cannot happen, full stop. Second, as an idealization: the best possible turbines and compressors run at constant entropy, which turns out to be exactly the adiabatic ideal from lesson seven, and next lesson grades real machines against that standard. Today we define the property, learn to compute it, and see, in two lines, why it sets the Carnot ceiling.

Defining the property

Here is the definition, and then the sense of it. For any infinitesimal step of a reversible process, the entropy change of a system is dS = delta Q / T: the heat transferred, divided by the absolute temperature at which it crosses the boundary. Entropy carries units of kJ/K, or per kilogram, s in kJ/kg K. Clausius proved that this quantity is a property: integrate delta Q / T along any reversible path between two states and you get the same answer, so a system in a given state has an entropy, as surely as it has a pressure. That is why steam tables carry s columns beside h, why quality arithmetic works unchanged (s = sf + x sfg inside the dome), and why you may speak of the entropy of a kilogram of steam without specifying how it got there.

The definition rations entropy by temperature, and the rationing is the physics. A kilojoule delivered at low temperature carries a lot of entropy; the same kilojoule at high temperature carries little. Feel it in the formula: 100 kJ leaving a 500 K reservoir removes 100/500 = 0.200 kJ/K, and the same 100 kJ arriving in a 300 K reservoir deposits 100/300 = 0.333 kJ/K. The transfer created 0.133 kJ/K of entropy out of nothing. Run the movie backwards, 100 kJ flowing cold to hot, and the universe's entropy would fall by 0.133 kJ/K. Hold that thought for one paragraph.

Key idea: Entropy is a property whose change is reversible heat over absolute temperature, dS = delta Q / T, so a kilojoule carries much entropy at low temperature and little at high.

The increase principle: the second law as arithmetic

Now the law itself, in its most powerful form. Real processes are irreversible, and irreversibility, friction, finite-difference heat transfer, free expansion, mixing, generates entropy. The books read: delta S(total) = delta S(system) + delta S(surroundings) = S(gen), where the entropy generation S(gen) is positive for every real process, zero for the reversible limit, and negative never. For an isolated system, which has no surroundings to trade with, this is the famous increase of entropy principle: the entropy of an isolated system never decreases. Unlike energy, entropy is not conserved. It is manufactured by every real event and destroyed by none.

This turns impossibility proofs into two-line arithmetic. Should heat flow from hot to cold? Our 500 K to 300 K transfer generated +0.133 kJ/K: allowed, and observed. The reverse would generate -0.133 kJ/K: forbidden, and never observed. The Clausius statement of last lesson has become a sign check. Notice also what the example teaches about waste: nothing was lost as energy, all 100 kJ arrived, yet something real was squandered, because energy at 300 K can do less for you than energy at 500 K. Entropy generation is the precise measure of that squandering, and a designer's scorecard: every kelvin of unnecessary temperature difference in a boiler, every throttle that could have been a turbine, shows up as S(gen) you paid for and got nothing from.

Key idea: Every real process generates entropy, delta S(total) = S(gen) greater than 0, so the entropy of the universe only climbs; a proposed process with negative total entropy change is impossible.

Computing entropy changes: three formulas

You need only three recipes for this course, all derived from dS = delta Q / T applied along reversible paths.

Solids and liquids (incompressible, specific heat c): warming from T1 to T2 gives delta S = m c ln(T2/T1). Two kilograms of water heated from 300 K to 350 K gains delta S = 2 x 4.18 x ln(350/300) = 8.36 x 0.1542 = 1.29 kJ/K. The logarithm rewards you for adding heat cold and penalizes adding it hot, exactly as the definition promised. Note the mandatory kelvins inside the ratio.

Phase change at constant temperature: boiling happens at fixed Tsat, so the integral collapses to delta S = Q/T = m hfg / Tsat. One kilogram of water boiling at 373.15 K gains 2257/373.15 = 6.05 kJ/kg K, and there is nothing to memorize here: look at any saturation table and you will find sfg at 100 C listed as exactly that number. The tables were built from this arithmetic.

Ideal gases, where temperature and pressure both matter: delta s = cp ln(T2/T1) - R ln(P2/P1). Heating raises entropy; compression to higher pressure lowers it (at fixed temperature, molecules confined to less room). Test it on lesson seven's adiabatic compression of air, 100 kPa and 300 K to 800 kPa and 543.4 K: delta s = 1.005 x ln(543.4/300) - 0.287 x ln(8) = 1.005 x 0.5942 - 0.287 x 2.0794 = 0.597 - 0.597 = 0.000 kJ/kg K. Zero, to three decimals. That quasi-equilibrium adiabatic process held entropy constant, which brings us to the most useful special case in engineering.

Key idea: delta S = m c ln(T2/T1) for incompressible substances, m hfg/Tsat for phase change, and delta s = cp ln(T2/T1) - R ln(P2/P1) for ideal gases.

Isentropic processes and the T-s diagram

A process that is both adiabatic (no heat, so no entropy carried in or out) and reversible (no entropy generated inside) keeps entropy constant: isentropic. This is the name lesson seven promised for the ideal compression and expansion formulas, and the calculation above is the proof of the connection: P v^k = constant is the ideal gas's isentropic path. Isentropic is the gold standard for turbines, compressors, pumps, and nozzles, machines built to move energy as work while touching heat as little as possible. Real ones generate some entropy through friction and turbulence; how close they come to the standard is next lesson's subject.

Entropy also earns its own graph. Plot temperature against specific entropy and you get the T-s diagram, the working canvas of cycle analysis from here to the end of the course. Its charms: reversible heat transfer appears as the area under the path (since delta Q = T dS), isentropic processes are clean vertical lines, and the steam dome drawn on T-s axes looks much like it did on T-v. On this canvas the Carnot cycle, two isothermals and two isentropics, is literally a rectangle: heat in is the area under the top edge, heat out the area under the bottom, net work the rectangle's area. Efficiency becomes geometry you can see: a taller rectangle (hotter TH, colder TL) wastes proportionally less of its area below.

And now the promised two-line derivation of Carnot's ceiling. An engine drawing QH at TH ingests entropy QH/TH. A cycle returns the working fluid to its start, so all of it must be expelled, and the only exit is heat rejection at TL: QL/TL must be at least QH/TH (equality only if nothing extra was generated inside). Therefore QL is at least QH x TL/TH, and eta = 1 - QL/QH can never exceed 1 - TL/TH. The Carnot limit is nothing but entropy bookkeeping: engines reject heat because they must reject entropy, and cold rejection is expensive precisely when TL/TH is large.

Key idea: Adiabatic plus reversible equals isentropic (vertical on the T-s diagram), and requiring an engine to expel its ingested entropy at TL is exactly what caps efficiency at 1 - TL/TH.

What entropy means, and what it does not

The molecular view, worked out by Ludwig Boltzmann, says entropy measures the number of microscopic arrangements consistent with what you observe macroscopically: S = k ln W, with k the Boltzmann constant and W the count of microstates. Heat spreading from hot to cold, gas filling a vacuum, cream swirling into coffee: each moves matter and energy toward conditions realizable in overwhelmingly more ways, and the increase principle becomes probability at astronomical scale. The popular gloss disorder is serviceable but slippery; energy dispersal is closer. And two cautions will keep you out of common traps. Entropy is not energy and is not lost energy: energy is strictly conserved while entropy grows, and what degrades is energy's usefulness, its temperature pedigree. And local decreases are entirely legal at a price: your freezer lowers the entropy of water into ice all day, while the coils behind it dump more entropy into the kitchen than the ice ever lost. Order here, paid for with greater disorder there: that is also, incidentally, how living things manage it.

