Open access peer-reviewed chapter

Minimal Dissipation Processes in Irreversible Thermodynamics and Their Applications

Written By

Margarita Anatolyevna Zaeva and Anatoly Mikhailovich Tsirlin

Submitted: November 13th, 2018 Reviewed: January 24th, 2019 Published: March 27th, 2019

DOI: 10.5772/intechopen.84703

Chapter metrics overview

683 Chapter Downloads

View Full Metrics


It is known that the maximum efficiency of conversion of thermal energy into mechanical work or separation work is achieved in reversible processes. If the intensity of the target flux is set, the processes in the thermodynamic system are irreversible. In this case, the role of reversible processes is played by the processes of minimal dissipation. The review presents the derivation of conditions for minimum dissipation in general form and their specification for heat and mass transfer processes with arbitrary dynamics. It is shown how these conditions follow the solution of problems on the optimal organization of two-flux and multiflux heat exchange. The algorithm for the synthesis of heat exchange systems with given water equivalents and the phase state of the flows is described. The form of the region of realizability of systems using thermal energy and the problem of choosing the order of separation of multicomponent mixtures with the minimum specific heat consumption are considered. It is shown that the efficiency of the rectification processes in the marginal productivity mode monotonously depends on the reversible efficiency, which makes it possible to ignore irreversible factors for choosing the order of separation in this mode.


  • entropy production
  • conditions of minimal dissipation
  • optimal heat transfer
  • multithreaded heat exchange system
  • rectification
  • separation of multicomponent mixtures
  • boundary of the realizability of thermal machines

1. Problems and methodology of finite-time thermodynamics

Applied thermodynamics originates from the work of Sadi Carnot in 1824 [1]. One of the problems of thermodynamics is the study of problems on the limiting possibilities of thermodynamic systems. For a long time, these tasks boiled down to finding the maximum efficiency of heat and refrigeration machines, separation systems, and various chemical processes. The solution of these problems led to the fact that the maximum efficiency value was determined in the case when the process under study was reversible. Reversibility will include processes in which the coefficients of heat and mass transfer are arbitrarily large or the fluxes of energy and matter in the system under study are arbitrarily small. With the development of nuclear energy, a new task was set—to obtain such a cycle of a heat engine that would correspond to its maximum power with certain fixed exchange ratios with sources. This task is due to the fact that the capital expenditures for the construction of nuclear power facilities are high with a relatively low cost of fuel spent. Variants of solving the problem of optimization thermodynamics were proposed in [2, 3].

Further development of finite-time thermodynamics was stimulated by a great deal of work of very many investigators. Here, we list names of just a few first researchers: R.S. Berry, B. Andresen, K.H. Hoffmann, P. Salamon, L.I. Rozonoer, and some others (see [1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32]).

Typical problems of optimization thermodynamics include the following: processes with minimal irreversibility; determination of the limiting possibilities of heat engines, cold cycles, and heat pumps (maximum power, maximum efficiency, many realizable modes); and analysis of the processes of separation of mixtures.

The general approach to solving problems is as follows. It is assumed that the whole system is divided into subsystems. In each subsystem, at any time moment, the deviations of the intensive variables from their average values over the volume are negligible. Consequently, the change of these variables (temperatures, pressures, etc.) occurs only at the boundaries of the subsystems, which means that the system as a whole is in a nonequilibrium state. This assumption makes it possible to use the equation of state in the description of individual subsystems, which are valid under equilibrium conditions, and ordinary differential equations can be used to describe the dynamics of the subsystems. The solution of extremal problems in this case is performed by methods of the optimal control theory for lumped parameter systems.

To study the limiting possibilities of thermodynamic systems, it is first necessary to make balance relations for matter, energy, and entropy. Moreover, the balance ratio for entropy includes dissipation σ, that is, the production of entropy. It characterizes the irreversibility of processes in the system. If all processes are reversible, then the dissipation is zero. If the processes are irreversible, then dissipation takes positive values. Dissipation depends on the dynamics of the processes. The set of realizability of the system in the parameter space of input and output streams is determined by the nonnegativity of dissipation. Reversible processes lie on the boundary of this set.

When a minimum possible dissipation is found as a function of flux intensities, then the inequality σσminholds in an arbitrary real system; this contracts the region of realization. In this formulation, the set obtained incorporates the effects of process dynamics, the magnitude of fluxes, and the system’s extent (due to the presence of heat and mass transfer coefficients).

In any real system, it is possible to narrow the realizability region if we find the minimum possible dissipation value as a function of the flux intensity (σσmin). This will take into account the dynamics, the flux intensity, and even the size of the installation through the coefficients of heat and mass transfer.

Then, from the balance equations, it is necessary to derive the connection between the system performance indicators and dissipation σ. Performance indicators usually monotonically deteriorate with the increase of σ. The best values of efficiency indicators are achieved in a reversible process, which allows using them similarly to the Carnot efficiency indicators.

Next, it is necessary to solve the problem of the organization of processes in such a way that, with the given constraints, the dissipation as a function of the flux intensities is minimal. This is the most difficult step in analyzing the capabilities of thermodynamic systems.

