COURSE OF ENERGY CONVERSION SECOND-LAW ANALYSIS OF POWER CYCLES These class notes are for the students of the course "Energy conversion A" at Politecnico di Milano. Anyone who finds inaccuracies or, anyhow, wishes to send comments to improve them is invited to the lecturer (gianluca.valenti@polimi.it), who thanks in advance. Second-law analysis of power cycles - Energy Conversion A – V7.0 Why these classnotes ............................................................................................................................ 5 1 Introduction ................................................................................................................................... 6 2 Thermodynamics references ......................................................................................................... 7 2.1 Definitions............................................................................................................................. 7 2.2 First reference: second law of thermodynamics (Carnot's theorem) .................................... 7 2.3 Second reference: "ideal cycle” and “Reversible cycle” ...................................................... 8 2.3.1 Example 1: how can a Rankine cycle be made reversible? .............................................. 8 1.1.1.1 Saturated vapor cycle ................................................................................................ 8 1.1.1.2 Vapor cycle with superheating................................................................................ 10 2.3.2 Example 2: how can a Joule cycle be made reversible? ................................................. 12 2.3.3 Example 3: Stirling cycle ................................................................................................ 15 2.4 Third reference: reversible heat pump between sources/sinks at constant temperature: .... 16 3 Energy sources for power plants ................................................................................................. 18 3.1 Energy Sources Definition .................................................................................................. 18 3.2 Non-reacting flows, maximum work, exergy ..................................................................... 18 3.2.1 Trilateral and trapezoidal cycles ..................................................................................... 20 3.3 Fossil and renewable fuels .................................................................................................. 23 3.3.1 Stoichiometry in combustion reactions ........................................................................... 23 3.3.2 Enthalpies of formation and enthalpy balance of combustion ........................................ 26 3.3.3 Adiabatic flame temperature ........................................................................................... 28 3.3.4 Heating value of the fuel ................................................................................................. 29 3.3.5 Recuperative preheating.................................................................................................. 31 3.3.6 Entropic balance for a reacting system ........................................................................... 33 3.3.7 Irreversibility generated in combustion .......................................................................... 35 3.3.8 Reversible work and chemical exergy ............................................................................ 35 3.3.9 Comparison of indexes ................................................................................................... 37 4 Irreversible processes and loss of useful work............................................................................ 39 4.1 Irreversibilities and useful work losses ............................................................................... 39 4.1.1 First example of irreversibility: heat transfer .................................................................. 39 4.1.2 Second example of irreversibility: Isenthalpic throttling ................................................ 41 4.2 General demonstration ........................................................................................................ 44 4.2.1 Heat transfer: replacement with a reversible engine + heat pump .................................. 45 4.2.2 Throttling: replacement with reversible isothermal expansion + heat pump .................. 46 5 The most common causes of irreversibility ................................................................................ 49 5.1 Losses in Heat transfer ........................................................................................................ 49 5.1.1 General definitions .......................................................................................................... 49 5.1.2 Entropy generated in an irreversible heat exchange ....................................................... 52 5.1.3 Considerations on the design of the components ............................................................ 54 5.2 Losses in Expansion and Compression ............................................................................... 55 5.2.1 Gas/vapor turbomachines: turbines and compressors ..................................................... 55 5.2.2 Hydraulic machines: Pumps............................................................................................ 58 5.2.3 Considerations on the design of the components ............................................................ 58 5.3 Losses due to Pressure drops .............................................................................................. 59 5.3.1 Pressure drops of a liquid ................................................................................................ 59 5.3.2 Pressure drops of a gas .................................................................................................... 59 5.3.3 Considerations on the design of the components ............................................................ 60 5.4 Losses due to Mixing .......................................................................................................... 60 5.5 Losses due to Chemical reactions ....................................................................................... 61 5.6 Other losses: self-consumptions, auxiliary systems............................................................ 61 6 Applying the general formula to analysis of power plants ......................................................... 62 2 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 6.1 Definition of the time interval ............................................................................................. 62 6.2 Definition of the boundaries of the system considered ....................................................... 62 6.3 Definition of the energy source and of material flows........................................................ 62 6.4 The "dead" state .................................................................................................................. 63 6.5 Definition of the efficiency of the power cycle .................................................................. 63 6.6 Number N of modeled processes ........................................................................................ 64 7 Second law analysis of a steam rankine cycle ............................................................................ 65 7.1 Initial definitions ................................................................................................................. 65 7.2 βπΌπ irreversibility of heat transfer in the condenser .......................................................... 66 7.3 βπΌπ fluid dynamic irreversibility in the pumps .................................................................. 70 7.4 βπΌπ irreversibility of heat transfer in the preheating line ................................................... 72 7.4.1 Indirect preheaters ........................................................................................................... 72 7.4.2 Deaerator ......................................................................................................................... 74 7.5 βπΌπ irreversibility of heat transfer in introducing heat into the cycle ................................ 75 7.5.1 βπΌππ Derating the thermal energy from π»∞ to π»πππ ................................................. 77 7.5.2 βπΌππ Introducing heat into the cycle starting from π»πππ of the working fluid .......... 