Common misconceptions

• Entropy is conserved like energy. No: entropy is generated by every real process and destroyed by none. Conservation belongs to energy; monotonic increase belongs to entropy.
• Entropy always increases, everywhere. The total for an isolated system (or the universe) never falls; any subsystem's entropy may drop freely if more is generated elsewhere, as in every freezer.
• Entropy is lost energy in disguise. Entropy has different units (kJ/K) and different books. In irreversible heat transfer, all the energy arrives; what is lost is its capacity to do work, and S(gen) is the meter that measures the loss.
• Isentropic just means adiabatic. It means adiabatic and reversible. A real insulated compressor is adiabatic yet raises entropy through friction; only the idealized frictionless version is isentropic.

Recap

• Entropy is a property defined through reversible heat transfer, dS = delta Q / T, tabulated like enthalpy, with s = sf + x sfg inside the dome.
• All real processes generate entropy: delta S(total) = S(gen) greater than 0, zero only in the reversible limit; negative total changes mark impossible processes.
• Recipes: m c ln(T2/T1) for solids and liquids, m hfg/Tsat for phase change, cp ln(T2/T1) - R ln(P2/P1) per kilogram for ideal gases.
• Isentropic (adiabatic + reversible) processes are vertical lines on the T-s diagram, where reversible heat is area under the path and Carnot is a rectangle.
• Entropy bookkeeping yields the Carnot bound in two lines and quantifies every design loss as S(gen).
• Microscopically, S = k ln W: increase is probability at scale; local decreases are paid for elsewhere.

Sources

1. OpenStax. (2016). Entropy. In University physics volume 2. Rice University. openstax.org
2. OpenStax. (2016). Entropy on a microscopic scale. In University physics volume 2. Rice University. openstax.org
3. Britannica. (n.d.). Entropy. In Encyclopaedia Britannica. britannica.com
4. National Institute of Standards and Technology. (n.d.). SI redefinition: Kelvin and the Boltzmann constant. NIST. nist.gov

The big picture

You now hold both halves of a grading system. The steady-flow energy equation says what a turbine or compressor does: moves enthalpy in or out as shaft work. The entropy lesson says what the best possible version looks like: isentropic, a vertical drop on the T-s diagram, no entropy generated. Real machines are neither frauds nor ideals; they are somewhere between, and industry describes where with a single number per machine, the isentropic efficiency: the ratio of actual performance to what the isentropic machine would have done between the same inlet state and the same exit pressure. Turbine builders quote 80 to 90 percent; compressor builders 75 to 85; nozzle designers 90 to 99. Today you learn to use those numbers, and the skill is the workhorse of every cycle analysis to come: nearly every real Rankine or Brayton problem is an ideal calculation followed by one multiplication or division.

One orientation before the formulas. The comparison is always run between the actual inlet state and the actual exit pressure: the isentropic twin starts where the real machine starts and exhausts where it exhausts, but travels at constant entropy. What differs is the exit temperature or quality, and therefore the work. Keep that picture, two paths diverging from one point down to one pressure, and the algebra never confuses you.

Turbines: the ratio that flatters honestly

A turbine exists to produce work, so its grade is eta(T) = w(actual) / w(isentropic): what you got over what was theoretically available. The actual machine, with its blade friction and turbulence, produces less, so the ratio sits below one, as an efficiency should.

Worked example, gas side. Hot air enters a turbine at 1000 K and 800 kPa and exhausts at 100 kPa; take eta(T) = 0.85 and cp = 1.005, k = 1.4. Step one, the isentropic twin: T2s = T1 x (P2/P1)^((k-1)/k) = 1000 x (100/800)^0.2857 = 1000 / 1.811 = 552 K. Step two, the ideal work: w(s) = cp (T1 - T2s) = 1.005 x 448 = 450 kJ/kg. Step three, the actual work: w(a) = 0.85 x 450 = 383 kJ/kg. Step four, the actual exit: the energy not converted to work stays in the stream, so T2a = T1 - w(a)/cp = 1000 - 381 = 619 K, warmer than the ideal 552 K. The exhaust of a real turbine is always hotter than its isentropic twin's; the shortfall did not vanish (first law forbids), it merely left as heat still in the gas instead of work on the shaft.

Steam works identically with tables in place of cp formulas. Steam enters at 3 MPa and 400 C, where the tables give h1 = 3230.9 kJ/kg and s1 = 6.9212 kJ/kg K, and exhausts at 50 kPa, where sf = 1.0910, sfg = 6.5029, hf = 340.5, hfg = 2305.4. The isentropic twin lands in the dome: x2s = (6.9212 - 1.0910)/6.5029 = 0.897, so h2s = 340.5 + 0.897 x 2305.4 = 2408 kJ/kg and w(s) = 3230.9 - 2408 = 823 kJ/kg. At eta(T) = 0.85, the real machine delivers w(a) = 700 kJ/kg and exhausts at h2a = 3230.9 - 700 = 2531 kJ/kg, which back-solves to an actual exit quality of 0.950: drier than the ideal 0.897. The pattern generalizes: real turbine exhaust is hotter, or in the dome less wet, than ideal. Working turbines actually appreciate that particular side effect, since liquid droplets erode blades; we return to it in the Rankine lesson.

Key idea: eta(T) = w(a)/w(s): fix the inlet, drop entropy vertically to the exit pressure for the ideal, then multiply by the efficiency, and put the unconverted energy back in the exhaust stream.

Compressors and pumps: the ratio flips

A compressor consumes work, so grading it by actual over ideal would produce numbers above one and reward bad machines. The definition flips: eta(C) = w(isentropic) / w(actual): the least work the job could possibly take, over what your machine actually drew. Again the ratio sits below one, and again the physics is that irreversibility hurts you in whichever direction you are working: producers produce less than ideal, consumers consume more.

Worked example, continuing the course's running numbers. Air enters a compressor at 100 kPa and 300 K and leaves at 800 kPa; take eta(C) = 0.80. The isentropic twin is lesson seven's calculation: T2s = 300 x 8^0.2857 = 543 K, so w(s) = cp (T2s - T1) = 1.005 x 243.4 = 245 kJ/kg. The actual draw is w(a) = 245 / 0.80 = 306 kJ/kg, and the actual exit temperature is T2a = 300 + 306/1.005 = 604 K, hotter than the ideal 543 K. Now close the loop with lesson ten's formula and audit the entropy: delta s = cp ln(604/300) - R ln(8) = 1.005 x 0.700 - 0.287 x 2.079 = 0.704 - 0.597 = +0.107 kJ/kg K. Positive, as the second law demands of any real adiabatic machine: the isentropic twin scored exactly zero, and those 0.107 units are the generation you paid 61 extra kilojoules per kilogram to produce.

Pumps are compressors for liquids, and liquids hand you a gift: specific volume a thousandfold smaller. The reversible steady-flow work to raise a fluid's pressure is w = v x delta P for an incompressible fluid, so pumping water from 10 kPa to 6000 kPa costs only w(s) = 0.00101 x 5990 = 6.05 kJ/kg. At a pump efficiency of 0.85 the actual is about 7.1 kJ/kg. Compare: compressing a gas through a mere 8-to-1 pressure ratio cost 306 kJ/kg above; the pump raised pressure 600-fold for a fortieth of the price. Squeezing liquid is cheap, squeezing gas is brutal, and that asymmetry is the structural secret of the steam power cycle you will analyze next lesson: it compresses its working fluid while liquid and expands it while gas, keeping nearly everything the turbine makes.