Consider the process of studying the limiting possibilities in more detail, and begin with thermodynamic balances. Thermodynamic balances show the relationship between the fluxes (matter, energy, and entropy) that the system exchanges with the environment and the changes in these values in the system [19]. Let us summarize all the fluxes, considering incoming fluxes as positive and outcoming fluxes as negative. Fluxes can be convective and diffusive. Convective fluxes are forced into the system and removed from it. The diffusive flux depends on the differences between the intensive variables of the system at the point where it enters and the intensive variables of the environment.

The energy balance shows the rate of change in the energy of a system, which is determined by the flux of energy that enters or is removed along with the convective fluxes of matter, the change in energy due to the diffusional exchange of matter, the currents of conductively transmitted heat, and the power of the work done. Material balance shows the change in the number of moles of substances in the system. Entropy balance shows the change in the entropy of the system, which occurs due to the influx of entropy together with the incoming substances, the influx or removal of heat, and the production of entropy due to the irreversibility of exchange processes.

If the system operates cyclically, the balances can be recorded on average for the equipment working cycle. In this case, the total change in energy, amount of matter, and entropy per cycle is zero, since the state of the system at the start and the end of the cycle is the same. Balances are transformed into a system of relations of averages over cycle-averaged components.

The equations of thermodynamic balances show the relationship between process efficiency indicators, external fluxes, and the structure of the system. The increase in entropy σcauses an increase in the entropy of output fluxes. At the same time, either the temperature of the fluxes at the outlet decreases, or the outlet flux of heat increases at a constant temperature. This leads to a reduction in the work of separation, the mechanical work produced by the system.

Consider the operation of a thermal machine that converts the heat received from a hot source with temperature T+into work. The working fluid gives a part of the energy to a cold source with a temperature T. The working fluid changes its state cyclically. As an indicator of efficiency, we will consider the thermal efficiency (η=p/q+)—the ratio of the work produced to the amount of heat collected from a hot source.

Let us denote the average intensity of the heat flux taken from the hot source q+and that given to the cold source, q. For the generated power p, we write the equations of energy balances:




Since the state of the working fluid either does not change in time (for steam and gas turbines) or changes cyclically (for steam engines), then there are zeros in the right parts of the equations.

Thermal efficiency η=p/q+follows from Eq. (1) in the form


Taking into account the fact that the Eq. (2) implies




Thermal efficiency ηis equal to Carnot efficiency when any irreversible phenomena are absent in the system.

The growth of σleads to a growth entropy of output streams; under other equal conditions, this growth reduces the flux temperature at the outlet or at a fixed temperature increases the waste heat flux. And in this and in another case, this leads to a decrease in the mechanical work produced by the system or the work of separation. Energy efficiency of thermodynamic system, characterized by the relation of useful work, produced in it, to the energy costs, reaches a maximum in the invertible processes, when σ=0.


2. Processes with a minimal dissipation

It is known that the maximum efficiency of conversion of thermal energy into mechanical work or separation work is achieved in reversible processes. If the intensity of the target flux is set, the processes in the thermodynamic system are irreversible. In this case, the role of reversible processes is played by the processes of minimal dissipation, so it is necessary to determine conditions under which thermodynamic processes exhibit minimal dissipation for a prescribed average intensity (prescribed averaged value of driving forces).

2.1 The minimal dissipation’s conditions

Consider two systems interacting with each other. Intensive variables for the ith system will be denoted by uiand extensive variables by xi. In general, these are vector variables. When systems are in contact, the difference between u1and u2leads to the appearance of flux Ju1u2. Function Jis continuous, is differentiable, and has the following properties:


for scalar u1and u2.

The difference between vectors u1and u2(of the same sign as flux Jj) leads to appearance of driving forces Xj. Each force defines exclusively u1jand u2j, satisfying conditions analogous to those in Eq. (6). Entropy production σ, which characterizes the process irreversibility, is equal to the scalar product of the flux vector and the driving force vector, and average value of entropy is described by the formula


where the independent variable lhas the interpretation of a space-time or a contacting area measure. The integrand of this functional is defined non-negatively.

We shall assume that in our algorithm (at least) one intensive variable appears, by definition u2l, which may assume its values from within a certain manifold V. Yet, because of the variability of extensive variables of the first subsystem dY1jdl=Jju1u2, the second variable changes in accordance with the formula


Average values of all or some selected fluxes are prescribed:


Further on, we consider only the case of a scalar flux. The problem for vector fluxes and its solution is considered with details in [32, 33].

The scalar flux problem involves minimizing of the integral


subject to constraining conditions:


The problem (10)(12) simplifies in an important case when the rate of change of variable u1is proportional to the flux:


In this case the condition of minimal dissipation assumes the form


whereas the condition of prescribed flux intensity can be written as


The value of u1Lis determined regardless of the optimal solution u2u1.