80 7.5.3 Binary cycles ................................................................................................................... 82 7.6 βπΌπ fluid dynamic irreversibility in the turbine ................................................................. 83 7.7 βπΌπ pressure drops in the liquid phase ............................................................................... 83 7.8 βπΌπ pressure drops in the vapor phase ............................................................................... 83 7.9 βπΌπ thermal losses .............................................................................................................. 84 7.10 βπΌπ mechanical/electrical losses ........................................................................................ 84 7.11 βπΌππ auxiliary losses ......................................................................................................... 84 7.12 Comparison of losses .......................................................................................................... 84 7.13 Other numerical examples .................................................................................................. 86 8 Effect of the thermodynamic properties of the working fluid on Rankine cycle performance .. 87 8.1 Sources at constant temperature - Non-regenerative cycle ................................................. 87 8.1.1 Effect of the molecular complexity ................................................................................. 88 8.2 Sources at constant temperature - Regenerative cycle ........................................................ 90 8.3 Sources at variable temperature .......................................................................................... 91 8.3.1 Simple molecules ............................................................................................................ 91 8.3.2 Complex molecules ......................................................................................................... 93 8.4 General conclusions ............................................................................................................ 95 8.4.1 Economics ....................................................................................................................... 95 8.4.2 Considerations concerning the choice of the working fluid ........................................... 96 8.4.3 Mixtures of fluids ............................................................................................................ 99 9 Second law analysis of an open-loop Brayton GAS cycle........................................................ 102 9.1 Criteria for choosing the compression ratio ...................................................................... 102 9.2 Simple cycle with non-cooled turbine .............................................................................. 104 9.2.1 π«πΌπ irreversibility due to pressure drops (in the air filtration system and in every other part of the cycle) ....................................................................................................................... 106 9.2.2 π«πΌπ fluid dynamic irreversibilities in the compressor ................................................. 107 9.2.3 π«πΌπ combustion irreversibility ..................................................................................... 107 9.2.4 π«πΌπ fluid dynamic irreversibilities of the turbine (adiabatic assumption) ................... 108 9.2.5 π«πΌπ losses due to the discharge of exhaust gas into the atmosphere (stack losses) ..... 108 9.2.6 π«πΌπ thermal losses ........................................................................................................ 109 9.2.7 π«πΌπ mechanical/electrical/auxiliary losses .................................................................. 109 9.2.8 Observations ................................................................................................................. 109 9.3 Recuperated cycle with non-cooled turbine ...................................................................... 110 3 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 9.3.1 π«πΌπ losses relating to the heat transfer in the recuperator ........................................... 111 9.3.2 Observations ................................................................................................................. 112 9.4 Gas cycle with cooled expansion ...................................................................................... 113 9.5 Integrating cooled gas turbines in combined cycles ......................................................... 120 9.6 Effects of real gas in the gas cycles .................................................................................. 123 9.7 Other numerical examples ................................................................................................ 124 4 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 WHY THESE CLASSNOTES Cycles are the most useful architecture for power generation. The definition of thermodynamic cycle is a series of processes that involve energy transfer, both in form of heat and mechanical work, and eventually returns the system to the initial state. The analysis of such systems is of paramount to design, rate and size the equipment needed, to assess the critical aspects, and highlight which components and/or processes can be improved in order to achieve higher outputs. In this frame of reference, it is useful to anticipate that the second law analysis aims at studying how a real cycle differs for the ideal, reversible cycle in the same condition, i.e. at studying the differences between the real cycle and the ideal one working with the same energy input. Second law analysis quantifies these differences and allows to understand which processes introduce the most irreversibilities and are thus the processes where the most improvement could be made. The scope of this class note is to provide the student with a solid theoretical basis of the effect of entropy generation in thermodynamic cycles and the ability to perform a second law analysis, or entropic analysis on the said thermodynamic cycles both in a preliminary and an accurate way. The structure of the present document is as follows: • • • • • • • • • Chapter 1 introduces the second law analysis concepts Chapter 2 shows the thermodynamic frame of second law analysis in general Chapter 3 discusses the second law analysis of energy sources Chapter 4 shows examples of irreversibilities Chapter 5 lists all possible irreversibilities in power plants Chapter 6 shows the thermodynamic frame of second law analysis for power plants Chapter 7 presents the list of possible losses in Rankine cycles Chapter 8 shows the effect of the working fluid on losses in Rankine cycles Chapter 9 presents the list of possible losses in Joule-Brayton cycles 5 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 1 INTRODUCTION The energy analysis of power systems requires a knowledge of technical and theoretical aspects in order to define properly a system, the surrounding hypotheses and the parameters of merit. The enormous variability of technologies, plant configurations and applications requires a methodology for thoroughly comparing different solutions. The objective of this chapter is to provide the student with the information and knowledge necessary for the first and second law analysis applied to power plants. In particular, the second law analysis, also called entropic analysis, allows identifying the main causes for loss of power output, and which processes shall be modified for improving the performance of the plant. The purpose of this methodology is to get the net work of the plant starting from the maximum reversible work, which can be obtained with a reversible system, minus the sum of the works lost due to irreversibility and, hence, to increases in entropy of the universe. ππ’ = ππππ£ − ∑ βππ (1.1) π The definition and calculation of the wasted works βππ passes through the second-law analysis of the different processes. This method will be applied to the two main power production plants, i.e. the Rankine cycle and the Brayton cycle. 6 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 2 THERMODYNAMICS REFERENCES This chapter illustrates the main reference cycles that are used when analyzing different power cycles. This is done in order to set a standard procedure that is thermodynamically consistent. It will also show with examples the difference between ideal cycles and reversible cycles. 