Key idea: eta(C) = w(s)/w(a), so actual input is ideal divided by efficiency; pump work is only v delta P because liquid specific volume is tiny, which is why cycles compress liquids when they can.

Nozzles, and the method in five steps

Nozzles convert enthalpy drop to kinetic energy, so their grade is the kinetic energy ratio: eta(N) = V2a^2 / V2s^2, actual exit kinetic energy over isentropic. Well-shaped nozzles are the honor students of this family, commonly 0.90 to 0.99, because a smoothly accelerating flow generates little entropy. A nozzle whose ideal calculation promises 447 m/s (lesson eight's steam example) and whose efficiency is 0.94 actually delivers V2a = 447 x sqrt(0.94) = 433 m/s.

Whatever the device, the method is the same five steps, worth writing out once and using forever. One: fix the inlet state and find its h and s (tables) or T (ideal gas). Two: build the isentropic twin, same entropy, actual exit pressure, and find its exit enthalpy or temperature. Three: compute the ideal work (or kinetic energy). Four: apply the efficiency the right way around, multiply for producers, divide for consumers. Five: locate the actual exit state by putting the difference back into the stream with the energy balance. Steps two and five are where sign and direction errors breed; the T-s sketch, ideal vertical line beside a slightly right-leaning actual path, catches them before arithmetic does.

Key idea: eta(N) compares exit kinetic energies, and every device analysis is the same five steps: inlet state, isentropic twin, ideal work, apply efficiency, back out the actual exit.

Reading the grade: what the numbers mean in practice

A closing word on what these percentages are worth in money and design. A utility-scale steam turbine converting 700 instead of 823 kJ/kg at 400 kg/s is forgoing about 49 MW, roughly a small city's demand, which is why manufacturers fight for single points of isentropic efficiency and why blade aerodynamics is a career. Compressor efficiency compounds differently: in a gas turbine, the compressor eats 40 to 60 percent of the turbine's gross output (the back-work ratio you will meet in the Brayton lesson), so a point of compressor efficiency is worth several points of plant output. And the efficiencies are not constants of nature: they sag off the design point, which is why machines are matched to duty and why part-load operation costs more than proportionally. When a datasheet says eta(T) = 0.87, it is quoting the design point of a machine whose real grade varies with every valve setting; treat it as this course does, as the honest one-number summary of a complicated, hard-won compromise.

Common misconceptions

• Isentropic efficiency compares actual and ideal machines running to the same exit temperature. The twin shares the inlet state and the exit pressure; its exit temperature (or quality) is what differs, and finding it is step two of every problem.
• For compressors, efficiency is actual over ideal. Flipped: eta(C) = ideal over actual, so that worse machines score lower. Producers multiply by eta; consumers divide.
• An insulated real compressor is isentropic since no heat crosses. It is adiabatic but irreversible: friction and turbulence generate entropy inside (our example generated 0.107 kJ/kg K), so entropy rises without any heat transfer.
• The turbine's lost work disappears. Energy is conserved: the shortfall remains in the exhaust as extra enthalpy, which is why real exhaust runs hotter (or drier) than the isentropic twin's.

Recap

• Isentropic efficiency grades a real adiabatic device against an isentropic twin sharing its inlet state and exit pressure.
• Turbines: eta(T) = w(a)/w(s); air example 450 ideal, 383 actual kJ/kg at 0.85; steam example 823 ideal, 700 actual at 0.85, exit quality rising from 0.897 to 0.950.
• Compressors: eta(C) = w(s)/w(a); air at 8-to-1 costs 245 ideal, 306 actual kJ/kg at 0.80, generating +0.107 kJ/kg K of entropy.
• Pumps obey w = v delta P: 6 kJ/kg to lift water 10 to 6000 kPa, a fortieth of gas compression work, the Rankine cycle's structural advantage.
• Nozzles: eta(N) = V2a^2/V2s^2, typically 0.90 to 0.99.
• Five steps always: inlet state, isentropic twin, ideal work, apply efficiency correctly, back out the actual exit.

Sources

1. Wikipedia. (n.d.). Isentropic process. Wikimedia Foundation. en.wikipedia.org
2. OpenStax. (2016). Adiabatic processes for an ideal gas. In University physics volume 2. Rice University. openstax.org
3. Engineering LibreTexts. (n.d.). Thermodynamics bookshelf. LibreTexts. eng.libretexts.org
4. National Institute of Standards and Technology. (n.d.). Thermophysical properties of fluid systems. NIST Chemistry WebBook. webbook.nist.gov

The big picture

Follow your wall outlet upstream and, more often than not, you arrive at the same machine: a loop of water being boiled, blasted through a turbine, condensed, and pumped around again. Coal plants, nuclear plants, many gas plants (as the bottom half of combined cycles), solar-thermal towers, and geothermal stations all run versions of this loop. It is the Rankine cycle, named for the Scottish engineer William Rankine who systematized steam power analysis in the 1850s, and it generates the majority of the world's electricity to this day. Today you will analyze it completely: four components, four processes, real steam table numbers end to end, and a thermal efficiency computed honestly. This is the lesson where the whole course assembles: property tables (module two), enthalpy balances on flow devices (module three), and isentropic ideals (module four) all working one problem.

Why water and steam at all? Three reasons worth stating before the machinery. Water is cheap, abundant, and chemically tame. Its enormous latent heat lets each kilogram carry huge energy through the loop, keeping machines compact. And, decisively, it condenses at the cold end, so the cycle compresses a liquid, and lesson eleven taught you what that is worth: pump work of a few kJ/kg against turbine output of over a thousand. Nearly everything the turbine makes goes out the wires.

The loop, component by component

Trace the ideal cycle with the conventional state numbers. State 1, pump inlet: saturated liquid at condenser pressure, cool and dense. Process 1 to 2, the pump: isentropic compression of liquid up to boiler pressure; tiny work in. Process 2 to 3, the boiler: constant-pressure heat addition, warming the liquid to saturation, boiling it across the dome, and superheating the vapor; this is where the fuel is spent. State 3, turbine inlet: high-pressure superheated steam, the cycle's energetic summit. Process 3 to 4, the turbine: isentropic expansion to condenser pressure, the great enthalpy drop that spins the generator. Process 4 to 1, the condenser: constant-pressure heat rejection to cooling water, collapsing the exhaust back to saturated liquid. On the T-s diagram the ideal cycle is two verticals (pump and turbine) joined by two constant-pressure paths, wrapped around the steam dome; the enclosed area is the net work per kilogram.

Key idea: Pump, boiler, turbine, condenser: compress liquid, add heat at high pressure, expand vapor through the turbine, reject heat at low pressure, repeat forever.

A full plant, computed

Now the numbers, for an ideal Rankine cycle with boiler at 6 MPa, turbine inlet at 500 C, and condenser at 10 kPa. All property values below come from the steam tables; pull them yourself from any table set or the NIST WebBook and follow along.

State 1 (saturated liquid, 10 kPa): h1 = 191.8 kJ/kg, v1 = 0.00101 m^3/kg. The condenser pressure of 10 kPa, a rough vacuum with Tsat = 45.8 C, is typical: the plant rejects heat barely above the cooling water's temperature. Pump, 1 to 2: w(pump) = v1 x (P2 - P1) = 0.00101 x (6000 - 10) = 6.05 kJ/kg, so h2 = 191.8 + 6.05 = 197.9 kJ/kg. Boiler, 2 to 3: at 6 MPa and 500 C the tables give h3 = 3422.2 kJ/kg and s3 = 6.8803 kJ/kg K, so q(in) = h3 - h2 = 3422.2 - 197.9 = 3224.3 kJ/kg. Turbine, 3 to 4: the isentropic expansion carries s4 = s3 = 6.8803 down to 10 kPa, where sf = 0.6493 and sfg = 7.5009. Quality: x4 = (6.8803 - 0.6493)/7.5009 = 0.831. Enthalpy: h4 = 191.8 + 0.831 x 2392.8 = 2179.5 kJ/kg. Turbine work: w(turb) = 3422.2 - 2179.5 = 1242.7 kJ/kg. Condenser, 4 to 1: q(out) = 2179.5 - 191.8 = 1987.7 kJ/kg, carried off by cooling water.

