Engineering Thermodynamics

Chapter 6
Entropy and Irreversibility

Bachelors Program in Mechnical Engineering
Engineering Thermodynamics   •   Bachelor

Entropy and Irreversibility

Introduction

The first law of thermodynamics expresses conservation of energy, but it does not indicate whether a process can occur naturally or how much of its energy is available for useful work. The second law supplies this directionality through the property called entropy. It explains why heat flows spontaneously from a hot body to a cold body, why friction degrades mechanical energy, why mixing is generally irreversible, and why no heat engine can convert all supplied heat into work.

Entropy analysis is central to the design and evaluation of turbines, compressors, pumps, heat exchangers, refrigeration systems, combustion equipment, and power cycles. For an engineer, entropy generation is especially important because it identifies where performance is being lost and where design improvements are most valuable.

Learning Objectives

By the end of this chapter, the reader should be able to:

  1. Define entropy and explain its physical significance.
  2. Use property relations to calculate entropy changes for ideal gases and incompressible substances.
  3. Apply entropy balances to closed systems and control volumes.
  4. Calculate entropy generation and determine whether a process is reversible or irreversible.
  5. Explain irreversibility caused by heat transfer through finite temperature differences, friction, mixing, and unrestrained expansion.
  6. Evaluate practical devices using isentropic efficiency and entropy-generation analysis.

1. The Second Law and the Need for Entropy

The first law for a closed system is

∆E = Q − W

where ∆E is the change in total energy of the system, Q is heat transfer to the system, and W is work done by the system. This equation permits many processes that do not occur in reality. For example, it does not prohibit heat flowing from a cold object to a hot object without work input.

The second law imposes this missing restriction. Two classical statements are equivalent:

  • Kelvin–Planck statement: No cyclic heat engine can receive heat from a single reservoir and convert it completely into work.
  • Clausius statement: No cyclic device can transfer heat from a lower-temperature reservoir to a higher-temperature reservoir without work input.

These statements establish that natural processes have a preferred direction. They also lead to the concept of entropy.

A reversible process is an ideal limiting process that can be reversed without leaving any net change in the system and surroundings. A real process is irreversible because of effects such as friction, unrestrained expansion, mixing of different substances, electrical resistance, chemical reaction, and heat transfer across a finite temperature difference.

2. Entropy as a Thermodynamic Property

For a reversible process, the differential change in entropy is defined by

dS = δQrev / T

where S is entropy, δQ_rev is infinitesimal reversible heat transfer, and T is the absolute temperature at the system boundary. Entropy is a property, so its change depends only on the initial and final equilibrium states, even when the actual process is irreversible.

For a finite change between states 1 and 2,

∆S = S₂ − S₁ = ∫₁² δQrev / T

The integral is evaluated along any convenient reversible path connecting the two states. It is not necessary for the actual process to be reversible.

Specific entropy is entropy per unit mass:

∆s = s₂ − s₁

where s has units of kJ·kg⁻¹·K⁻¹ in engineering applications. Total entropy S has units of kJ·K⁻¹. Since entropy is an extensive property, total entropy is additive, whereas specific entropy is an intensive property.

Entropy and the Clausius Inequality

For any cycle, the Clausius inequality is

∮ δQ / T ≤ 0

Equality applies to a reversible cycle; strict inequality applies to an irreversible cycle. This inequality provides the mathematical basis for the entropy balance and the second-law inequality.

3. The Entropy Balance

The general entropy balance for a system is

∆Ssystem = ∑(Qk / Tk) + Sgen

For a finite process, Q_k is heat transfer at boundary temperature T_k, and S_gen is entropy generated internally. The equation assumes that heat transfer occurs at a well-defined boundary temperature. More generally, for a differential process,

dSsystem = δQ / Tboundary + dSgen

The second law requires

Sgen ≥ 0

with S_gen = 0 only for a reversible process. Entropy generation cannot be negative. A calculated negative value usually indicates an incorrect sign convention, inconsistent property data, or an invalid assumption.

For an isolated system, Q = 0 and W does not transfer entropy across the boundary. Therefore,

∆Sisolated system = Sgen ≥ 0

This is the entropy form of the second law.