2.1 DEFINITIONS It is useful to refer to some basic concepts of thermodynamics, which should be well known to those students who have attended the courses preliminary to the energy conversion course. • Power cycle: a thermodynamic cycle whose purpose is to convert thermal energy into mechanical work. In the temperature-entropy plane, a power cycle works clockwise. • Efficiency: is defined as the ratio between the useful work (net mechanical energy supplied outside) and the heat entering the cycle 2.2 FIRST REFERENCE: SECOND LAW OF THERMODYNAMICS (CARNOT'S THEOREM) Carnot's theorem states that “all reversible power cycles operating between a heat source at a constant temperature ππππ₯ and a heat sink at a constant temperature ππππ have the same efficiency, equal to: πππππππ‘ = 1 − ππππ ππππ₯ (2.1) Fig. 2.1 - Diagram of the Carnot cycle operating between two sources at a constant temperature The three conditions necessary so that a thermodynamic cycle (not necessarily a Carnot cycle) has the Carnot efficiency are: 1. Reversibility: all processes must be reversible, therefore devoid of any dissipative process that causes an increase in the entropy of the universe; 2. Heat source (also called heat sink) with constant temperature equal to ππππ₯ and therefore with infinite heat capacity; 3. Cold source (also called cold sink) with constant temperature equal to ππππ and therefore with infinite heat capacity. If even one of these conditions fails, the Carnot efficiency cannot be achieved. Therefore, the Carnot efficiency increases the more the source and heat sink move away, and asymptotically tends to 1 for ππππ₯ tending to infinity. 7 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 2.3 SECOND REFERENCE: "IDEAL CYCLE” AND “REVERSIBLE CYCLE” A cycle obtained with ideal components is an ideal cycle. This includes: • machines that operate in the absence of friction and irreversible fluid dynamic processes (turbulence, boundary layer separations, mixing); • heat exchangers that work with infinite surface, and therefore with an infinitesimal temperature difference at least in one section; • components that have no pressure drops. An ideal cycle can operate with “real” fluids, i.e. with fluids characterized by volumetric behavior appreciably different from the ideal gas equation of state. The “reversible” cycle, on the other hand, requires that all transformations are reversible, including the introduction and the rejection of energy by heat interaction. When applying the second law of thermodynamics to power cycles, the following recommendations should be kept in mind: • Do not confuse ideal cycles (ideal machines, infinite surface heat exchangers, no pressure drops, real fluids) with reversible cycles (all processes must be reversible, including the heat exchanges with external sources) • The Carnot efficiency is achieved only if the source and heat sink are at constant temperature. If the temperature is variable, it will be possible to define a new reversible cycle called Lorentz cycle, presented later. The Lorentz cycle has efficiency lower than that of Carnot. The following examples will clarify the meaning of the recommendations. 2.3.1 Example 1: how can a Rankine cycle be made reversible? 1.1.1.1 Saturated vapor cycle The ideal cycle with saturated vapor shown in Fig. 2.2 is considered. The cycle will have a lower efficiency than the Carnot cycle operating between the same temperatures even assuming that the heat source is at a constant temperature equal to the evaporation temperature, and that the heat sink is at a constant temperature equal to the condensing temperature. CYCLE LAYOUT 4 Wout Qin • 1-2 isentropic compression considering incompressible liquid • 2-3 introduction of heat to bring the condensate from the condensation to evaporation temperature • 3-4 introduction of heat to bring the saturated liquid to saturated vapor conditions • 4-5 isentropic expansion of the vapor from the evaporation to condensation pressure • 5-1 condensation of the vapor which brings the working fluid back to the saturated liquid conditions 5 3 Qout 2 1 Win 3 1≡2 4 5 Fig. 2.2 - Ideal cycle with saturated vapor 8 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 Despite being made with ideal components (pump and turbine without fluid dynamic losses, infinite surface heat exchangers), the saturated Rankine cycle is in fact not reversible as there is an irreversible process in introducing heat. The preheating of the water takes place between the hot source and the working fluid of the cycle with heat transfer under finite difference in temperature, and therefore causes irreversibility. To make this cycle reversible, the irreversibilities have to be eliminated. This can be achieved with an infinite series of infinitesimal bleeds that preheat the condensate as shown in Fig. 2.3 and Fig. 2.4 in an infinite series of infinite surface heat exchangers. In such cases, all of the heat is introduced only in the evaporation stage at constant temperature, and therefore the cycle reaches the Carnot cycle efficiency. In practice, this is not feasible because it would require an infinite number of components operating with infinitesimal flows, accompanied by unsustainable technological complications and an economic burden. In industrial practice, the regeneration is achieved with a finite number of regenerative bleeds, each of which allows the pressurized fluid to be heated by a certain finite βπ. In each of these components the heat transfer takes place under finite temperature differences. Therefore, there is a production of irreversibility and a reduction in efficiency with π < πππππππ‘ . CYCLE LAYOUT 4 Wout Qin 3 5 2 1 Qout An infinite number of regenerative bleeds (each with infinitesimal flow rate) from the turbine is assumed: each bleed goes to an infinite surface heat exchanger, where it gives up the heat of the vapor fraction phase change to the condensate. The preheaters are in cascade and the condensed vapor coming out from a preheater enters into the previous preheater following the path of the condensate, together with the vapor. Win 3 4 1≡2 5 Fig. 2.3 - Ideal cycle with saturated vapor with ideal regenerative preheating. T T T Q Q 9 of 124 Q Second-law analysis of power cycles - Energy Conversion A – V7.0 Fig. 2.4 - Reduction of the mean temperature difference in the preheating process obtained by adopting an increasing number of regenerators from left to right 1.1.1.2 Vapor cycle with superheating If the cycle with superheated vapor in Fig. 2.5 operates with a heat source of infinite heat capacity at the superheating temperature, the entire heat introduction phase takes place irreversibly. CYCLE LAYOUT 5 4 Wout Qin 3 6 2 1 • 1-2 isentropic compression of a liquid • 2-3 introduction of heat to bring the condensate from the condensation to evaporation temperature • 3-4 introduction of heat to bring the saturated liquid to the superheated vapor conditions • 4-5 expansion of the vapor from the evaporation to condensation pressure • 5-1 condensation of the vapor which brings the working fluid back to the saturated liquid conditions Qout Win 5 3 1≡2 4 6 Fig. 2.5 - Ideal cycle with superheated vapor To make the cycle reversible, it is necessary to imagine: • the adoption of an infinite number of reheatings: therefore endless turbine stages, each of which performs an infinitesimal expansion, and an infinite number of heat exchangers for introducing thermal energy into the cycle. Alternatively, it is possible to think of an isotherm expansion; • a recuperation that heats the working fluid from the condensing temperature up to the superheating temperature, with a counter-current heat exchanger that cools the vapor leaving the last turbine stage. The resulting cycle under these assumptions is depicted in Fig. 2.6. 