Now the verdicts. Net work: w(net) = 1242.7 - 6.05 = 1236.6 kJ/kg; check it against q(in) - q(out) = 3224.3 - 1987.7 = 1236.6, and the books balance. Thermal efficiency: eta = 1236.6 / 3224.3 = 0.384, thirty-eight percent, squarely in the range of real subcritical steam plants. The back-work ratio, pump work over turbine work, is 6.05/1242.7 = 0.005: half of one percent, the liquid-compression dividend in person. And scale: a 500 MW plant needs mdot = 500,000 kW / 1236.6 kJ/kg = 404 kg/s of steam, roughly a bathtub of water every second, boiled, blasted, condensed, and boiled again around the clock. Meanwhile 404 x 1987.7 = 803 MW of heat pours into the cooling water: the plant makes 500 MW of electricity and 800 MW of warm river.

Key idea: The worked cycle delivers w(net) = 1236.6 kJ/kg from q(in) = 3224.3 kJ/kg for eta = 38.4 percent, with a back-work ratio of half a percent and two units of heat rejected for every unit of electricity.

Reading the result: the Carnot gap and the moisture problem

Grade this plant against lesson nine. Its hottest temperature is 773 K and its coldest 319 K, so the Carnot ceiling is 1 - 319/773 = 58.7 percent, and even our fully idealized cycle scored 38.4. Why the gap, with no friction anywhere? Because Rankine heat addition is not isothermal: much of q(in) enters while the liquid warms from 46 C toward 276 C, far below the peak temperature, dragging the average temperature of heat addition well under 773 K. The cycle's improvements are all campaigns to raise that average or lower the rejection temperature. Superheat more: hotter turbine inlets raise the average directly; the frontier is blade metallurgy, with modern ultra-supercritical plants running 600 C class steam and topping 45 percent. Raise boiler pressure: boiling happens hotter, again lifting the average. Lower condenser pressure: every kPa closer to vacuum drops the rejection temperature; the limit is the cooling water nature provides, which is why plants sit beside rivers, seas, and cooling towers.

One improvement needs its own paragraph because our own numbers demand it. Look at x4 = 0.831: the turbine exhaust is 17 percent liquid by mass. Water droplets moving at steam speeds sandblast turbine blades, and practice keeps exhaust quality above roughly 0.88 to 0.90. Raising boiler pressure alone makes this worse (the expansion line slides deeper into the dome). The standard cure is reheat: expand the steam partway in a high-pressure turbine, pipe it back to the boiler for a second heating to around the original temperature, then finish the expansion in a low-pressure turbine. Reheat dries the exhaust, adds a few points of efficiency, and is fitted to virtually every large plant. Its sibling regeneration bleeds a little steam from the turbine to preheat boiler feedwater in feedwater heaters, raising the average temperature of external heat addition; large plants chain five to eight of them. With reheat, regeneration, and modern steam conditions, the humble loop we computed reaches the mid-40s in percent efficiency, about as far as pure steam has ever been pushed.

Key idea: Rankine efficiency trails Carnot because heat enters at a low average temperature; superheat, higher boiler pressure, lower condenser pressure, reheat, and regeneration all attack that average, while reheat also cures the wet-exhaust problem our x4 = 0.831 exemplifies.

Where you meet this cycle

The Rankine cycle wears many fuels. A coal or biomass plant burns fuel under the boiler; a nuclear plant replaces the flame with a reactor core (and, running cooler for safety, accepts about 33 percent efficiency); a solar-thermal plant focuses sunlight on the boiler; a geothermal plant taps hot rock, often boiling a low-temperature organic fluid instead of water in what is called an organic Rankine cycle; a combined-cycle gas plant uses a jet-engine exhaust as its heat source, a marriage the final module will celebrate properly. The analysis you just performed transfers to all of them unchanged: find the four states, difference the enthalpies, mind the quality at the turbine exit. When the news mentions a 1000 MW plant, you can now unpack the sentence: about 404 kg/s of steam per 500 MW, a condenser pulling vacuum against the local river, and an efficiency whose every point was fought for against the second law.

Common misconceptions

• The boiler makes the power. The boiler spends the fuel; the turbine extracts the work. Power is made where enthalpy drops, and the drop is engineered by the pressure difference the pump and condenser maintain.
• Cooling towers emit smoke. The white plume is condensing water vapor from the heat-rejection side, the visible face of q(out); the second law obliges every heat engine to have one, in some form.
• Pump work is negligible so it could be omitted. It is small (0.5 percent here) but structural: without the pump maintaining boiler pressure there is no cycle. Neglecting its magnitude in an estimate is fine; forgetting its role is fatal.
• Higher boiler pressure is pure gain. Alone, it wets the exhaust and endangers the blades; it ships with reheat precisely to fix what it breaks.

Recap

• Rankine cycle: pump (isentropic, liquid), boiler (constant-pressure heat in), turbine (isentropic expansion), condenser (constant-pressure heat out).
• Worked plant (6 MPa, 500 C, 10 kPa): w(pump) = 6.05, q(in) = 3224.3, w(turb) = 1242.7, q(out) = 1987.7, w(net) = 1236.6 kJ/kg, eta = 38.4 percent, x4 = 0.831.
• Back-work ratio 0.5 percent, because the cycle compresses liquid; a 500 MW plant circulates about 404 kg/s of steam and rejects about 800 MW of heat.
• Efficiency rises with hotter average heat addition (superheat, pressure, regeneration) and colder rejection (condenser vacuum); reheat dries the exhaust and protects blades.
• The same loop, differently fueled, runs coal, nuclear, solar-thermal, geothermal, and the steam half of combined-cycle plants.

Sources

1. Wikipedia. (n.d.). Rankine cycle. Wikimedia Foundation. en.wikipedia.org
2. U.S. Energy Information Administration. (n.d.). How electricity is generated. EIA Energy Explained. eia.gov
3. Britannica. (n.d.). Steam turbine. In Encyclopaedia Britannica. britannica.com
4. National Institute of Standards and Technology. (n.d.). Thermophysical properties of fluid systems. NIST Chemistry WebBook. webbook.nist.gov

The big picture

Open your refrigerator and consider what is being accomplished: heat is flowing, steadily and on purpose, from a cold box to a warmer kitchen, precisely the direction Clausius says it will never go on its own. The machine that forces the issue is the vapor-compression cycle, and it may be the most consequential loop of pipe on Earth. It preserves the food supply, makes vaccines and much of modern medicine deliverable, cools the data centers this course is served from, and, as air conditioning, made entire climates habitable for billions. Running backwards as a heat pump, it is also the leading technology for low-carbon heating. One cycle, four components, and you already own every tool needed to analyze it: throttles and heat exchangers from module three, isentropic compression from module four, and property tables read exactly as you read steam's.

Conceptually the machine is a Rankine cycle thrown into reverse: work goes in instead of out, and heat is lifted from cold to hot instead of falling hot to cold. But one component differs, and the difference teaches a good lesson in engineering economics, so watch for the missing turbine as we walk the loop.