Control-Volume Entropy Balance

For a control volume, the rate form is

dScv/dt = ∑(Q̇k/Tk) + ∑ṁin sin − ∑ṁout sout + Ṡgen

where S_cv is entropy within the control volume, Q̇_k is heat-transfer rate at boundary temperature T_k, ṁ is mass-flow rate, s is specific entropy, and Ṡ_gen is entropy-generation rate.

At steady state, dS_cv/dt = 0, so

0 = ∑(Q̇k/Tk) + ∑ṁin sin − ∑ṁout sout + Ṡgen

For a one-inlet, one-outlet device,

Ṡgen = ṁ(s₂ − s₁) − ∑(Q̇k/Tk)

The sign convention is that heat transfer into the control volume is positive.

4. Entropy Relations for Common Substances

4.1 Incompressible Substances

For a substance with approximately constant specific heat c, the entropy change is

∆s = c ln(T₂/T₁)

where T₁ and T₂ are absolute temperatures in kelvin and c is the specific heat, usually c_p or c for a liquid or solid. If the temperature is constant, the entropy change due to phase change is

∆s = h₂₁/Tsat

where h₂₁ is the latent enthalpy change at saturation temperature T_sat. For vaporization, this becomes ∆s = h_fg/T_sat.

4.2 Ideal Gases

For an ideal gas with constant specific heats,

s₂ − s₁ = cp ln(T₂/T₁) − R ln(p₂/p₁)

or, equivalently,

s₂ − s₁ = cv ln(T₂/T₁) + R ln(v₂/v₁)

Here c_p and c_v are constant-pressure and constant-volume specific heats, R is the gas constant, p is absolute pressure, and v is specific volume. The relation follows from the ideal-gas equation of state and the thermodynamic property relations.

If pressure is constant,

∆s = cp ln(T₂/T₁)

If temperature is constant,

∆s = −R ln(p₂/p₁) = R ln(v₂/v₁)

For accurate work over large temperature ranges, temperature-dependent specific heats or ideal-gas entropy tables should be used rather than constant average values.

4.3 General Simple Compressible Substances

Two useful Gibbs relations are

T ds = du + p dv

and

T ds = dh − v dp

where u is specific internal energy, h is specific enthalpy, p is pressure, and v is specific volume. These relations apply to a simple compressible substance in equilibrium and are useful for deriving entropy changes from property data.

For a liquid or solid, the approximation v ≈ constant often gives

ds ≈ c dT/T − v dp/T

The pressure term is commonly negligible for liquids under moderate pressure changes, but it should not be ignored automatically in high-pressure applications.

5. Entropy Generation and Irreversibility

Entropy generation measures the extent of thermodynamic imperfection. It is not the same as entropy transfer by heat or mass. Entropy may cross a system boundary, whereas entropy generation occurs inside the system or control volume because of irreversibility.

The principal causes include:

  • heat transfer through a finite temperature difference;
  • fluid friction and viscous dissipation;
  • unrestrained expansion;
  • mixing of fluids or gases with different states;
  • electrical resistance;
  • inelastic deformation;
  • chemical reactions and combustion;
  • shock waves and throttling.

A useful engineering measure is the irreversibility, also called lost work or exergy destruction:

I = T₀ Sgen

where I is irreversibility and T₀ is the temperature of the surrounding environment, both expressed consistently. If S_gen is in kJ·K⁻¹, I is in kJ when T₀ is in kelvin. In rate form,

Ī = T₀ Ṡgen

This is the Gouy–Stodola theorem. It states that the maximum useful work destroyed by irreversibility equals the ambient absolute temperature multiplied by entropy generation. It applies when the environment is characterized by a uniform reference temperature T₀ and when the work potential is evaluated relative to that environment.

Heat Transfer Through a Finite Temperature Difference

Suppose heat Q flows from a reservoir at T_H to one at T_L, where T_H > T_L. The total entropy generation is

Sgen = Q(1/TL − 1/TH)

This is positive because 1/T_L > 1/T_H. Heat transfer would be reversible only if it occurred through an infinitesimal temperature difference, requiring an idealized sequence of intermediate temperatures.