10 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 CYCLE LAYOUT Qin Wout 4 3 2 5 Qout 1 Win 4 3 An infinite series of reheatings during the expansion that makes it isotherm and a recuperation which should take the condensate from the condensing temperature up to the state of superheated vapor are introduced 1≡2 5 Fig. 2.6 - “Carnotized” diagram of an ideal cycle with superheated vapor plant If on one hand an isothermal expansion is theoretically possible, a reversible recuperation is not feasible even from the conceptual point of view. That is because the specific heat at constant pressure of the pressurized side is different (much greater, moreover infinite during the phase change) than that of the low pressure side. This is also represented by the slope of the isobaric curves in the Ts plane, π which is equal to . Under these conditions, even if adopting an infinite surface exchanger, leading to ππ the cancellation of βπ at least in one point, the pressurized current could not be taken to the maximum temperature of the cycle. Fig. 2.7 shows the TQ diagram of the infinite surface recuperated heat exchanger. Remember that in this plane the slope of the curves is inversely proportional to the heat capacity πΆ (defined as the product of the flow rate on a mass basis πΜ and the specific heat at constant pressure ππ of the flows affected by the heat transfer). For the evaporation section, it is equal to zero given that ππ tends to infinity, while in the vapor phase it is always greater than in the liquid phase1. Therefore, the process cannot be made reversible due to the different slope of the high and low pressure isobaric lines, even in the presence of infinite surfaces. The heat available from the cooling of the vapor is only sufficient to preheat the condensate and to evaporate the vapor partially, even assuming 1 Also remember that the difference in slope between the liquid and vapor phases depends on the complexity of the fluid. The simpler the fluid, the greater, in relative terms, the correction of the specific heat passing from liquid to saturated vapor. While for complex fluids characterized by a high number of degrees of freedom, the specific heat of the vapor is about equal to that of the liquid. The proportion between the heat introduced during phase change and that introduced during preheating and superheating is instead mainly a function of the molecular weight. The greater the molecular weight is, the less ββππ£π generally is. Therefore, the lower the relative weight of the evaporation phase is with respect to the preheating phase. 11 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 cooling up to the condensation temperature. The rest of the heat necessary to get superheated vapor must still be supplied from the outside. Two irreversible heat transfer processes are always present in those systems: • recuperation • introduction of heat to complete evaporation and achieve superheating. T Q Fig. 2.7 - Temperature-exchanged thermal power diagram for an ideal recuperation 2.3.2 Example 2: how can a Joule cycle be made reversible? An ideal Joule cycle has a lower efficiency than a Carnot cycle operating between sources of infinite heat capacity with same ππππ₯ and ππππ . This is due to the presence of two irreversible processes: the introduction and release of heat as shown in Fig. 2.8. This cycle was improved by introducing a recuperator that uses the heat discharged by the turbine to preheat the compressed gas. Even if reversible, this component however does not make the cycle totally reversible by itself because a certain amount of heat must always be exchanged with the sink under finite temperature differences. In addition, the recuperator may be reversible only if it has an infinite surface and if the two currents have equal heat capacities throughout the temperature change. The first condition is not accessible from a technical and economic point of view, but it is conceptually feasible. The second, on the other hand, can be precluded a priori if the cycle shows an internal combustion system or if the gas presents different real fluid effects on the high and low-pressure sides. In both cases, if the heat transferred is equal, the temperature change on the two sides will be different. Thus, there will be certainly finite βπ in some sections of the heat exchanger. The first case is representative of a gas turbine cycle in which both the flow rate and the gas composition (and the specific heat) are different on the two sides of the heat exchanger. The most relevant factor is given by the difference in flow rate since downstream of the compressor a substantial part of the compressed air is extracted for cooling high-pressure blades and for diluting exhaust gases in the combustor. The mass flow at the outlet of the turbine is thus larger that the mass flow at the outlet of the compressor. The second factor can be represented, for example, by a closed gas cycle operating close to the saturation dome. In this case the real fluid effects become important on the high pressure side, increasing the specific heat of the fluid which will therefore have a lower temperature change, heat transferred being equal. 12 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 CYCLE LAYOUT Qin 2 3 Wout 3 • 1-2 isentropic compression of an ideal gas • 2-3 introduction of heat to bring the compressed gas to the maximum temperature of the cycle (coinciding with that of the source) • 3-4 isentropic expansion of an ideal gas • 4-1 transfer of heat to bring the expanded gas to the minimum temperature (coinciding with that of the heat sink) 4 1 Qout 4 2 1 Fig. 2.8 - Ideal Joule cycle Fig. 2.9 - Diagrams of heat transfer for a recuperator of an open gas cycle (left) and for a closed gas cycle with real fluid effects (right) To eliminate the irreversibilities of heat transfer with the heat sinks, instead, three alternatives are conceptually viable: • Make an isotherm compression at the temperature ππππ and an isotherm expansion at the temperature ππππ₯ (Fig. 2.10). These processes are theoretically achievable with an infinite plant cost and, at the same time, the adoption of a recuperator2. This cycle is called Ericsson cycle. 2 In practice, in gas cycles the idioms regeneration and recovery are used without distinction. Originally, recuperator in English indicates a surface heat exchanger, regenerator a heat accumulating device, a path alternatively from a hot current and a cold current. 13 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 • Decrease the compression ratio (up to the unit value) and simultaneously use a recuperator (Fig. 2.11). • Increase the compression ratio until all the gas is entirely heated in the compression phase. All these solutions maximize the efficiency that becomes coincident with that of Carnot. However, the last two lead to a useful work of the cycle equal to zero since, in both cases, the compression and expansion works are coincident and the introduction of heat is zero. CYCLE LAYOUT Qin 3 Wout 1 4 3 2 Qout 4 2 1 • 1-2 isotherm compression of an ideal gas at ππππ • 2-3 introduction of heat to bring the compressed gas to the maximum cycle temperature (coinciding with that of the source) by means of a reversible recuperator • 3-4 isotherm expansion of an ideal gas at ππππ₯ • 4-1 transfer of heat to bring the expanded gas to the minimum temperature (coinciding with that of the heat sink) by means of a reversible recuperator Fig. 2.10 - Reversible Joule cycle obtained with isotherm compression and expansion (called Ericsson cycle). CYCLE LAYOUT Qin 3 3 Wout 2 1 6 4 5 6 4 Qout 2 1 5 • 1-2 isentropic compression of an ideal gas (T2= T1) • 2-6 preheating of the compressed gas at the expense of the expanded gas • 6-3 introduction of heat to bring the compressed gas to the maximum cycle temperature (coinciding with that of the source): since T3=T6, the heat introduced is infinitesimal • 3-4 isentropic expansion of an ideal gas (T3= T4) • 4-5 recuperative cooling of the expanded gas • 5-1 transfer of heat to bring the expanded gas to the minimum temperature (coinciding with that of the heat sink); the heat transferred is infinitesimal (T5=T1) Fig. 2.11 - Reversible Joule cycle obtained with compression ratios tending to one and infinite surface recuperator. 14 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 3 2 CYCLE LAYOUT Qin 2 3 Wout 4 1 1 • 1-2 isentropic compression of an ideal gas (T2 coincides with T3) • 2-3 introduction of heat to bring the compressed gas to the maximum cycle temperature (coinciding with that of the source): since T3=T2, the heat introduced is infinitesimal • 3-4 isentropic expansion of an ideal gas (T3= T4) • 4-1 transfer of heat to bring the expanded gas to the minimum temperature (coinciding with that of the heat sink); the heat transferred is infinitesimal (T4=T1) 4 Qout NOTE: All the temperature increase occurs due to isentropic compression Fig. 2.12 - Reversible Joule cycle obtained with high compression ratios that lead to making the heat contribution from the outside infinitesimal. 2.3.3 Example 3: Stirling cycle The ideal (or limit, in this case the two definitions coincide) Stirling cycle is composed of two isochores and two isotherms as shown in Fig. 2.13. The heat is introduced at constant temperature ππππ₯ and is transferred at constant temperature ππππ . Both processes take place with infinitesimal temperature differences, so they are reversible. The recuperation has the same limitations seen in the previous example, and it does not present irreversibilities only if the gas is an ideal gas on both sides of the exchanger. With this condition, the two curves in the TQ diagram are superimposed since the specific heat depends only on the temperature. The efficiency of the Stirling cycle in this case is identical to that of the Carnot cycle operating between the same temperatures. 