The loop, and the trick at its heart

The trick is one you already know from lesson five: a liquid's boiling temperature is set by its pressure. Choose a fluid, the refrigerant, whose pressure-temperature curve lets it boil below the cold box's temperature at one convenient pressure and condense above the kitchen's temperature at another, and the second law will do the local heat transfers for you, downhill both times. The cycle is just plumbing that holds the fluid at the right pressure in the right room.

Walk it with the standard numbering. State 1 to 2, the compressor: cool refrigerant vapor from the cold side is compressed, ideally isentropically, to the high pressure; its temperature jumps well above the surroundings. This is where the electricity is spent. 2 to 3, the condenser: the hot, high-pressure vapor flows through the coil on the machine's warm side (the grille behind your refrigerator), rejects heat to the room, and condenses to liquid: heat flowing downhill, hot refrigerant to cooler kitchen. 3 to 4, the expansion valve: the high-pressure liquid squeezes through a throttle, and lesson eight told you what happens: enthalpy is conserved while pressure collapses, part of the liquid flashes to vapor, and the latent heat of that flash chills the remaining mixture far below the cold box's temperature. 4 to 1, the evaporator: the frigid two-phase mixture winds through the coil inside the cold space, boiling as it absorbs heat from the food: downhill again, warmish food to colder refrigerant. The vapor returns to the compressor and the loop closes. Cold is never created; heat is merely given two easy downhill steps and one paid uphill ride.

Now, the missing turbine. In Rankine, expansion from high to low pressure happened through a turbine that harvested work. Here the same pressure drop happens through a dumb valve that harvests nothing, and the reason is a cost-benefit verdict: the recoverable work from expanding a liquid is tiny (v delta P again, a few kJ/kg), the fluid is two-phase and hard on machinery, and a capillary tube costs a dollar. Engineers spend irreversibility where reversibility does not pay. It is a deliberate, priced-in loss, and you will see it in the COP shortfall below.

Key idea: The refrigerant boils cold at low pressure (absorbing QL) and condenses warm at high pressure (rejecting QH); the compressor pays the uphill toll, and a throttle, chosen over a turbine on economics, provides the pressure drop.

The cycle computed: a refrigerator on paper

Our working fluid is R-134a, the standard automotive and domestic refrigerant of recent decades, with properties from its tables (or the NIST WebBook). Let the evaporator run at 140 kPa, where R-134a boils at -18.8 C, cold enough to serve a freezer compartment, and the condenser at 800 kPa, where it condenses at 31.3 C, warm enough to reject heat to a summer kitchen.

State 1 (saturated vapor, 140 kPa): h1 = 239.2 kJ/kg. Compressor, 1 to 2 (isentropic to 800 kPa): the superheat tables give h2 = 275.4 kJ/kg, so w(in) = 275.4 - 239.2 = 36.2 kJ/kg. State 3 (saturated liquid, 800 kPa): h3 = 95.5 kJ/kg, so the condenser rejected q(H) = h2 - h3 = 275.4 - 95.5 = 179.9 kJ/kg. Throttle, 3 to 4: h4 = h3 = 95.5 kJ/kg, arriving in the evaporator as a cold mixture at -18.8 C. Evaporator, 4 to 1: the refrigeration effect is q(L) = h1 - h4 = 239.2 - 95.5 = 143.7 kJ/kg. Audit the books: q(L) + w(in) = 143.7 + 36.2 = 179.9 = q(H). Every joule pulled from the food plus every joule of compressor work exits through the grille, which is why the back of a refrigerator is warm and why you cannot cool a kitchen by leaving the refrigerator door open: the machine would dump back everything it removes, plus its own work.

The verdict: COP(R) = q(L)/w(in) = 143.7/36.2 = 3.97. Each kilojoule of electricity moves nearly four kilojoules of heat out of the cold box. For scale, a mid-size unit providing 5 kW of cooling (about 1.4 tons of refrigeration, the industry's odd unit: one ton = 3.517 kW, the melting rate of a ton of ice per day) circulates mdot = 5/143.7 = 0.035 kg/s of refrigerant and draws about 0.035 x 36.2 = 1.26 kW at the plug. Grade it against Carnot between 254.4 K and 304.5 K: COP(max) = 254.4/(304.5 - 254.4) = 5.08. Our ideal cycle scores 3.97, and the gap is structural, mostly the throttle's generated entropy plus the superheated spike of hot vapor entering the condenser; real machines with imperfect compressors land lower still, typically 2 to 4.

Key idea: For the worked R-134a cycle, q(L) = 143.7, w(in) = 36.2, q(H) = 179.9 kJ/kg and COP = 3.97 against a Carnot ceiling of 5.08, the shortfall being the priced-in cost of the throttle and superheat.

The same machine, resold: heat pumps

Now the accounting flip with outsized consequences. Aim the condenser into the house and the evaporator at the winter outdoors, and the identical hardware becomes a heat pump whose product is q(H). Its score: COP(HP) = q(H)/w(in) = 179.9/36.2 = 4.97, which is COP(R) + 1 exactly as lesson nine promised, the machine collecting credit for its own work as well as the heat it moves. Read that number the way a utility bill does: one kilowatt-hour of electricity delivers nearly five kilowatt-hours of heat indoors, four of them mined from the cold outside air. An electric resistance heater, by definition, delivers exactly one. This factor-of-several advantage is why heat pumps are central to decarbonizing buildings, and a reversing valve that swaps the two coils' roles gives you the familiar year-round machine: air conditioner in July, heater in January.

The physics also forecasts the fine print. COP falls as the temperature lift grows, so a heat pump's advantage shrinks on the coldest days (and frost on the outdoor coil forces defrost cycles); modern cold-climate units mitigate this with variable-speed compressors and improved refrigerants, holding useful COPs well below freezing. Ground-source machines buy a steadier, milder source temperature by burying the outdoor coil. Every one of these design choices is the Carnot formula TL/(TH - TL) negotiating with a climate.

Key idea: The same cycle heating instead of cooling scores COP(HP) = COP(R) + 1 = 4.97: one unit of electricity delivers about five of heat, an advantage that shrinks as the outdoor-to-indoor temperature lift grows.

A word on the working fluids

Refrigerant choice is a story of trade-offs revised by consequences. Early machines used ammonia, sulfur dioxide, and ethers: thermodynamically excellent, variously toxic or flammable (ammonia still rules industrial plants, where trained crews manage it). The 1930s brought chlorofluorocarbons, nonflammable, nontoxic, seemingly perfect, until the 1980s discovery that their chlorine was thinning the stratospheric ozone layer led to the Montreal Protocol, the landmark 1987 treaty phasing them out worldwide. Their replacements, hydrofluorocarbons like our R-134a, spare the ozone but are potent greenhouse gases, and are themselves being phased down in favor of low-global-warming choices: hydrofluoroolefins, propane in small sealed units, and CO2 in supermarkets and heat pumps. The thermodynamics you learned today is indifferent to the fluid; the tables change, the enthalpies change, the cycle and the method do not.

Common misconceptions

• A refrigerator makes cold. It moves heat. The evaporator absorbs heat from the food into boiling refrigerant, and everything absorbed, plus the compressor's work, is rejected out the back: q(H) = q(L) + w, as our 179.9 = 143.7 + 36.2 confirmed.
• Leaving the refrigerator door open cools the kitchen. The machine dumps its q(H) into the same room, so the net effect is heating the kitchen by exactly the compressor's power draw.
• A heat pump COP of 5 means 500 percent efficiency, which must be a scam. It moves four units of outdoor heat with one of work; nothing is created. The second law's actual limit here was 5.08 for cooling duty, and the claim sits below it.
• The expansion valve is where the work is recovered. The valve recovers nothing; it is a deliberate throttle chosen because liquid expansion work is too small to be worth a machine. The cold comes from flash evaporation at constant enthalpy.