If the flux is proportional to the driving force with constant coefficient α, then the minimum entropy production equals


2.2 Minimal dissipation’s conditions of selected processes

Consider the conditions for the minimum dissipation of heat exchange. Let us take the temperature of the body being heated as the controlling intense variable. The driving force in the minimum dissipation problem is


whereas the heat flux is qT1T2. In the majority of cases, we may assume the energy balance in the form


where c1T1is the heat capacity of the hot source.

If the process takes place in time, then the parameter lhas the meaning of time, and the parameter L—the duration of the process. If a pipe heat exchanger is considered, in which the hot flux temperature changes from section to section, the value of cis the water equivalent of the flux, and Lis the length of the heat exchanger.

In agreement with conditions (14), (15) describing the minimum dissipation subject a prescribed average intensity of heat flux q¯, we can obtain a condition of minimum dissipation for an arbitrary law of heat transfer:


The first of these conditions determines T2T1λ2, second—T1L, and third—constant λ2.

For the Newtonian law of heat transfer


with a constant heat capacity (water equivalent) c, we obtain from conditions (18)(20)


Therefore, for an arbitrary lof an optimal process, the ratio T1T2should be constant. This constant equals


As it follows from Eq. (20), T1L=T10q¯L/c. Finally, the condition (20) leads to the following equality:


Substituting Eqs. (23) and (24) into the expression


minimal entropy production is obtained in the form


Table 1 presents analogous conditions of minimal dissipation for some well-known processes and corresponding expressions for minimal entropy production.

ProcessConditions of minimal dissipation and entropy production
Heat transfer q=αT2T1T1lT2l=1βαL; σmin=β2αLβ
Vector flux, linearly depending on driving forces J=LXX=const. J=J¯; σmin=J¯TL1J¯
One-sided isothermal mass transfer gc1c2=kc1lc2lc2l=c1l+m2c1lm+m4
c1Lc10GdC1k1c12c1m+m4m2=L; σmin=c1Lc10RG1c12lnc1dc1c1+m2c1m+m24
Two-sided isothermal equimolar mass transfergc1/gc2=mc2l1c2lc1l1c1l
c10=C10,c1L=c1L; σmin=R0Lgc1c2lnc11c2c21c1dl

Table 1.

Conditions of minimal dissipation in thermodynamic processes.

As shown in [34], the conditions of minimal dissipation make it significantly easier to estimate the limiting possibilities of thermodynamic systems. In a system with multithreaded heat exchange [35], the total heat load q and the total heat transfer coefficient α¯are fixed. At the input of the system, kheating fluxes with temperatures Ti0and water equivalents Wicome in. It is necessary to choose the parameters of heat fluxes, the structure of the system, and the distribution of heat transfer coefficients.

The conditions under which the minimum possible production of the entropy of the σ¯trait is reached are also defined in [34]: (1) At each point of contact of the heating and heated streams, the minimum dissipation conditions must be satisfied. (2) Temperatures of heating fluxes at the outlet of the system should be equal to each other, as well as the temperature of the heated fluxes at the outlet. (3) Heating fluxes, in which the inlet temperature is less than the calculated temperatureT¯, do not participate in heat exchange.

Computational relations for Newtonian heat transfer are


The system in which the entropy production calculated with parameters of all fluxes


is lower than a certain value cannot exist in reality.

Analogous relations can easily be obtained in the case when the inlet parameters of heated fluxes are prescribed.


3. Synthesis of heat exchange systems

In [36] the problem of the limiting possibilities of the heat exchange system (“ideal” heat exchange) was considered. The minimum possible entropy production σwas found in the system with the given values of water equivalents and input temperatures of hot or cold fluxes and given the total heat load and the total heat transfer coefficient. It is shown that for the case when the heat flux is proportional to the temperature difference (Newtonian dynamics), this irreversibility limit can be reached if at each point of contact the ratio of the absolute temperatures of the fluxes is the same, and their temperatures at the outlet of the system are the same for all fluxes whose input temperatures are fixed (hot or cold).

Conditions of ideal heat exchange impose very strict requirements on the characteristics of the system:

—Each double-flux cell must be a counter-flux heat exchanger.

—The ratio of the water equivalents of the hot and cold flux in it should be equal to the ratio in degrees Kelvin of the temperature of the cold flux at the outlet of the heat exchange cell to the temperature of the hot flux at its inlet—conditions of thermodynamic consistency.

—This ratio and its corresponding minimum possible entropy production at fixed temperatures and water equivalents of hot fluxes are related to their inlet temperatures Ti0, the water equivalents Wi, and the total heat transfer coefficient Kas:


—The temperature of the hot streams at the outlet should be the same and, as it follows from the conditions of the energy balance, is equal to:


—Hot fluxes with initial temperatures less than T¯+do not participate in the heat exchange system.

If a part of the hot fluxes condenses in the process of heat transfer, then in the expression for m(Eq. (24)), the water equivalent of the corresponding term tends to infinity. Assign the index kto the condensing fluxes and find the limit.

WklnTk0lnT¯+=WklnTk0lnTk0qkWkwhen Wktends to infinity. Using L’Hospital’s rule to disclose the uncertainties, we find that


Here, it is taken into account that the temperature Tk0is equal to the condensation temperature and the thermal load is the product of the flux rate of the latent heat of vaporization.