3 2 • 1-2 reversible isotherm compression of an ideal gas (T1= T2) • 2-3 introduction of heat to bring the compressed gas to the maximum cycle temperature (coinciding with that of the source) at constant volume, at the expense of the expanded gas (recuperation) • 3-4 reversible isotherm expansion of an ideal gas (T3=T4) with introduction of heat at the maximum cycle temperature (coinciding with that of the source) • 4-1 transfer of heat to the compressed gas to bring the expanded gas to the minimum temperature (coinciding with that of the heat sink) at constant volume 4 1 Fig. 2.13 - Representation in the temperature-specific entropy plane of the Stirling cycle 15 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 2.4 THIRD REFERENCE: REVERSIBLE HEAT PUMP BETWEEN SOURCES/SINKS AT CONSTANT TEMPERATURE: The thermodynamic cycles can operate following: • a clockwise path as in Fig. 2.14a: it is the case of the direct or power cycles. The heat π1 enters the cycle (and therefore transfers from a heat source to the working fluid) in one or more transformations at high temperature. The heat π2 exits the cycle (and therefore transfers from the working fluid to the heat sink) at low temperature. The useful effect of a power cycle is the conversion of part of the heat introduced into work ππ’ . The parameter for identifying the energy performance of a power cycle is the first law efficiency: ππΌ = • ππ’ π1 (2.2) a counterclockwise path as in Fig. 2.14b: it is the case of the inverse cycles. The heat π2 enters the cycle (and therefore transfers from a heat source to the working fluid) in one or more transformations at low temperature. The heat π1 exits the cycle (and therefore transfers from the working fluid to a heat sink) at a higher temperature. Depending on the application, the useful effect of an inverse cycle is the removal of heat from the low temperature source (refrigeration cycle) or the transfer of heat to the heat sink at a higher temperature (heat pump cycle). In both cases these heat transfers are achieved at the expense of the work ππ’ introduced into the cycle. In some cases, it may happen that both effects are to be considered "useful". One easy example to be remembered is a cycle used by a machine that dispenses ice (using cold) and hot water (using heat). Another case, more significant in terms of energy, is an inverse cycle designed for the climate control of a building which simultaneously has areas where cooling is required and areas where heating is required. In these cases, the use of heat and/or cold is often partial. T T s s Fig. 2.14 - Direct reversible cycles (or power cycles) and indirect cycles (or refrigeration, or with heat pump) in the temperature-specific entropy plane The parameter to identify the energy performance of an inverse cycle is the coefficient of performance (πΆππ), defined as: πΆππ = ππ₯ ππ’ 16 of 124 (2.3) Second-law analysis of power cycles - Energy Conversion A – V7.0 where ππ₯ is the heat taken from the cold source (π2 ) in the case of a refrigeration cycle, or the heat released to the heat sink (π1 ) in the case of a heat pump. For the first law of thermodynamics: πΆπππ»π = π1 π2 + ππ’ = = πΆπππΆπ»πΏ + 1 ππ’ ππ’ (2.4) In the case of reversible cycles (and only for them), the ratios between the transferred heat flows (π1 and π2 ) and ππ’ are obviously independent from the direction of the cycle, therefore: ππΌ = 1 1 = πΆπππ»π πΆπππΆπ»πΏ + 1 (2.5) If the reversible cycles operate with sources/heat sinks at a constant temperature as in Fig. 2.15: ππππ ) ππππ₯ ππππ₯ πΆπππ»π = ππππ₯ − ππππ ππππ πΆπππΆπ»πΏ = ππππ₯ − ππππ ππΌ = 1 − ( (2.6) (2.7) (2.8) For real cycles these relations no longer apply. In particular, the πΆπππ»π is less than the ratio 1/ππΌ . T Qin Qout Wout Win Qout Qin s Fig. 2.15 - Reversible Carnot cycle in the T-S plane. It can be covered in a clockwise or counterclockwise direction, without changes of the ratios between the work exchanged with the outside and the heats exchanged with the two heat sources. It must be also remembered that the area of the cycle in the TS plane has no physical meaning for real cycles, unlike what takes place for ideal or limit cycles. Similarly, the area under a real transformation (∫ π ππ) does not represent the heat transferred during the transformation. 17 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 3 ENERGY SOURCES FOR POWER PLANTS This chapter introduces the calculation of the potential of different energy sources. A classification is first given. Then the definition of exergy for an energy source is given. A chemical and entropic analysis of fuels is performed, and combustion as a mean of transferring energy is analyzed. 3.1 ENERGY SOURCES DEFINITION The power production plants can be powered by a wide variety of energy sources. Some of the main ones are: • non-reacting flows: recovery of waste heat, geothermal energy; • fossil and renewable fuels: solid (like coal and biomass), liquid (like gasoline, diesel, bioethanol, and biodiesel) and gaseous (like natural gas, syngas, biogas, and biomethane); • solar energy, etc. Each of these sources can be modelled in a different way, and some can be considered with infinite heat capacity while others cannot. Other sources, although available at very high temperature (sometimes modeled as an infinite temperature: solar radiation, combustion flue gases) require the use of a heat-transfer fluid to transfer heat from the energy source to the working fluid, so they refer to the first type. 3.2 NON-REACTING FLOWS, MAXIMUM WORK, EXERGY The maximum work obtainable from a flow rate of a fluid in thermodynamic conditions different from the environment is that obtained with any reversible path (or formed by one or more reversible processes) that brings the fluid in equilibrium with the environment3. If all the processes are reversible, there is no increase in entropy of the universe and, as a result, there is no loss of useful work. An energy source constituted by a constant flow rate πΜ of air with pressure ππ₯ and temperature ππ₯ is assumed. It is assumed that the air is an ideal gas with ππ constant with the temperature. The purpose of this hypothesis is the analytical steps, but the same considerations can be made for any fluid. Let's consider the simplest reversible path4 in Fig. 3.1: 1. a reversible adiabatic expansion (isentropic) from point x up to condition A at temperature π0 and pressure ππ΄ < π0 2. a reversible isotherm compression from condition A up to ambient pressure p0 For all the transformations that will be considered here and in the following paragraphs, it will be assumed that: a) The flow is one-dimensional at the input and output sections, so it is permissible to identify the kinematic and thermodynamic properties of the fluid with average values in these sections b) The flow is stationary and each quantity is independent from time. In analytical ππ terms, = 0 ππ‘ 3 Equilibrium of the fluid with the environment means that the source is brought to the same temperature and pressure of the dead state. As we will see in the section dedicated to fuels, the maximum work can be obtained when the equilibrium considers also the chemical contribution (i.e. same composition). 4 Another chain of reversible processes might be: (i) an isotherm compression up to a point z with a ππ§ greater than ππ₯ such as to get an entropy equal to π 0 and (ii) an isentropic expansion up to π0 . In this case a Carnot cycle that reversibly recovers the heat released by the compressor is necessary. On the other hand, an isentropic expansion up to π0 and the subsequent isobaric cooling up to π0 is not a reversible path since the heat transfer with the cold sink would take place with the production of irreversibility. In this case a trilateral cycle that reversibly recovers the heat released by the compressor would be necessary. 