Recap

• Vapor-compression loop: compressor (work in), condenser (q(H) out, vapor to liquid), expansion valve (h constant, pressure and temperature collapse), evaporator (q(L) in, liquid boils).
• The refrigerant's saturation curve is chosen so it boils below the cold space and condenses above the warm space at practical pressures.
• Worked R-134a cycle (140 kPa to 800 kPa): q(L) = 143.7, w(in) = 36.2, q(H) = 179.9 kJ/kg, COP(R) = 3.97, Carnot ceiling 5.08.
• One ton of refrigeration = 3.517 kW; a 5 kW unit circulates about 0.035 kg/s of refrigerant.
• As a heat pump the same cycle delivers COP(HP) = COP(R) + 1 = 4.97, the case for heat pump heating; performance falls with temperature lift.
• Refrigerant history runs ammonia to CFCs to HFCs to low-GWP fluids, consequences driving each revision; the analysis method survives every change.

Sources

1. U.S. Department of Energy. (n.d.). Heat pump systems. Energy Saver. energy.gov
2. Wikipedia. (n.d.). Vapor-compression refrigeration. Wikimedia Foundation. en.wikipedia.org
3. Britannica. (n.d.). Refrigeration. In Encyclopaedia Britannica. britannica.com
4. OpenStax. (2016). Refrigerators and heat pumps. In University physics volume 2. Rice University. openstax.org

The big picture

Steam plants keep their fire outside the working fluid; the engines of transportation carry the fire inside. In a car engine or a jet turbine, the working fluid is the air, fuel burns directly in it, and the hot products push the piston or spin the turbine themselves. No boiler, no condenser, no tonnage of water: just air, fuel, and geometry, which is why these engines are light enough to fly. Today we analyze the three great gas cycles: Otto, the spark-ignition gasoline engine; Diesel, its compression-ignition sibling; and Brayton, the gas turbine behind jets and much of the power grid. Each yields a closed-form efficiency with the ideal gas toolkit from lesson seven, and each formula tells a design story you can read in any showroom or airport.

To analyze combustion engines without a course in chemistry, engineers use the air-standard idealization: treat the working fluid as pure air behaving as an ideal gas, replace combustion with heat addition from outside, replace the exhaust-and-intake breath with heat rejection that closes the loop, and make the compressions and expansions isentropic. The cold-air-standard version fixes cp, cv, and k at their room-temperature values (1.005, 0.718, 1.4), which is what we will use. The idealization overstates real efficiencies but preserves exactly what we want: how efficiency depends on the ratios a designer actually chooses.

The Otto cycle: the gasoline engine's soul

Nikolaus Otto's 1876 engine breathed in four strokes, and its thermodynamic skeleton is four processes. 1 to 2: isentropic compression of the air-fuel charge from maximum to minimum cylinder volume; the compression ratio r = V(max)/V(min) is the cycle's master parameter. 2 to 3: the spark fires and combustion races through the mixture so fast the piston barely moves: modeled as constant-volume heat addition, pressure and temperature spiking. 3 to 4: isentropic expansion, the power stroke, hot gas shoving the piston and doing the work you drive on. 4 to 1: constant-volume heat rejection, standing in for the exhaust blowdown. Run the first law around the loop with the lesson-seven relations and everything collapses onto one clean result:

eta(Otto) = 1 - 1/r^(k-1).

Nothing but the compression ratio and k. Feel the numbers: r = 8 gives eta = 1 - 1/8^0.4 = 1 - 1/2.297 = 0.565, and r = 10 gives 1 - 1/2.512 = 0.602. Squeeze harder, waste less: compression before burning is what lets the expansion afterward extract more of the heat. The compression also sets the pre-flame temperature, T2 = T1 x r^(k-1) = 300 x 2.297 = 689 K for r = 8, with combustion then vaulting the gas past 2000 K for the power stroke.

So why not r = 20? Because the formula's rising curve runs into chemistry: compress a gasoline-air mixture too far and it detonates spontaneously ahead of the spark, the destructive rattle called knock, which hammers pistons and caps practical ratios near 9 to 12 (octane rating measures a fuel's knock resistance). And why do real gasoline engines deliver 25 to 35 percent rather than our 56? The air-standard model ignored combustion time, heat lost to cylinder walls and coolant, friction, and the pumping cost of breathing, which together claim the difference. The formula still earns its keep: it says, correctly, that every increment of r helps, which is why engine designers have spent a century inching it upward with better fuels, chamber shapes, and electronic timing.

Key idea: The Otto cycle's efficiency is 1 - 1/r^(k-1), rising with compression ratio alone: 56.5 percent at r = 8 ideally, with knock capping r and real losses roughly halving the ideal.

The Diesel cycle: compression as the ignition system

Rudolf Diesel's 1890s insight dodges knock entirely: compress only air, which cannot detonate for want of fuel, then inject fuel into air that lesson seven showed reaches about 953 K at r = 18, hot enough to ignite it on contact. No spark plug, no knock limit, and compression ratios of 15 to 22 against gasoline's 9 to 12. Because injection and burning proceed while the piston retreats, the heat addition is modeled at constant pressure rather than constant volume, and that one change puts a second parameter in the efficiency: the cutoff ratio rc, the volume ratio across the burn (how long injection lasts):

eta(Diesel) = 1 - (1/r^(k-1)) x (rc^k - 1) / (k (rc - 1)).

The bracketed factor exceeds 1 whenever rc exceeds 1, so at equal compression ratio a Diesel scores slightly below an Otto. But the ratios are never equal: try r = 18 with rc = 2. The pieces: 18^0.4 = 3.178; 2^1.4 = 2.639; so eta = 1 - (1/3.178) x (2.639 - 1)/(1.4 x 1) = 1 - 0.3147 x 1.171 = 1 - 0.368 = 0.632. Sixty-three percent ideal, comfortably above the gasoline engine's 56, and the same ordering survives into practice: real diesels deliver roughly 35 to 45 percent, the best large marine engines exceeding 50, which is why trucks, ships, trains, and generators, machines whose fuel bill is a business expense, run diesel. The trade-offs are familiar from the roadside: heavier construction to hold the pressures, and combustion at high temperature and pressure that historically favored soot and NOx, driving the modern aftertreatment stack.

Key idea: Diesel compresses air alone, ignites by compression heat, and accepts a small constant-pressure penalty (the cutoff-ratio factor) in exchange for far higher compression ratios: eta = 63 percent ideal at r = 18, rc = 2.

The Brayton cycle: the engine that flies

Replace reciprocating pistons with steady-flow machinery and you get the gas turbine, whose skeleton is the Brayton cycle: a rotating compressor raises incoming air's pressure (1 to 2, isentropic); a combustor burns fuel in it at essentially constant pressure (2 to 3); a turbine expands the hot gas (3 to 4, isentropic), driving both the compressor and the useful load; and the exhaust-to-intake breath closes the loop as constant-pressure heat rejection (4 to 1). The master parameter is the pressure ratio rp = P2/P1, and the ideal efficiency is the Otto formula in pressure clothing:

eta(Brayton) = 1 - 1/rp^((k-1)/k).