Friction

Friction converts organized mechanical energy into internal energy. If an insulated object slides to rest, the total entropy increase is

Sgen = ∆Sobject + ∆Ssurface

For a constant-temperature reservoir receiving frictional heat Q_f, S_gen = Q_f/T₀. In fluid systems, viscous friction produces a pressure drop and entropy increase. For steady adiabatic flow with negligible changes in kinetic and potential energy, friction causes h₂ ≈ h₁ but s₂ > s₁.

Mixing

When two fluids at different temperatures, compositions, or chemical potentials mix, the final equilibrium state generally has greater entropy than the separated initial state. Mixing is irreversible unless the substances are identical and initially at the same thermodynamic state.

For ideal gases mixed isothermally at the same pressure, the entropy of mixing is

∆Smix = −R∑ ni ln(yi)

where n_i is the amount of component i, y_i is its final mole fraction, and R is the universal gas constant. Since 0 < y_i ≤ 1, the expression is nonnegative. This relation assumes ideal-gas behavior and no chemical reaction.

Unrestrained Expansion

In free or unrestrained expansion, a gas expands into a vacuum. There is no opposing pressure, so boundary work is zero. If the container is insulated, heat transfer is also zero. The first law then gives ∆U = 0. For an ideal gas, internal energy depends only on temperature, so T₂ = T₁. However, volume increases and entropy does not remain constant:

∆s = R ln(v₂/v₁) > 0

The process is therefore irreversible even though Q = 0 and W = 0. This example demonstrates why the first law alone cannot detect irreversibility.

6. Reversible and Isentropic Processes

A process is isentropic if

∆s = 0

An isentropic process is reversible and adiabatic. The converse is important: an adiabatic process is not necessarily isentropic, because friction or another irreversibility may generate entropy.

For an ideal gas with constant specific heats undergoing a reversible adiabatic process,

p vk = constant
T v^(k−1) = constant
T₂/T₁ = (p₂/p₁)^((k−1)/k)

where k = c_p/c_v. These equations require ideal-gas behavior, constant specific heats, and a reversible adiabatic process. They should not be applied to a throttling valve, a frictional duct, or any process with significant heat transfer.

7. Isentropic Efficiencies of Devices

Isentropic efficiencies compare actual device performance with an ideal reversible adiabatic reference.

For a turbine,

ηt = (h₁ − h₂,actual)/(h₁ − h₂s)

where state 2s is the exit state obtained by setting s₂s = s₁ at the actual exit pressure. The turbine efficiency is normally less than unity.

For a compressor,

ηc = (h₂s − h₁)/(h₂,actual − h₁)

The actual compressor requires more work than the isentropic compressor, so η_c is also less than unity.

For a pump handling an incompressible liquid, a similar expression based on enthalpy or shaft work may be used. These definitions assume steady operation and clearly specified inlet and exit pressures. They are not measures of entropy generation directly, but lower efficiency generally corresponds to greater entropy generation.

8. Worked Examples

Example 1: Entropy Generation During Heat Transfer

A quantity of heat Q = 500 kJ flows from a reservoir at 800 K to a reservoir at 300 K. Determine the entropy generation and the irreversibility relative to an environment at T₀ = 300 K.

For the two-reservoir system,

Sgen = Q(1/TL − 1/TH)

Therefore,

Sgen = 500(1/300 − 1/800)
= 500(0.003333 − 0.001250)
= 1.042 kJ/K

The irreversibility is

I = T₀ Sgen = 300 × 1.042 = 312.5 kJ

Thus, 1.042 kJ/K of entropy is generated and 312.5 kJ of potential work is destroyed. The result demonstrates why heat exchangers should minimize unnecessary temperature differences while still meeting practical size and cost constraints.

Example 2: Entropy Change of an Ideal Gas

Air is modeled as an ideal gas with c_p = 1.005 kJ·kg⁻¹·K⁻¹ and R = 0.287 kJ·kg⁻¹·K⁻¹. It changes from T₁ = 300 K, p₁ = 100 kPa to T₂ = 600 K, p₂ = 500 kPa. Determine the specific entropy change.

Use

∆s = cp ln(T₂/T₁) − R ln(p₂/p₁)

Substitution gives

∆s = 1.005 ln(600/300) − 0.287 ln(500/100)

Using ln 2 = 0.6931 and ln 5 = 1.6094,

∆s = 1.005(0.6931) − 0.287(1.6094)
= 0.6966 − 0.4619
= 0.235 kJ·kg⁻¹·K⁻¹

The entropy increases because the temperature effect exceeds the entropy decrease associated with compression.