18 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 Adopting the sign conventions best suited to energy conversion (positive heat if entering the system, positive work if going out of the system), the following equation is obtained: π1 2 π2 2 β1 + + ππ§1 + π = β2 + + ππ§2 + π 2 2 (3.1) The further hypothesis of null (or in any case negligible) changes in kinetic (π 2 /2) and gravitational energy (ππ§) can be applied in most of the components used in conversion plants Fig. 3.1 - Sequence of ideal and reversible transformations necessary for the production of the maximum power extractable from a flow of gas not in equilibrium with the dead state, and representation in the Ts diagram The power obtained from the isentropic and adiabatic expansion (π = 0) from π₯ to A is equal to: πΜπ₯→π΄ = πΜππ π΅ = πΜ(βπ₯(ππ₯,ππ₯) − βπ΄(π0 ,ππ΄) ) (3.2) The power spent in the isotherm and reversible compression (transfer under infinitesimal temperature differences and without friction) from A to 0 is instead equal to: πΜπ΄→0 = −πΜπΆππ = −πΜπΆππ = −πΜπ0 (π π΄(π0 ,ππ΄ ) − π 0(π0 ,π0) ) (3.3) The resulting total power is therefore equal to: πΜπ = πΜππ π΅ − πΜπΆππ = πΜπ₯→π΄ + πΜπ΄→0 = πΜ(βπ₯(ππ₯,ππ₯) − βπ΄(π0 ,ππ΄ ) )−πΜπ0 (π π΄(π0 ,ππ΄ ) − π 0(π0 ,π0) ) (3.4) where being π π΄ = π π₯ and βπ΄ = β0 it turns out πΜπ = πΜ[(βπ₯ − π0 π π₯ ) − (β0 − π0 π 0 )] (3.5) where the terms in round brackets are also known as specific exergy, a state function of the fluid. πΜπ = πΜ[ππ₯ − π0 ] 19 of 124 (3.6) Second-law analysis of power cycles - Energy Conversion A – V7.0 The result obtained for this particular example is generally valid. Therefore, the specific exergy has the physical meaning of the capacity to produce mechanical work (or maximum obtainable work) for unit mass of a fluid in the absence of changes in kinetic and potential energy, from a fluid that, as it is able to exchange heat only with the external environment, evolves from an initial generic state to a final state in thermodynamic equilibrium with the environment. 3.2.1 Trilateral and trapezoidal cycles Now a look will be taken at how reversible work can be assessed with the assumption of a flow at a temperature π > π0 and pressure equal to that of reference, which is cooled up to the condition of equilibrium with the environment. It is also assumed that the source is at constant heat capacity. It is possible to imagine a reversible system with the following processes: a) a "trilateral" reversible cycle, sometimes said Lorenz’ cycle, consisting of: 1. a step of heat introduction at variable temperature, from π0 to ππ₯ , in which the source is reversibly cooled from the temperature ππ₯ to temperature π0 . The following has to be assumed: (i) a counter-current heat exchanger in which the working fluid and the source have the same thermal capacity at each temperature, (ii) of infinite surface, (iii) adiabatic outwards and (iv) without pressure drops 2. an isentropic expansion step that cools the working fluid of the cycle up to the temperature π0 3. a reversible isotherm compression step at temperature π0 , which reversibly releases heat to the environment b) a mixing step in thermodynamic equilibrium (equal π, π) of the source. The production of entropy resulting from the irreversible phenomenon of a release into the environment of a fluid in thermodynamic equilibrium, but with a chemical composition different from the ambient, will be discussed in the chemical exergy topic. T reversible cycle 2 2 1 1 3 3 s Fig. 3.2 - Use of a trilateral cycle for extracting the maximum mechanical power from the cooling of a hot flow. Therefore, the reversible power is equal to the exergy difference of the fluid in point x and in equilibrium with the environment conditions πΜπππ£ = πΜ[(βπ₯ − π0 π π₯ ) − (β0 − π0 π 0 )] πΜπππ£ = πΜ[(βπ₯ − β0 ) − π0 (π π₯ − π 0 )] (3.7) (3.8) If a process is considered isobaric, then the terms of enthalpy difference and entropy difference are a function of the temperature only. 20 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 πβ = ππ (π)ππ (3.9) ππ (π) (3.10) ππ = ππ π If it is possible to consider constant specific heat (or at least to calculate an average value), the heat capacity of the source is constant, and the following is found: ππ₯ πΜπππ£ = πΜ [ππ (ππ₯ − π0 ) − π0 ππ ππ ( )] π0 (3.11) By gathering ππ (ππ₯ − π0 ) and observing that πΜππ (ππ₯ − π0 ) = πΜ 1 , it turns out that: πΜπππ£ = πΜ1 [1 − π0 π ] (3.12) (ππ₯ − π0 )⁄ππ ( π₯ ) π0 π where the term (ππ₯ − π0 )⁄ππ ( π₯) is called mean logarithmic temperature of the hot source ππππ(ππ₯|π0) . π0 The term in square brackets is called the efficiency for the trilateral cycle: π ππππ = 1 − π0 (3.13) ππππ(ππ₯|π0) The efficiency thus found can be considered as the efficiency of an equivalent Carnot cycle operating between ambient temperature and the mean log temperature of heat introduction. This relation may be used both for gaseous and liquid streams in which the cooling is isobaric and the heat capacities can be considered constant. A similar result is also obtained by limiting the cooling of the source up to a temperature ππ¦ higher than the ambient temperature. The chain of reversible transformations to calculate the expression of the maximum obtainable work requires: (i) isentropic expansion from ππ₯ to π0 , (ii) isothermal compression at π0 up to the entropy π π¦ , (iii) isentropic compression from π0 to ππ¦ . T reversible cycle 2 1 2 1 4 4 3 3 s 21 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 Fig. 3.3 – Adoption of a trapezoidal cycle for extracting the maximum mechanical power from the cooling of a hot flow with a minimum temperature limit. It is therefore equal to πΜπππ£ = πΜππ π΅ − πΜπΆππ1 − πΜπΆππ2 Μ ππππ£ = πΜ[(βπ₯ − β0 ) − π0 (π π₯ − π π¦ ) − (βπ¦ − β0 )] πΜπππ£ = πΜ[(βπ₯ − βπ¦ ) − π0 (π π₯ − π π¦ )] πΜπππ£ = πΜ[ππ₯ − ππ¦ ] (3.14) (3.15) (3.16) (3.17) With the hypothesis of constant heat capacity, the efficiency of the trapezoidal cycle that reversibly receives heat from the heat source is obtained with a procedure similar to that used for the trilateral cycle. π ππππ = 1 − π0 (3.18) ππππ(ππ₯|ππ¦) Therefore, it is clear that for heat capacities of the source tending to infinite, the source becomes isothermal, ππ¦ → ππ₯ , and the trilateral cycle efficiency obviously tends to that of the Carnot cycle. Finally, the most general case is the one with both hot and cold sources with variable temperature. The reference cycle is the mixtilinear cycle shown in Fig. 3.4, whose efficiency can be calculated starting from the variation in exergy of the hot and cold sources calculated according to: π0 βπ΅Μβ = πΜππ (1 − ) ππππ(ππππ₯2|ππππ₯1) π0 βπ΅Μπ = πΜππ’π‘ (1 − ) ππππ(ππππ2|ππππ1) (3.19) (3.20) where exergy variations are calculated from the Lorentz efficiency (Eq. (3.13)). Then, the reversible power that can be produced is given by the difference between the two sources exergy variations: πΜπππ£ = βπ΅Μπ − βπ΅Μπ (3.21) With βπ΅Μβ and βπ΅Μπ positive. From the first principle of thermodynamics, πΜππ = πΜππ’π‘ + πΜπππ£ and the πΜ efficiency is defined as π = πππ£, it turns out that πΜππ’π‘ = πΜππ (1 − π) and: πΜππ πΜπππ£ = πΜππ (1 − π0 ππππ(ππππ₯2|ππππ₯1 ) ) − πΜππ (1 − π) (1 − π0 ππππ(ππππ2|ππππ1) ) (3.22) By gathering πΜππ the following expressions can be obtained π = (1 − π0 ππππ(ππππ₯2 |ππππ₯1 ) −1+ π0 ππππ(ππππ2|ππππ1) 22 of 124 ) + π (1 − π0 ππππ(ππππ2|ππππ1) ) (3.23) Second-law analysis of power cycles - Energy Conversion A – V7.0 π( π0 ππππ(ππππ2 |ππππ1 ) )=( π0 ππππ(ππππ2|ππππ1 ) − π0 ππππ(π πππ₯2 |ππππ₯1 ) ππππ(ππππ2|ππππ1) ππππ,f π = (1 − ) = (1 − ) ππππ(ππππ₯2|ππππ₯1) ππππ,c ) (3.24) (3.25) T ππππ₯ ππππ₯2 ππππ ππππ₯1,2 ππππ₯1 ππππ2 π0 ππππ ππππ1,2 ππππ1 s Fig. 3.4 – Quadlateral mixtilinear cycle 3.3 FOSSIL AND RENEWABLE FUELS The energy source most common for power production processes are the fuels (identified by the chemical composition, thermodynamic conditions, etc.) which after combustion reactions can make a certain amount of heat available. The energy content of the fuel can be described by four values specific to the mass unit5: • Lower Heating Value • Higher Heating Value • Reversible Work • Chemical exergy. Their theoretical definition will be provided in this chapter, and the main differences between these indexes will be highlighted. In addition, practical recommendations for their use in the analysis of power cycles will be discussed. 3.3.1 Stoichiometry in combustion reactions Combustion reactions are exothermic oxidation reactions that transform the reactants into combustion products. The reactants are formed by the fuel (compound that contains chemical species able to oxidize) and oxidant (compound that provides the oxygen necessary for oxidation). The products of combustion are obtained at a temperature higher than that of the reagents, making useful heat available to thermodynamic power production processes. A combustion reaction always respects the mass balance and the atomic balance, while the number of moles may change in accordance with the chemical species destroyed and formed during the reaction. The atomic chemical species involved in the combustion reactions are mainly carbon and hydrogen, which form typical fuels in the largest portion. There may be traces of sulfur and nitrogen, depending on their composition. 