Thus, the expression for min the presence of condensing fluxes will be rewritten in the form:


In a multithreaded system integrated with the technological process, the values of water equivalents of both hot and cold fluxes are set, and often their outlet temperatures are set. Therefore, the performance of the ideal heat exchange system cannot be achieved. It is natural to set the task of synthesis of the heat exchange system of the minimum irreversibility at more rigid restrictions on characteristics of streams. The conditions of ideal heat transfer can only serve as a “guiding star” like Carnot’s efficiency for thermal machines, and the value of the ratio σto the real production of entropy in the designed system is an indicator of its thermodynamic perfection.

Next, we propose the calculated relations for the bottom estimate of the minimum dissipation in the system with the above restrictions and the synthesis of a hypothetical system in which such an estimate is implemented.

Consider a multithreaded heat exchange system containing a set of hot (index i) and cold fluxes (index j), with given water equivalents Wi,Wj. For each of the cold (heated) fluxes, its inlet and outlet temperatures are set to Tj0and T¯j>Tj0.

For hot (cooled) fluxes, except for water equivalents, their temperatures at the inlet to the heat exchanger Ti0are set. If some flux in the system changes its phase state, then for it except water equivalents, the flux rate gi,gjand heat of vaporization (condensation) ri,rjare fixed. The ambient temperature will be denoted as T0.

Under these conditions, the thermal load of the system is equal to the total energy required for heating all cold fluxes and is determined by the equality:


The difference in the conditions imposed on the hot and cold fluxes is due to the fact that for cold fluxes leaving the system with a temperature less than a predetermined one, heating is required, i.e., additional energy costs, and for hot ones, if their outlet temperature is greater than a predetermined one, cooling is required, which is much easier.

Entropy production is the difference between the total entropy of outgoing fluxes and the total entropy of incoming fluxes. Initially, we assume that all fluxes enter and leave the system in the same phase state, the pressure change in the system is small, and the heat capacity is constant. Then, the change in the entropy of each flux is the product of its water equivalent by the logarithm of the ratio of its inlet and outlet temperatures in degrees Kelvin [37]. So, it follows from the conditions of the thermodynamic entropy balance that:


The first of these terms is negative, the second is positive, and their sum is always greater than σ>0.

Note that all variables determining the value of the entropy growth of cold fluxes are given by the conditions of the problem, so that the minimum entropy production corresponds to the minimum at a given thermal load of the first summand by temperatures T¯i.

The formal statement will take the form:


The Lagrangian of this problem


The conditions of its stationarity in T¯ilead to equality:


Thus, in any water equivalents and input temperatures of hot streams, minimum dissipation corresponds to such an organization of heat transfer for which the temperatures of hot streams at the exit are the same.

In general, coolant fluxes at the system inlet can have different phase states: vapor, liquid, or vapor-liquid mixture. The same states can be at the output of the stream.

—If the flux does not change its phase state, but changes only the temperature, then we assume that its temperature at the input to the cell Tk0, the water equivalent Wk, and for cold fluxes the temperature at the output T¯kare known. The temperature of the hot streams at the output of the T¯+system is selectable (see Eq. (30)).

—If the cold flux changes its phase state so that at the inlet it is a liquid at boiling point and at the outlet it is saturated with steam (let us define it as “evaporating”), the weight flux rate gj, the boiling point Tbj, and the heat of vaporization rjare given. The same is true for hot “condensing” streams. They have a saturated steam state at the inlet and a liquid state at the boiling point at the outlet.

Thus, the first step in the synthesis algorithm of heat exchange systems is the preparation of initial data, in which actual fluxes and their characteristics are converted into calculated fluxes. They can be of two types: those that do not change their phase state (heated and cooled) and those that change it at the boiling point (evaporating and condensing). End-to-end fluxes are not included in the calculation. To calculate the total heat load production, use the following expression:


Minimum dissipation implies fulfillment of the “counterflow principle”: the cold streams with higher temperatures must be in contact with the hot flux with a higher temperature. The latter requirement, as well as the equality of temperatures of hot streams at the outlet, corresponds to the conditions of the ideal heat transfer [36].

As the hot fluxes move from one contact cell to the next, their temperature changes due to the recoil of the heat flux. At the output of the system, the heat flux given by them is q¯, and the temperature is T¯+. Let us denote by qthe given heat load in some intermediate state of hot fluxes. As the hot streams cool down, it changes from zero to q¯.

In this case, we assume that when the hot flux with the highest input temperature (first) is cooled to a temperature of T20, the first and second fluxes are combined, so that their water equivalents are summed. A similar union occurs with the third flux, etc., until the temperature of the equivalent hot flux drops to the previously calculated formula (30) T¯+. If a condensing flux is at a certain temperature Tbiin the number of hot fluxes, the temperature of the equivalent flux is constant and equal to Tbiuntil the equivalent flux transfers the heat of condensation qbi=gbiri. The dependence of the temperature of the hot flux equivalent on the given heat load T+q, we will call the contact temperature of the hot fluxes.