Example 3: Free Expansion of an Ideal Gas

An insulated rigid vessel is divided into two equal compartments. One contains an ideal gas and the other is evacuated. When the partition is removed, the gas fills the entire vessel. If the initial volume is 0.10 m³ and the final volume is 0.20 m³, determine the specific entropy change for nitrogen with R = 0.2968 kJ·kg⁻¹·K⁻¹.

Because the gas is ideal and the vessel is insulated and rigid, ∆U = 0, so temperature remains constant. Thus,

∆s = R ln(v₂/v₁) = 0.2968 ln(0.20/0.10)
∆s = 0.2968(0.6931) = 0.2057 kJ·kg⁻¹·K⁻¹

The entropy increases even though heat transfer and work are both zero. The process is irreversible because the gas expands against no opposing pressure and cannot be restored without external effects.

9. Entropy Analysis of Engineering Devices

Turbines and Expanders

An adiabatic turbine operating at steady state with negligible kinetic and potential energy changes has

Ṡgen = ṁ(s₂ − s₁)

Since Ṡ_gen ≥ 0, the actual exit entropy must satisfy s₂ ≥ s₁. A turbine with excessive friction, leakage, or nonuniform flow has larger entropy generation and produces less work than an ideal turbine.

Compressors and Pumps

For an adiabatic compressor,

Ṡgen = ṁ(s₂ − s₁) ≥ 0

The entropy increase reflects friction, turbulence, leakage, and heat transfer effects. Cooling can reduce the required compression work, but heat transfer itself must be included in the entropy balance.

Throttling Valves

A steady throttling valve is commonly modeled with negligible heat transfer, shaft work, and changes in kinetic and potential energy. The steady-flow energy equation gives

h₁ = h₂

The entropy balance then shows

s₂ ≥ s₁

A throttling process is isenthalpic, not isentropic. Confusing these two conditions is a frequent error.

Heat Exchangers

For an insulated heat exchanger with no mixing between streams,

Ṡgen = ∑ṁout sout − ∑ṁin sin ≥ 0

Entropy generation results mainly from finite temperature differences and pressure losses. A balanced design must consider heat-transfer area, pressure drop, capital cost, and operating efficiency.

10. Applied Case: Diagnosing Losses in a Steam-Turbine Installation

This is an illustrative engineering case.

Background

A small industrial plant operates a steam turbine to generate electricity and process heat. Operators report that actual power output is lower than the value predicted from the turbine design model.

Problem

Measurements show that the inlet pressure and temperature are close to design values, but the measured exhaust entropy is substantially higher than the isentropic prediction. The plant must identify the dominant losses without replacing the turbine immediately.

Analysis

The engineering team applies a steady-flow entropy balance. Assuming one inlet, one outlet, and negligible heat transfer,

Ṡgen = ṁ(s₂ − s₁)

They compare the measured entropy increase with values from historical operation. A progressive increase is associated with deposits on blade surfaces and partial seal leakage. The pressure ratio alone cannot identify these losses; the entropy increase provides a direct second-law indicator.

The team also calculates turbine isentropic efficiency:

ηt = (h₁ − h₂,actual)/(h₁ − h₂s)

A lower-than-design efficiency confirms that the available enthalpy drop is not being converted effectively into shaft work.

Approach and Solution

The plant schedules a controlled inspection, cleans deposits, repairs seals, and verifies inlet measurement calibration. It also introduces routine monitoring of inlet and outlet pressure, temperature, mass flow, and estimated entropy generation. Operating procedures are revised to avoid prolonged wet-steam operation, which can accelerate blade erosion.

Outcome

After maintenance, the measured exit entropy decreases and turbine efficiency improves. The plant obtains more power from the same steam flow, reducing the amount of supplementary purchased electricity.

Learning Points

  1. Entropy generation is a diagnostic quantity, not merely a theoretical property.
  2. Pressure and temperature measurements should be combined with reliable property data.
  3. Efficiency comparisons require a clearly defined isentropic reference state.
  4. Maintenance decisions can be prioritized by locating components with the greatest entropy generation.