5 Generally, reference is made to the Sm3 for gaseous fuels, which is always a value that identifies the mass. Indeed, in 1m3 at standard conditions (T=0°C, p=1bar) 44.03 moles are always contained. Therefore, once the gas composition is known, the mass is as well. 23 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 The basic combustion reactions for a mole of these elements with molecular oxygen π2 are: πΆ + π2 → πΆπ2 1 1 π» + π2 → π»2 π 4 2 A combustion reaction of a fuel of chemical composition πΆπ π»π is now analyzed. The total combustion π of a mole of this compound will require π + moles of π2 . 4 ππΆ + ππ2 → ππΆπ2 π π ππ» + π2 → π»2 π 4 2 This reaction is called stoichiometric since there is no trace of the reagents in combustion products as they have been completely consumed. The ratio between oxidant mass and fuel mass for a combustion reaction is called stoichiometric ratio and is expressed as: πΌ= ππ2 ππΆπ π»π (3.26) For example, the combustion of a gas mixture with composition on a mass-basis π¦π (methane 60%, ethane 30%, propane 10%) can be calculated with the following steps: 1. calculation of the composition on a molar-basis of the mixture: the kilomoles of each compound are found considering a base of 1 kg of fuel as: 0.6 = 0.0375 kmol → 75.3% 16 π¦π ππ 0.3 ππ = = πΆ2 π»6 → = 0.01 kmol → 20.1% πππ πππ 30 0.1 πΆ π» → = 0.0023 kmol → 4.6% 3 8 { 44 πΆπ»4 → 2. the moles of oxygen required for complete combustion of the reacting compound moles are calculated ππ2 ,π | π π‘πππ πΆπ»4 → 0.0375 (1 + 0.25 β 4) = 0.1125 kmol ππ (2 + 0.25 β 6) = 0.05 kmol = ππ (ππ + ) {πΆ2 π»6 → 0.01 4 πΆ3 π»8 → 0.0023 (3 + 0.25 β 8) = 0.016 kmol 3. the flow rate on a mass-basis of oxygen required is calculated πΌπ π‘ = ππ2 | π π‘πππ = πππ2 ∑ππ2 ,π | π π‘πππ = 3.883 kg π The flue gas produced shall consist solely of πΆπ2 and π»2 π with a composition on a molar-basis given by: 24 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 πΆπ»4 → 0.0375 (1) = 0.0375 ππΆπ2 | = ∑ ππ (ππ ) = ∑ {πΆ2 π»6 → 0.01 (2) = 0.02 = 0.0644 kmol π π‘πππ πΆ3 π»8 → 0.0023(3) = 0.0069 π π → 36.1% πΆπ»4 → 0.0375 (0.5 β 4) = 0.075 ππ ππ»2 π | = ∑ ππ ( ) = ∑ {πΆ2 π»6 → 0.01 (0.5 β 6) = 0.03 = 0.1142 kmol π π‘πππ 2 πΆ3 π»8 → 0.0023 (0.5 β 8) = 0.0092 π π → 63.9% And in mass-basis terms ππΆπ2 | π π‘πππ = πππΆπ2 ππΆπ2 | ππ»2π | π π‘πππ = πππ»2 π ππ»2π | π π‘πππ π π‘πππ = 2.834 kg → 58% = 2.0556 kg → 42% That allows the mass balance of the reaction to be checked. It can be noted that the combustion π products are always damp, and have a greater water content the greater the ratio in the fuel. π If the reaction takes place with ambient air instead of pure oxygen in the combustion products, the presence of nitrogen will also have to be considered. Assuming a simplified composition of ambient air equal to 79% π2 and 21% π2 on a molar-basis, the value πΌπ π‘ can be calculated as: πΌπ π‘πππ = ππ2 | π π‘πππ + ππ2 | π π‘πππ = πππ2 ∑ππ2 ,π | π π‘πππ π + πππ2 ∑ 3.76 ππ2,π | π π‘πππ π where the factor 3.76 is given by the ratio between the mole fractions of nitrogen and oxygen in the ambient air. If the reaction is not stoichiometric, a flow rate of oxidant different from that calculated above is used, and it is useful to define the combustion ratio as: π= πΌ πΌπ π‘πππ =1+π (3.27) Two cases are possible: • combustion in excess of oxygen π > 1, π > 0: more oxidant than the stoichiometric is introduced and total combustion of the fuel is obtained. The combustion products will have the same content on a mass-basis of the stoichiometric case for the species πΆπ2 and π»2 π, plus the unreacted oxygen and non-reacting nitrogen • combustion lacking oxygen π < 1, π < 0: less oxidant than the stoichiometric is introduced and total combustion of the fuel is not obtained. The content πΆπ2 and π»2 π in the combustion products will be lower than that of the stoichiometric case. Added to this will be the unreacted fuel, non-reacting nitrogen and the presence of incomplete combustion products, such as πΆπ. Depending on the type of fuel and the technological solutions adopted, it will be more or less expedient to use a certain excess of oxygen: • In coal dust plants, the heterogeneous combustion between coal and combustion air requires a certain degree of excess air, about 6%, in order to be certain that the reaction will be completed, 25 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 • • 3.3.2 and to have no unburnt materials in the flue gas or in the ash collection hopper. For gas boiler plants, the excess air drops to 3% since the combustion is in the homogeneous phase. In gas turbines, on the other hand, an extremely high combustion ratio is used even if the fuel is gaseous. That is because it is necessary to dilute the flue gas produced in the combustor, to limit the formation of NOx and reduce the wall temperature of the combustion chamber. In internal combustion engines a lack of oxygen is instead used to keep down the emissions. In this case, there is the presence of CO in the flue gas since it is a not wholly oxidized species. Enthalpies of formation and enthalpy balance of combustion It is possible to write the energy balance for reacting systems considering that the system evolves from the reactants to the products in a continuous reactor that, in the most general case, can exchange heat and work with the external environment with π − π = π»π − π»π reagents combustion products (fuel) +1 (combustion products) (comburent) W Fig. 3.5 - Enthalpy balance of the first law for a reacting system or π − π = ∑ ππ βΜπ(π,π)π − ∑ ππ βΜπ(ππ,ππ)π π∈π (3.28) π∈π where the enthalpies must be calculated as βΜ(π,π) = βΜ0 + ΔβΜ(π,π) , considering the enthalpy of reference βΜ0 and ΔβΜ(π,π) due to the deviation from the thermodynamic state of reference. The second term ΔβΜ(π,π) is the increase of enthalpy given by the difference in temperature and pressure for the pure fluid with respect to the reference conditions. Its value can be derived from reliable thermodynamic tables or equations of state since it must also consider effects of real fluid, phase transitions, etc. In this discussion the assumption of ideal gas will be adopted since it can be applied to the combustion products and reagents in gaseous form, without introducing excessive errors. In this specific case of ideal gas, the term ΔβΜ(π,π) = ΔβΜ(π) and it takes on the form: π ΔβΜ(π) = ∫ ππ0 (π)ππ π0 26 of 124 (3.29) Second-law analysis of power cycles - Energy Conversion A – V7.0 In the study of a combustion reaction, however, special attention has to be paid to the reference values adopted for the various thermodynamic quantities. Unlike what has been seen so far on enthalpy balances of non-reacting systems, the value of the enthalpy of reference β0 cannot be chosen arbitrarily but must correspond to very specific properties of the compound. In fact, for non-reacting systems the choice of βΜ0 was totally irrelevant because, if the difference in enthalpy between an initial and a final state were to be calculated, this term always simplified and its choice was therefore arbitrary. On the contrary, in the case of reacting systems in which chemical species disappear and form, this value cannot be chosen without considering the type of compound, the reactions necessary for its formation and the thermodynamic state of reference. In this case the reference enthalpy is referred as standard enthalpy of formation βΜ°π . The first step is to define the standard reference state, which by convention has π0 = 25°πΆ and π0 = 1 atm. All the molecules formed by a single element in the stable form at this thermodynamic condition have a βΜ°π = 0 by convention. The compounds that fall within this category are monatomic molecules, such as graphite πΆ or the noble gases (He, Ar, etc.), and diatomic molecules such as π»2 , π2 , π2 and not, for example, their atomic species that are much more unstable in the standard reference conditions. For all the other compounds, the reference enthalpy is equal to the standard enthalpy of formation starting from the stable compounds, i.e. the extractable thermal power starting from reactants in standard conditions and bringing the products back to the same conditions. The values of the enthalpies of formation can be obtained experimentally through statistical thermodynamics or spectroscopy methods. Take, for example, the formation reaction of πΆπ2 starting from its constituents in a stable form (πΆ and π2 ) under the standard conditions. Have the reaction take place in a reactor that then brings the product to the initial standard conditions. A certain amount of heat, which in the specific case is equal to 393.52 kJ/kmol, must be released into the environment since the reaction is exothermic. The enthalpy of formation of