Cold fluxes are ordered by their outlet temperature, so that j=1corresponds to the flux with the highest output temperature. For cold fluxes, the value of qis equal to the current required heat load, i.e., the heat they need to obtain to satisfy the conditions imposed on their temperature and output state. The greater the q, the lower the cold flux temperature corresponding to this value. In this case, we assume that when the temperature of the first flux decreases, two events are possible:

  1. Its temperature will drop to the set temperature at the output of the second stream.

  2. Its temperature will drop to its initial temperature.

In the first case, the first cold flux is calculated combined with the second. In the second case, it is excluded from the system and transferred to the heating of the second stream. This procedure continues until an equivalent cold flux reaches the lowest cold flux temperature at the system inlet. The number of threads included in the equivalent cold flux is changed by adding fluxes with lower temperatures at the outlet and due to the exclusion from streams with the highest temperatures at the entrance. But each value of qcorresponds to the value of Tqof the contact temperature of cold fluxes.

The dependencies of the current contact temperatures can be calculated from energy balance conditions similar to the expression (25). For equivalent hot flux:


where S+T+is the set of indices of hot fluxes for which the inlet temperature is greater than the current contact temperature (Ti0>T+q).

Similarly, for the contact temperature of the equivalent cold flux, we have:


where STis the set of indices of cold fluxes for which the contact temperature Tsatisfies the inequality T¯j>Tq>Tj0.

The curves of the current contact temperatures decrease monotonically with the growth of q, with T+q>Tq. On each of these curves, the points (nodes) are selected, in which either the composition of the fluxes entering the equivalent flux changes or the condensation/evaporation process takes place. In the latter case, horizontal sections appear on the curves. On the curve Tq, there can be vertical jumps if the flux temperature Tj0>T¯j1.

The interval δqνfrom one of the nodes on any of the contact curves to the nearest node on the same or another curve is characterized by the same composition and phase state of the contacting fluxes. We will call it the homogeneity interval.

For each such interval of δqν, three combinations of contacting fluxes are possible:

  1. Both equivalent fluxes change their phase states.

  2. The hot equivalent flux is cooled and the cold is heated.

  3. One of the fluxes changes its phase state, and the other is cooled or heated.

Contact temperature curves provide all the data necessary to calculate the heat transfer coefficient of the cell in which the contact is made:

—Water equivalents of W+,Wis equal to the sum of the equivalents of water fluxes which are part of the equivalent contacting fluxes.

—Temperatures of equivalent fluxes at the inlet and outlet of the interval of homogeneity is known.

—The thermal load of such a computational cell is δqν.

Depending on which of these contact combinations is implemented, it is possible to select the type of cell hydrodynamics and find Kν. Finding the heat transfer coefficient Kνfor each νinterval of homogeneity and summing these coefficients over all intervals, we obtain the total coefficient K, which can be achieved by organizing the countercurrent heat exchange of equivalent fluxes. In turn, knowledge of the heat load q¯and the total heat transfer coefficient allows us to calculate the minimum possible entropy production σby the formula (29) and estimate the degree of thermodynamic perfection of the constructed system as η=σσ0, where σ0is total entropy production in the system.


4. The region of realizability of systems

An irreversible factor affecting machine power or pump performance is finite heat transfer coefficients αibetween heat sources and the working fluid. We substitute the minimum possible entropy production into Eq. (5) and obtain the condition determining the maximum power:


As σminq+increases faster than q+, there exist a maximum of power in the region of realizability.

To minimize the production of entropy, it is necessary that with each contact of the working medium with the sources the conditions of minimum dissipation, which depend on the dynamics of heat transfer, are met. For a source of infinite capacity and the temperature of the working fluid in contact with, it should be constant. For Newtonian dynamics, the ratio of working fluid temperature and sources should have been constant. So, if the temperature of the source changes due to the final capacity, then the temperature of the working fluid should change, remaining proportional to the temperature of the source.

For sources of infinite capacity, the optimal cycle of a heat machine with maximum power for any heat transfer dynamics should consist of two isotherms and two adiabats, and it turned out that the efficiency corresponding to the maximum power (it is called the Novikov-Curzon-Ahlborn, ηnca) is only a function of the Carnot efficiency:


The maximum difference between ηCand ηncais achieved when the ratio of absolute temperature of the hot and cold sources is 0.25, when ηC=0.75.

For power that is less than the maximum possible, the maximum efficiency of the heat machine is equal to


In this case, αis expressed as


As p0, the efficiency ηmaxtends to Carnot efficiency and as


Corresponding thermal efficiency approaches the efficiency value obtained by Novikov, Curzon, and Ahlborn (42).

The nature of the set of realization modes is shown in Figure 1.

Figure 1.

Power of thermal machine as a function of driving heat flux.

Similar results can be obtained for the heat pumps. Since the flux of costs is mechanical energy, the set of realizable modes has the form of a convex upward and unbounded parabola.