11. Common Errors and Best Practices

A frequent error is treating entropy as heat. Entropy is a property, while heat is energy transferred across a boundary. Another error is assuming that every adiabatic process is isentropic. Adiabaticity eliminates heat transfer, but friction and other irreversibilities may still generate entropy.

Additional best practices include:

  • Use kelvin, not degrees Celsius, in logarithmic entropy equations.
  • Use absolute pressure in ideal-gas entropy relations.
  • Establish heat and mass-flow sign conventions before writing balances.
  • Distinguish total entropy from specific entropy and maintain units.
  • Use property tables or validated software for real fluids and phase-change states.
  • Check that calculated entropy generation is nonnegative.
  • Report assumptions such as steady state, negligible kinetic energy, or constant specific heats.
  • Use entropy generation together with energy balances; neither balance alone is sufficient for complete performance assessment.

12. Limitations and Contemporary Engineering Practice

Entropy analysis identifies the magnitude of thermodynamic losses but does not by itself determine the cheapest or most practical design. A lower-entropy design may require greater surface area, more complex controls, expensive materials, or higher maintenance. Engineering decisions must therefore combine second-law analysis with economic, environmental, safety, and reliability considerations.

Modern plants increasingly use digital sensors and supervisory control systems to estimate real-time entropy generation and exergy destruction. Data-driven fault detection can identify deviations from expected turbine, compressor, or heat-exchanger behavior. However, AI-based predictions must be checked against conservation laws, calibrated instruments, valid property models, and physically plausible bounds. An algorithm that reports negative entropy generation should be treated as a model or measurement fault, not as a new thermodynamic result.

Chapter Summary

Entropy is a thermodynamic property that quantifies energy dispersal and establishes the direction of real processes. For a reversible process, dS = δQ_rev/T; for an actual process, the entropy balance includes entropy generation. The second law requires S_gen ≥ 0, with equality only for a reversible process.

Entropy changes can be evaluated using property tables, ideal-gas relations, and incompressible-substance approximations. Closed-system and control-volume entropy balances allow engineers to analyze turbines, compressors, valves, pumps, heat exchangers, and mixing devices. Important irreversibilities arise from finite-temperature-difference heat transfer, friction, mixing, and unrestrained expansion.

The Gouy–Stodola theorem, I = T₀S_gen, converts entropy generation into lost work relative to the environment. Isentropic efficiencies provide practical comparisons between actual and ideal turbines, compressors, and pumps. Reliable entropy analysis requires consistent units, absolute temperature, correct property data, explicit assumptions, and verification that entropy generation is nonnegative.

Key Terms

Entropy: A thermodynamic property that measures energy dispersal and process directionality.

Specific entropy: Entropy per unit mass, usually measured in kJ·kg⁻¹·K⁻¹.

Entropy generation: Entropy produced by irreversibilities within a system or control volume.

Reversible process: An ideal process that can be reversed without net changes to system and surroundings.

Irreversible process: A real process involving entropy generation.

Second law of thermodynamics: Law establishing process direction and limiting heat-engine performance.

Clausius inequality: The inequality ∮δQ/T ≤ 0 for any thermodynamic cycle.

Isentropic process: A process with constant entropy.

Adiabatic process: A process with no heat transfer.

Irreversibility: Lost work caused by entropy generation.

Gouy–Stodola theorem: Relation I = T₀S_gen between lost work and entropy generation.

Unrestrained expansion: Expansion into a vacuum with no opposing pressure.

Mixing irreversibility: Entropy generation caused by combining different substances or states.

Isentropic efficiency: Ratio comparing actual device performance with ideal reversible adiabatic performance.

Throttling: Flow through a restriction, commonly modeled as an isenthalpic process.

Control volume: A specified region through which mass may flow.

Finite temperature difference: A temperature gradient that causes irreversible heat transfer.

Property relation: Thermodynamic equation relating state variables and properties.

Review Questions

Short Answer Questions

  1. What is the physical meaning of entropy generation?
  2. State the Clausius inequality and explain when equality applies.
  3. Why is an adiabatic process not necessarily isentropic?
  4. What is the difference between an isenthalpic and an isentropic process?
  5. Why does entropy increase during the free expansion of an ideal gas?