Work a full machine with rp = 8, intake at 300 K, and turbine inlet at 1300 K. Compressor exit: T2 = 300 x 8^0.2857 = 543 K, costing w(comp) = 1.005 x 243.4 = 244.6 kJ/kg. Heat added: q(in) = 1.005 x (1300 - 543.4) = 760.4 kJ/kg. Turbine exit: T4 = 1300/1.811 = 717.7 K, yielding w(turb) = 1.005 x 582.3 = 585.2 kJ/kg. Net work: 585.2 - 244.6 = 340.6 kJ/kg, and eta = 340.6/760.4 = 0.448, agreeing with the formula's 1 - 1/8^0.2857 = 0.448 to the decimal, as it must.

Now the number that distinguishes the gas turbine's whole character: the back-work ratio, 244.6/585.2 = 0.42. The compressor eats forty-two percent of everything the turbine produces, against one half of one percent for the Rankine pump, because compressing gas is brutal (lesson eleven's asymmetry made law). This is why component efficiencies are life and death for gas turbines: shave the turbine's output by 15 percent and the compressor's appetite grows by 18, and the net can halve or vanish; early jet development stalled for exactly this reason until compressors got good. It is also why the turbine inlet temperature T3 is the industry's obsession: hotter T3 grows the turbine's output against a fixed compressor bill, so blades run in gas hotter than their own melting points, kept alive by internal cooling air and ceramic coatings, at 1700 K and beyond in modern machines.

Aircraft complete the story: a jet engine is a Brayton cycle whose turbine extracts only enough work to run the compressor and fan, leaving the rest of the expansion to a nozzle (lesson eight) that converts it to a high-velocity exhaust jet and thrust. And that 717.7 K exhaust in our example is not a rounding error; it is a resource, hot enough to raise steam, which is precisely the opening the final lesson exploits in the combined cycle.

Key idea: Brayton efficiency is 1 - 1/rp^((k-1)/k) (44.8 percent at rp = 8 ideally), but a back-work ratio near 0.42 makes component efficiency and turbine inlet temperature the whole game.

Reading the three formulas side by side

Set the three results next to each other and the family resemblance is the lesson. Every gas cycle's efficiency is 1 minus a temperature-ratio penalty, and every design lever, compression ratio, cutoff ratio, pressure ratio, turbine inlet temperature, works by raising the temperature at which heat enters relative to where it leaves: Carnot's logic wearing three different suits of hardware. The differences are chemistry and machinery: Otto's premixed charge ignites elegantly but knocks; Diesel's air-only compression unlocks higher ratios at the cost of weight; Brayton's steady flow scales to enormous power and flight but pays a compressor tax. When you next choose a car, watch a truck climb a grade, or feel a jet spool up, you are watching these three formulas negotiate with their constraints.

Common misconceptions

• The spark plug's timing is a detail; compression is just squeezing. Compression ratio is the single lever in the Otto efficiency formula; it is the point of the design, and knock is the chemical wall it runs into.
• Diesels are efficient because diesel fuel contains more energy. Diesel fuel's volumetric energy density is only modestly higher; the efficiency edge comes from compression ratios of 15 to 22 that spark engines cannot survive.
• A gas turbine's net output is basically its turbine's output. The compressor consumes 40 to 60 percent of gross turbine work; the machine sells only the difference, which is why component efficiencies dominate the design.
• Ideal-cycle efficiencies describe real engines. They are ceilings under the air-standard idealization; wall heat loss, finite combustion, friction, and breathing costs bring a 56 percent ideal Otto to a 25 to 35 percent car engine. The formulas' value is showing which levers matter.

Recap

• Air-standard analysis: air as ideal gas, combustion as heat addition, exhaust as heat rejection, isentropic strokes, constant specific heats.
• Otto: eta = 1 - 1/r^(k-1); 56.5 percent at r = 8, 60.2 at r = 10; knock caps r near 9 to 12 for gasoline.
• Diesel: compress air only, ignite by 900 K-plus compression heat; eta = 1 - (1/r^(k-1)) x (rc^k - 1)/(k(rc - 1)) = 63.2 percent at r = 18, rc = 2; higher r beats the cutoff penalty.
• Brayton: eta = 1 - 1/rp^((k-1)/k); worked machine at rp = 8, T3 = 1300 K gives w(net) = 340.6 kJ/kg, eta = 44.8 percent, back-work ratio 0.42.
• Turbine inlet temperature and component efficiencies rule gas turbines; jets swap shaft work for nozzle thrust; hot Brayton exhaust invites a steam bottoming cycle.

Sources

1. NASA Glenn Research Center. (n.d.). Otto cycle. Beginner's Guide to Aeronautics. grc.nasa.gov
2. NASA Glenn Research Center. (n.d.). Brayton cycle. Beginner's Guide to Aeronautics. grc.nasa.gov
3. Britannica. (n.d.). Diesel engine. In Encyclopaedia Britannica. britannica.com
4. U.S. Department of Energy. (n.d.). How gas turbine power plants work. Office of Fossil Energy and Carbon Management. energy.gov

The big picture

This is the capstone, and its purpose is to turn everything you have learned into a working instrument panel for the actual world. The skills are all in hand: energy balances, cycle efficiencies, coefficients of performance, and the Carnot ceilings that govern them. What remains is to connect them the way reality does: in chains. No machine works alone. The electricity in your kettle came from a flame through a boiler through a turbine through a generator through hundreds of kilometers of wire, and every link took its cut. Today we follow such chains end to end, meet the two great architectures engineers invented to fatten them (combined cycles and cogeneration), settle two questions of immediate household relevance (how should you heat a home, and what does an electric car really save?), and finish with the habit of mind this course most wants to leave you: second-law thinking, the practice of asking not just how much energy, but how useful.

Efficiencies multiply, and chains are only as fat as their thinnest links

The rule for chains is simple and unforgiving: overall efficiency is the product of the links. Consider coal-fired electricity driving a factory motor. The plant converts fuel heat to electricity at 38 percent (you computed why in the Rankine lesson). Transmission and distribution deliver about 95 percent of what enters the wires. The motor converts perhaps 90 percent of arriving electricity to shaft work. Fuel to spinning shaft: 0.38 x 0.95 x 0.90 = 0.32. Two thirds of the coal's energy never reaches the task, and the arithmetic tells you exactly where it went: the plant's condenser dwarfs every other loss combined. This is the pattern nationwide: the Lawrence Livermore National Laboratory's famous energy flow charts show roughly two thirds of all primary energy the United States consumes ending as rejected heat, overwhelmingly from thermal power generation and combustion transportation, the two second-law tollbooths you now understand from the inside.

The product rule also directs repair effort. Improving the 95 percent link to 97 buys almost nothing; improving the 38 percent link to 60 transforms the chain. So engineers attack the conversion step, and the two best attacks are worth knowing by name.

Key idea: Chained conversions multiply: 0.38 x 0.95 x 0.90 = 0.32 from coal pile to motor shaft, and the thermal conversion link is where most of civilization's energy is lost.

Attack one: the combined cycle

Recall two numbers from the cycles module. A gas turbine at pressure ratio 8 exhausted at 717.7 K, and you were promised that heat was a resource. A Rankine boiler, meanwhile, wants nothing more than a steady stream of 700 K gas to boil water with. Pipe the Brayton exhaust into a heat recovery steam generator, run a steam cycle on it, and you have the combined cycle: two engines eating one fire, the second dining on the first's rejected heat. The arithmetic of stacking is lovely: if the gas turbine converts fraction eta(B) of the fuel and the steam cycle converts fraction eta(R) of what the gas turbine rejects, the total is eta(cc) = eta(B) + eta(R) x (1 - eta(B)). With eta(B) = 0.45 and eta(R) = 0.35: 0.45 + 0.35 x 0.55 = 0.64. Modern combined-cycle plants routinely reach 60 percent and the record machines exceed 63, against roughly 40 for the best simple cycles: the most efficient large-scale conversion of fuel heat to electricity humanity has achieved. When natural gas displaced coal across much of the power sector, it arrived riding this architecture.