5. Rectification processes

In the separation process, energy is spending on getting the work of separation. The work of separation can be obtained as an increase in the free energy of the streams leaving the system compared to the energy of the mixture flux at the system inlet. The energy expended can be thermal or mechanical. In systems of separation with thermal energy, the set of realizable modes coincides in the form with heat engines. In this case, the rectification processes will be the most important and energy-intensive. In the section below, the process of thermal separation of a two-component mixture is considered, and considerations which allow one to proceed to the determination of the order of separation of multicomponent mixtures are obtained.

Let the following parameters be defined for a mixture of two components: qi,Ti, si,pi, hi,xi,μi—molar consumption, temperature, molar entropy, pressure, enthalpy, concentration of key component, and its chemical potential in ith stream. Assume the index i=0for the separated stream, index i=1for the stream of the enriched key component (for which x1>x0), and index i=2for the stream cleared of the key component x2<x0. Heat flux q++, brought (coming) from a source of temperature T+, is supplied to the separation system, whereas heat flux qis rejected to a source of temperature T. Equations of thermodynamic balances are of the form:


The ratio of target mass flux g1and heat flux q+may be accepted as the thermal efficiency of the separation process:


Using material balances of Eq. (46), we shall express g2in terms of g1and introduce coefficient a=x1x0/x0x2. Then, the second flux satisfies g2=ag1. Eq. (46) assumes the forms


Here, Δs01=s0s1,Δs02=s0s2are the entropy increases, whereas Δh01=h0h1,Δh02=h0h2are enthalpy increases in corresponding streams. It is mandatory that the concentrations of the key component in streams are prescribed.

We transform Eq. (48) to the form q=q++g1Δh01+aΔh02and substitute the expression obtained into Eq. (49). The frontier of the realization set is characterized by the following equation:


Here, F=TΔs01+aΔs02Δh01ah02. The increases of enthalpy and entropy contained in Fhave the forms


The entropy of mixing per one mole of mixture is:


Note that the ratio T-/F depends on reversible factors only. In the reversible process, the entropy production σis zero, and the thermal efficiency reaches a maximum equal to the multiplier at the heat flux in Eq. (42).

As a productivity you can take any of the streams, even the stream of a separated mixture, because with given compositions of the streams they are proportional.

A reversible estimate of the thermal efficiency of the separation process and the shape of the border of the realizability region can be clarified by finding the minimum possible for a given productivity and dynamics of heat and mass transfer value σand its dependence on the coefficients of dynamics and heat flux.

If the dynamics of heat transfer can be approximated by the Fourier law and the mass transfer flux is proportional to the difference of chemical potentials, then the minimum dissipation is proportional to the square of the cost of heat. The boundary of the set of realizable modes in this case has a parabolic form


Then, the efficiency of a separation column in the maximum productivity mode is equal to one half of the reversible efficiency:


Qualitative expressions linking characteristic coefficients aand bwith parameters of the separation column were obtained [38]:


Here, AGis the molar reversible work of mixture separation, equal to the difference between molar free energy of streams leaving the column and the free energy of raw stream, TD,TBare the temperatures in the condenser and the kettle of the column, ris the molar evaporation heat, βD,βBare the coefficients of heat exchange in the condenser and the kettle, and kis the effective coefficient of mass transfer for column height.

The coefficients aand bcan be found not only by Eq. (55) but also by the results of measurements on the current column. This allows us to solve many problems associated with finding a set of realizability, including the problem of choosing the order of separation. It is important that the efficiency in maximum performance mode depends only on the reversible efficiency. The following condition, sufficient for to be independent of a, is valid.

The sufficiency condition for independence ofηofa[34]:

If the partial derivativeσabqqdepends continuously on some scalar functionzaqand the ratioσabqqis a function ofz, then the thermal efficiency in the maximum productivity mode is defined exclusively in terms of variablesb, characterizing the reversible process. The condition is satisfied for thermal machines and for binary rectification.

In Figure 2 shows an example of the boundaries of realizable sets in cases where σabqdoes not depend on a.

With decreasing dynamic coefficients, the entropy production increases. The set of realizable modes is compressed, while the maximum performance points with a corresponding heat flux remain on a straight line with a slope of η.

Figure 2.

Nature of change of the realization frontier with the irreversibility increase.

5.1 Order of separation: rule of temperature multipliers

We arrange the substances according to the property γused for separation (boiling point in the rectification processes). We normalize γso that it is in the range from 0 to 1. The order of separation of substances may be direct or reverse. In the case of a direct order, the stream with the components γ<γ1is first separated, and then, the stream with large values of γis divided into two sub-streams so that the first one has γ1γ<γ2and the second one γγ2. In the case of reverse order, the stream with the components γ>γ2is first separated, and then, the stream with lower values of γis divided into two sub-streams so that the first one has γ1γ<γ2and the second one γγ1.