Long Answer Questions

  1. Derive the ideal-gas entropy-change relation using temperature and pressure and state its assumptions.
  2. Explain the entropy balance for a steady-flow control volume and apply it conceptually to a turbine.
  3. Discuss the sources of irreversibility in a heat exchanger and methods for reducing them.
  4. Explain the Gouy–Stodola theorem and its significance in engineering design.
  5. Compare reversible heat transfer, frictional dissipation, mixing, and unrestrained expansion as sources of entropy generation.

Analytical / Application Questions

  1. A heat quantity of 250 kJ flows from a reservoir at 600 K to one at 300 K. Calculate entropy generation and the irreversibility relative to an environment at 300 K.
  2. Air with c_p = 1.005 kJ·kg⁻¹·K⁻¹ and R = 0.287 kJ·kg⁻¹·K⁻¹ changes from 300 K and 100 kPa to 450 K and 300 kPa. Determine its specific entropy change.
  3. An adiabatic turbine has inlet entropy 6.8 kJ·kg⁻¹·K⁻¹ and measured exit entropy 7.1 kJ·kg⁻¹·K⁻¹ at a mass flow rate of 4 kg/s. Determine the entropy-generation rate and interpret the result.

Multiple Choice Questions with Answers

Question 1

Which condition is required for a reversible process?

A. Entropy generation is zero

B. Heat transfer is always zero

C. Pressure remains constant

D. Internal energy remains constant

Correct Answer: A. Entropy generation is zero

Question 2

The entropy balance for an isolated system reduces to which relation?

A. ∆S = 0 for every process

B. ∆S = S_gen ≥ 0

C. ∆S = Q/T

D. ∆S = −S_gen

Correct Answer: B. ∆S = S_gen ≥ 0

Question 3

For an ideal gas with constant specific heats, the specific entropy change is

A. c_p ln(p₂/p₁) − R ln(T₂/T₁)

B. c_p ln(T₂/T₁) − R ln(p₂/p₁)

C. R ln(T₂/T₁) − c_p ln(p₂/p₁)

D. c_v ln(p₂/p₁) + R ln(T₂/T₁)

Correct Answer: B. c_p ln(T₂/T₁) − R ln(p₂/p₁)

Question 4

A throttling valve is ideally modeled as a process with constant

A. Entropy

B. Temperature

C. Enthalpy

D. Pressure

Correct Answer: C. Enthalpy

Question 5

Heat Q flows from a reservoir at T_H to one at T_L, where T_H > T_L. The entropy generation is

A. Q(1/T_H − 1/T_L)

B. Q(T_H − T_L)

C. Q/(T_H T_L)

D. Q(1/T_L − 1/T_H)

Correct Answer: D. Q(1/T_L − 1/T_H)

Question 6

In an insulated free expansion of an ideal gas, which statement is correct?

A. Temperature decreases necessarily

B. Entropy decreases

C. Internal energy remains constant and entropy increases

D. Boundary work is positive

Correct Answer: C. Internal energy remains constant and entropy increases

Question 7

The Gouy–Stodola theorem states that irreversibility equals

A. S_gen/T₀

B. T₀S_gen

C. Q/T₀

D. W/T₀

Correct Answer: B. T₀S_gen

Question 8

For a steady, adiabatic, one-inlet, one-outlet turbine with negligible changes in kinetic and potential energy, entropy generation rate is

A. ṁ(s₁ − s₂)

B. ṁ(s₂ − s₁)

C. ṁ(h₂ − h₁)

D. ṁ(p₂ − p₁)

Correct Answer: B. ṁ(s₂ − s₁)

Question 9

Which process is necessarily irreversible under ordinary conditions?

A. Quasistatic isothermal expansion

B. Reversible adiabatic compression

C. Unrestrained expansion into a vacuum

D. Reversible phase change

Correct Answer: C. Unrestrained expansion into a vacuum

Question 10

In a real adiabatic compressor, entropy generation is associated primarily with

A. Friction, turbulence, and leakage

B. A decrease in mass flow to zero

C. Perfectly reversible heat transfer

D. Constant pressure operation only

Correct Answer: A. Friction, turbulence, and leakage

Engineering Thermodynamics
Chapter 6 of 15