Attack two shares the philosophy but changes the customer. Cogeneration, or combined heat and power, notices that the second law only forbids converting rejected heat to work; it says nothing against using it as heat. A university campus or district system that pipes its power plant's condenser heat into buildings turns q(out) from waste into product, pushing total fuel utilization, work plus useful heat, to 80 percent or more. The efficiency number needs honest reading (work and heat are different-quality products, as the next section insists), but the fuel savings are real and large, which is why hospitals, campuses, refineries, and Scandinavian cities run on CHP.

Key idea: Combined cycles stack a steam cycle on gas turbine exhaust, eta(cc) = eta(B) + eta(R)(1 - eta(B)), reaching about 60 percent; cogeneration sells the rejected heat itself, pushing fuel utilization past 80.

Two household verdicts: heating, and driving

Now aim the tools at your own utility bill. How to heat a home? Compare three options per unit of fuel burned at the source, taking a fossil grid at 40 percent for the electric paths. Electric resistance heat: 0.40 x 1.00 = 0.40 units of indoor heat per unit of fuel; a gas furnace burning the fuel on site: 0.90 to 0.95, and the furnace seems to win. But the heat pump changes the game: at a seasonal COP of 3.5, it delivers 0.40 x 3.5 = 1.40 units of heat per unit of source fuel, mining the difference from outdoor air, and it beats the furnace by half again even through a fossil-fired grid. Feed it cleaner electricity and the margin only widens. This one calculation, resistance 0.4, furnace 0.9, heat pump 1.4, is quietly reorganizing the heating industry, and you have just performed it from first principles. It also explains the fine print you can now anticipate: the advantage narrows in extreme cold as the COP sags with rising temperature lift, exactly as Carnot's TL/(TH - TL) predicts.

What does an electric car save? A gasoline drivetrain converts about 25 to 30 percent of tank energy to wheels: the Otto lesson told you why the ceiling is where it is, and accessories, part-load operation, and friction do the rest. The electric path from a fossil plant: 0.40 (generation) x 0.90 (charging and battery) x 0.90 (motor and electronics) = 0.32, plus regenerative braking clawing back city-driving losses no combustion car can touch. Through today's mixed grid the EV runs comparable-to-better on primary energy and improves automatically as the grid does, while the combustion car's chain is frozen at the pump. The deeper lesson is the method: never compare machines at one link; compare whole chains, from source to task, and state your grid assumptions out loud.

Key idea: Whole-chain accounting crowns the heat pump (1.4 units of heat per unit of source fuel versus 0.9 for a furnace) and shows electric drivetrains matching or beating gasoline through even a fossil grid, with headroom as generation cleans.

Second-law thinking: quality, matching, and the audit kit

Behind all these verdicts stands the course's deepest idea: energy has quality as well as quantity. A kilojoule of shaft work or electricity can become anything, including several kilojoules of moved heat. A kilojoule of 2000 K combustion gas can still become work at up to Carnot's rate. A kilojoule of 30 C water can heat your feet and nothing else. Engineers formalize quality as exergy, the maximum work extractable from energy in a given environment, and though we will not compute exergy budgets here, the qualitative habit is the payoff: match the quality of the source to the quality of the task. Burning a 2000 K flame to make 20 C room air is thermodynamic malpractice, legal under the first law and wasteful under the second; the flame deserved to make electricity first (combined cycle) or the room deserved a heat pump. Every architecture in this lesson, stacking cycles, selling rejected heat, pumping heat instead of making it, is that one sentence wearing different hardware.

And the habit doubles as a fraud detector. Assemble the audit kit you have been building all course. First law: outputs cannot exceed inputs around a cycle, so any device claiming to multiply energy outright is dead on arrival. Second law: engine efficiency cannot beat 1 - TL/TH between its actual temperatures, and heat mover COP cannot beat TL/(TH - TL); a claimed 80 percent engine between 600 K and 300 K fails in one line. Chains: efficiencies multiply, so a long chain of plausible links can still be a poor system, and a marketer quoting one shiny link is answering a question you did not ask. Units: kilowatts are a rate and kilowatt-hours an amount, and confusing them, innocently or otherwise, is the commonest energy error in public life. A COP above 1 is legal (heat is moved); an efficiency above 1 is not (work would be created). That kit, plus the reflex of drawing a boundary before arguing, is the durable residue of this course.

Key idea: Energy degrades in quality even when conserved in quantity, so match source quality to task and audit every claim with the two laws, the product rule for chains, and clean units.

Where you now stand

Take stock of what you can do that you could not fifteen lessons ago. You can fix the state of steam or air from two properties and a table. You can balance energy on a tank, a turbine, or a whole plant, and grade real machines against their isentropic twins. You can compute what a power plant, refrigerator, or engine can possibly deliver from its temperatures alone, and how close its cycle actually comes. You can follow a kilojoule from coal seam or wind farm to kettle or wheel, and say where it thinned. This toolkit is nearly two centuries old and has never needed recall or revision; it will price technologies that do not exist yet, because whatever they are, they will obey the two laws. Physics offers few better investments. Wherever your engineering goes next, heat transfer, fluids, propulsion, energy policy, batteries, you now carry the laws that all of them answer to.

Common misconceptions

• Efficiencies along a chain add or average. They multiply: 0.38 x 0.95 x 0.90 = 0.32, and the chain is dominated by its weakest conversion link, not helped by its strongest.
• Waste heat is gone. Rejected heat is low-quality, not nonexistent: combined cycles convert part of it to more work, and cogeneration delivers it as useful heat; only its capacity for work has degraded.
• A heat pump COP of 3.5 must be marketing fiction. Moving heat is not creating energy; the second-law ceiling for typical winter lifts is well above 3.5, and the measured performance is routine.
• Electric cars merely relocate the exhaust pipe. Partly true through a fossil grid, and the chains still come out comparable or better; decisively false as generation decarbonizes, because the electric chain upgrades at the source while the gasoline chain cannot.

Recap

• Overall efficiency is the product of link efficiencies; nationally, about two thirds of primary energy ends as rejected heat, concentrated at thermal conversion links.
• Combined cycles run a Rankine cycle on Brayton exhaust: eta(cc) = eta(B) + eta(R)(1 - eta(B)), about 0.64 for 0.45 and 0.35, and real plants exceed 60 percent.
• Cogeneration sells rejected heat as a product, lifting fuel utilization past 80 percent for campuses and cities.
• Per unit of source fuel: resistance heating delivers about 0.4, a furnace 0.9, a heat pump at COP 3.5 about 1.4; electric drivetrains match or beat gasoline whole-chain and improve with the grid.
• Second-law thinking: energy quality (exergy) matters as much as quantity; match source to task; audit claims with the two laws, multiplied chains, and kW-versus-kWh discipline.

Sources

1. Lawrence Livermore National Laboratory. (n.d.). Energy flow charts. LLNL. flowcharts.llnl.gov
2. U.S. Energy Information Administration. (n.d.). Energy explained. EIA. eia.gov
3. U.S. Department of Energy. (n.d.). Heat pump systems. Energy Saver. energy.gov
4. U.S. Department of Energy. (n.d.). How gas turbine power plants work. Office of Fossil Energy and Carbon Management. energy.gov