Let Adenote the work of separation of the mixture. The Aincludes the factor RT0, where Ris the universal gas constant and T0is the ambient temperature. Then, the entropy of the separated fluxes will decrease in proportion to the work of separation A. In a reversible case, a decrease in the entropy of the material flux should be compensated by an increase in the entropy of the heat flux. If T+and Tare the temperatures of heat supply and removal, then the increase in entropy of the heat flux can be expressed as δsq=q1/T1/T+. The total cost of heat depends on the order of separation, while the total work does not depend on the order of separation. At each separation stage, the heat flux is proportional to the separation work at this stage, with the proportionality multiplier KT=T+TT+T. Temperature factors are determined by the choice of the separation boundary. Denote by KT1and KT2the temperature factors corresponding to the separation boundary γ1and γ2, respectively, and by A21and A22the work of separation of the mixture in the second stage in the direct and inverse order of separation. The total cost of heat per mole of the input mixture in the direct and reverse order of separation can be written as:


To determine the separation order, it is necessary to calculate the difference:


If the result of the calculation Eq. (57) is negative, then it is reasonable to choose a direct separation order. If the result is positive—a reverse order.

In the case of a multistage system, this rule applies to each of two successive stages. It is easy to see that the expression in square brackets in Eq. (57) is non-negative. From here follows the rule of temperature multipliers (see [39]): The separation boundaries must be chosen so that the temperature multipliers do not decrease from stage to stage. In the case when the separation efficiency in the maximum performance mode depends only on the reversible efficiency, the rule of temperature multipliers is also valid. It is important that the information that is needed to calculate temperature factors is much more accessible and accurate than the information on the dynamics of the processes in the column.


6. Conclusions

This chapter discusses the problems of optimization of thermodynamics and methods of analysis of systems and describes the types of thermodynamic balances, the relationship between the performance of the process, and the production of entropy. Also, it is shown that in the absence of irreversibility, the thermal efficiency is equal to the Carnot efficiency.

The conditions are found under which the thermodynamic processes at a given average intensity have minimal dissipation, expressions for determining the minimum dissipation and entropy with the Newtonian heat transfer law are obtained, and expressions for the cases of vector flux, one-sided isothermal, and two-sided equimolar mass transfer are given.

The synthesis algorithm makes it possible to build heat exchange systems with minimal irreversibility, in which restrictions on water equivalents, temperatures, and phase states of the flows are fulfilled, which imply combining the fluxes into two equivalent ones. The nature of the set of realizable modes of heat engines and pumps is described. It is shown that the efficiency corresponding to the maximum power mode does not depend on heat transfer coefficients, but is only a function of the Carnot efficiency.

Separation processes are considered, and estimates of the thermal efficiency of the separation process and the shape of the realizable area boundary are obtained for them. It is shown that the efficiency in the mode of maximum performance depends only on the reversible efficiency. The rule of temperature multipliers is described, which allows to determine the separation order in multistage systems.


  1. 1. Carnot S. Reflections on the motive power of fire and on machines fitted to develop that power. In: Thurston RH, editor. Ecole Polytechnique. No. 55. Chez Bachelier, Libraire, Quai des Augustins: A Paris; 1824
  2. 2. Novikov II. The efficiency of atomic power stations. Atomnaya Energiya. 1957;3(11):409; English translation in Journal of Nuclear Energy. Part B. 1958;7(2):25-128
  3. 3. Reitlinger HB. Sur l’utilisation de la chaleur dans les machines a feu. Liege: Vaillant-Carmanne; 1929
  4. 4. Andresen B, Salamon P, Berry RS. Thermodynamics in finite time: Extremals for imperfect heat engines. The Journal of Chemical Physics. 1977;66(4):1571-1577
  5. 5. Andresen B. Finite-Time Thermodynamics. Copenhagen: University of Copenhagen; 1983
  6. 6. Andresen B. Finite time thermodynamics. Current trends in finite time thermodynamics. Angewandte Chemie, International Edition. 2011;50:2690-2704
  7. 7. Bejan A. Entropy generation minimization: The new thermodynamics of finite size devices and finite time process. Journal of Applied Physics. 1996;79:1191-1218
  8. 8. Berry RS, Kazakov VA, Sieniutycz S, Szwast Z, Tsirlin AM. Thermodynamic Optimization of Finite Time Processes. Chichester: Wiley; 1999
  9. 9. Chen J, Yan Z, Lin G, Andresen B. On the Curzon-Ahlborn efficiency and its connection with the efficiencies of real heat engines. Energy Conversion and Management. 2001;42:173-181
  10. 10. Chen L, Wu C, Sun F. Finite time thermodynamic optimization or entropy generation minimization of energy systems. Journal of Non-Equilibrium Thermodynamics. 1999;24:327-359
  11. 11. Curzon FL, Ahlburn B. Efficiency of a Carnot engine at maximum power output. American Journal of Physics. 1975;43:22-24
  12. 12. Hoffman KH, Watowich SJ, Berry RS. Optimal paths for thermodynamic systems: the ideal Diesel cycle. Journal of Applied Physics. 1985;58(6):2125-2134
  13. 13. Linezky SB, Rodnjansky LE, Tsirlin AM. Optimal cycles of chillers and heat pumps. Izvestiya An SSSR—Energetika i Transport. 1985;5(6):42-49
  14. 14. Mironova VA, Amelkin SA, Tsirlin AM. Mathematical Methods of Thermodynamics with a Finite Time. M: Ximija; 2000
  15. 15. Moloshnikov BE, Tsirlin AM. The thermodynamically optimal concentration profiles in the problems of isothermal irreversible mass transfer. Theoretical Foundations of Chemical Engineering. 1990;(24):129-137
  16. 16. Mironova V, Tsirlin A, Kazakov V, Berry RS. Finite-time thermodynamics: Exergy and optimization of time-constrained processes. Journal of Applied Physics. 1994;(76):629-636
  17. 17. Ondrechen MJ, Andresen B, Mozurkewich M, Berry RS. Maximum work from a finite reservoir by sequential Carnot cycles. The American Journal of Physiology. 1981;49:681
  18. 18. Ondrechen MJ, Berry RS, Andresen B. Thermodynamics in finite time: A chemically driven engine. The Journal of Chemical Physics. 1980;72(9):5118-5124
  19. 19. Ondrechen MJ, Berry RS, Andresen B. Thermodynamics in finite time: Processes with temperature-dependent chemical reactions. The Journal of Chemical Physics. 1980;73(11):5838-5843
  20. 20. Orlov VA, Pozonoer LI. Estimates of the efficiency of the thermodynamic processes of controlled substances based on the energy balance equations and entropy. In: X-Union Conference on Control. M.: Nauka; 1986
  21. 21. Rozonoer LI, Tsirlin AM. Optimal control of thermodynamic systems. Automation and Remote Control. 1983;44(1):70-79;44(2):88-101;44(3):50-64
  22. 22. Salamon P, Nitzan A. Finite time optimizations of a Newton’s law Carnot cycle. The Journal of Chemical Physics. 1981;74(6):3546-3560
  23. 23. Salamon P, Band YB, Kafri O. Maximum power from a cycling working fluid. Journal of Applied Physics. 1982;53(1):197-202
  24. 24. Salamon P, Hoffman KH, Schubert S, Berry RS, Andresen B. What conditions make minimum entropy production equivalent to maximum power production? Journal of Non-Equilibrium Thermodynamics. 2001;26(1):73-84
  25. 25. Salamon P, Nitzan A, Andresen B, Berry RS. Minimum entropy production and the optimization of heat engines. Physical Review A. 1980;21:2115-2129
  26. 26. Salamon P, Nulton JD, Siragusa G, Andresen TR, Limon A. Principles of control thermodynamics. The International Journal of Energy Research. 2001;26(3):307-319
  27. 27. Salamon P. Physics versus engineering of finite-time thermodynamic models and optimizations. In: Bejan A, Mamut E, editors. Thermodynamic Optimization of Complex Energy Systems. Dordrecht, The Netherlands: Kluwer Academic Publishers; 1999. pp. 421-424
  28. 28. Salamon P, Hoffman K-H, Tsirlin AM. Optimal control in a quantum cooling problem. Applied Mathematics Letters. 2012;(25):1263-1266
  29. 29. Salamon P, Hoffman K-H, Rezek K-H Y. Maximum work in minimum time from a conservative quantum system. Chemical Physics. 2009;(11):1027-1032
  30. 30. Sieniutycz S, Jezowski J. Energy Optimization in Process Systems and Fuel Cells. Oxford: Elsevier; 2013
  31. 31. Sieniutycz S. Thermodynamic Approaches in Engineering Systems. Oxford: Elsevier; 2016
  32. 32. Spirkl W, Ries H. Optimal finite-time endoreversible processes Physical Review E. 1995;52(4):3455-3459
  33. 33. Tsirlin AM, Mironova VA, Amelkin SA. Processes minimal dissipation. Theoretical Foundations of Chemical Engineering. 1997;31(6):649-658
  34. 34. Sieniutycz S, Tsirlin A. Finding limiting possibilities of thermodynamic systems by optimization. Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences. 2017;375(2088):20160219
  35. 35. Tsirlin AM, Ackremenkov AA. The optimal organization of heating and cooling systems. Theoretical Foundations of Chemical Engineering. 2012;46(1):109-114
  36. 36. Tsirlin AM. Ideal heat exchange systems. Journal of Engineering Physics and Thermophysics. 2017;90(5)
  37. 37. Kondepudi D, Prigogine I. Modern Thermodynamics. From Heat Engines to Dissipative Structures. John Wiley & Sons; 1998
  38. 38. Tsirlin AM, Sukin IA. Finite-time thermodynamics: The maximal productivity of binary distillation and selection of optimal separation sequence for an ideal ternary mixture. Journal of Non-Equilibrium Thermodynamics. 2014;39(1):13
  39. 39. Tsirlin AM, Balunov AI, Sukin IA. Estimates of energy consumption and selection of optimal distillation sequence for multicomponent distillation. Theoretical Foundations of Chemical Engineering.2016;50(3):250-259


  • †Ad memoriam Anatoly V. Zaev

Written By

Margarita Anatolyevna Zaeva and Anatoly Mikhailovich Tsirlin

Submitted: November 13th, 2018 Reviewed: January 24th, 2019 Published: March 27th, 2019