πΆπ2 is therefore equal to βΜ°π = −393.52 kJ/kmol, since for convention negative and positive sign is adopted for exothermic and endothermic reactions, respectively. Some compounds may have different aggregating forms at the reference conditions, and they are generally provided with two different values. For example, for water two values of enthalpy of formation are provided: (i) one referring to liquid water βΜ°π,π»2 ππ which therefore considers the possibility of recovering the enthalpy of condensation, and the other (ii) referring to water in the vapor state at the reference conditions βΜπ,π»2 ππ . It follows that βΜ°π,π»2 ππ = βΜ°π,π»2 ππ + ββΜππππ,π»2 π(π0 ) . Once the reference enthalpy, i.e. the standard enthalpy of formation for a compound in a reacting system, has been defined, the enthalpy in a certain thermodynamic state at temperature π and pressure π can be written, like: βΜ(π,π) = βΜ°π + ΔβΜ(π) (3.30) Using the assumption of ideal gas. Referring to the enthalpy balance, the input conditions may be different from one compound to the next (i.e. the different reagents may not be in thermodynamic equilibrium). While the conditions at the discharge are considered common to all the species present in the exhaust gases. With the assumption that also the reagents enter at the same conditions, the balance of the first law on a mass-basis can be rewritten in the following manner. 27 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 π − π = ∑ ππ (β°π,π + Δβπ,ππ ) − ∑ ππ (β°π,π + Δβπ,ππ ) π∈π (3.31) π∈π In order to be solved, this relation requires many considerations: • combustor type, i.e. whether or not it is possible to extract/introduce heat and work. • the possibility to neglect components of kinetic and potential energy. • the thermodynamic state of the reagents, i.e. if they are in the liquid, gaseous or solid state. • the thermodynamic state of the products. In particular, the aggregation state of the water contained in the flue gas, that may be in the liquid, vapor or partially condensed state according to the phase equilibrium at a certain temperature. • whether or not the reactants are pre-mixed. Once these assumptions are defined, it is possible to define quantities such as the adiabatic flame temperature and the heating value of the fuel. 3.3.3 Adiabatic flame temperature Let’s consider a complete adiabatic combustion of the unit of fuel mass, to which corresponds a mass of oxidant air given by the value of the stoichiometric ratio πΌπ π‘πππ . Starting from the equation and the previous assumptions, the following is considered: (i) an adiabatic combustion chamber (ii) the absence of exchanged work (iii) the perfect mixing between the reaction products at temperature ππ (iv) the perfect mixing between the reagents at temperature ππ (v) an isobaric process at pressure π0 reagents adiabatic combustor combustion products (fuel) +1 (combustion products) (comburent) Fig. 3.6 - First law enthalpy balance for an adiabatic reacting system and without extracting work to determine the adiabatic flame temperature It is possible to calculate the adiabatic flame temperature ππ΄πΉ as that temperature of the combustion products that guarantees the first law balance for an adiabatic system. It is found by means of an iterative calculation, given the implicit bond between enthalpy and temperature for ideal polyatomic gases. ∑ ππ (β°π,π + Δβπ,ππ΄πΉ ) = ∑ ππ (β°π,π + Δβπ,ππ ) π∈π π∈π 28 of 124 (3.32) Second-law analysis of power cycles - Energy Conversion A – V7.0 The adiabatic flame temperature therefore depends on: • the temperature of the reactants ππ : as the temperature of the reactants increases, the adiabatic flame temperature increases, which is generally defined with ππ΄πΉ,ππ . As will be seen, it can be raised at will by means of recuperative preheating. • from the ratio πΌ between the air flow rate and the fuel flow rate: as the air flow rate increases starting from the minimum flow rate (the one that corresponds to the stoichiometric ratio πΌπ π‘ for which complete combustion can be achieved), the adiabatic flame temperature decreases. That is because the heat generated by the combustion reaction becomes diluted over a greater flow rate of exhaust gases. 3.3.4 Heating value of the fuel With reference to the previous case, consider now an isobaric heat recovery downstream of the combustion process of a unit of fuel mass at pressure π0 . It is also assumed that this recovery brings the products back from a temperature ππ΄πΉ,ππ to that of the reagents ππ not necessarily equal to π0 . The obtainable thermal power will be given by the difference of enthalpy of the products and reagents at the same conditions of ππ , π0 . Fig. 3.7 - Adiabatic combustion followed by isobaric cooling The enthalpy balance expressed on a mass-basis, considering π = 0 and introducing the hypothesis of ideal gases, is therefore: π = ∑ ππ (β°π,π + Δβπ,ππ ) − ∑ ππ (β°π,π + Δβπ,ππ ) π∈π π = ∑ ππ β°π,π − ∑ ππ β°π,π + π∈π π∈π π∈π ππ 0 (π)ππ ∑ ππ ∫ ππ,π π 0 π∈π (3.33) ππ 0 − ∑ ππ ∫ ππ,π (π)ππ π∈π π0 (3.34) where all the terms of Δβ can in part be simplified by introducing the hypothesis that the sum of the heat capacities of the reagents and products is coincident at every temperature in the interval π0 ÷ ππ . With these hypotheses, the previous relation is simplified as: π = ∑ ππ β°π,π − ∑ ππ β°π,π π∈π (3.35) π∈π That depends only on the enthalpies of formation of the species involved in the combustion reaction. 29 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 The extractable thermal power is called heating value π»π. It is equal to – π and is the thermal power that can be obtained by cooling the flue gas from the adiabatic flame temperature up to ππ : π»π = −π = π»π − π»π = ∑ ππ β°π,π − ∑ ππ β°π,π π∈π (3.36) π∈π ππ΄πΉ,ππ π»π = πππ’ππ (πΌ + 1) ∫ ππ0 (π)ππ (3.37) ππ This reaction heat depends on the assumptions on the thermodynamic state of the products and reagents and, in particular, on the state of aggregation of the water in the exhaust gases. It does not depend on the degree of dilution of the oxidant since the species not involved in the reaction give no net contribution in terms of β°π ; the dilution has an effect only on the adiabatic flame temperature, whose reduction is balanced by an increase in πΌ. Conceptually, the reaction heat depends on the temperature ππ since the sum of the heat capacities of the reagents and products are not exactly the same, but this change is modest. Depending on whether or not the enthalpy of condensation of the water in the combustion products is recovered, the following can be defined: • Lower Heating Value (LHV), defined as the heat that can be obtained from the unit of mass with a path that considers: 1. a complete combustion of the unit of mass of the fuel, without heat transfer (adiabatic combustion), starting from reagents at a reference temperature ππ (by convention, for natural gas, ππ = π0 =25°C6) until combustion products are obtained at the adiabatic flame temperature ππ΄πΉ,ππ . 2. an isobaric cooling of the combustion products from the adiabatic flame temperature ππ΄πΉ,ππ up to the reference temperature ππ , assuming that all the H2O in the combustion products is at the vapor state πΏπ»π = (πΌ + 1) ∫ ππ΄πΉ,ππ ππ0 (π)ππ = π»π − π»π (3.38) ππ in which the coefficient πΌ must be greater than the stoichiometric one, and π»π is calculated using β°π,π»2 π(π) . • Higher Heating Value (HHV), defined as the heat that can be obtained with the same path described above for the LHV, with the only difference that in this case it is assumed that all the H2O in the combustion products is at the liquid state. ππ΄πΉ,ππ π»π»π = (πΌ + 1) ∫ ππ ππ0 (π)ππ + ππ»2 π ββππππ,π»2 π(π0 ) ππΉ (3.39) in which π»π is calculated using β°π,π»2 ππ . Both hypotheses at the base of the definition of LHV and HHV are unrealistic (and therefore unreproducible in practice). If the combustion products are cooled to 25°C at the pressure of reference 6 Recent European convention: the reference was previously at 15°C. 30 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 (usually atmospheric), the content of H2O will be partially at the vapor state and partially at the liquid state, with a mole fraction in gaseous mixture given by the condition, π₯π»2 ππ π0 = ππ ππ‘(π0 ) . Both LHV and HHV are usually calculated from the chemical composition of the fuel once the thermodynamic properties of all components are known. For instance, to determine the heating value of natural gas, the starting point is a gas chromatograph analysis that identifies its chemical composition. The LHV is then calculated using the tabulated values of the enthalpies of formation. As an alternative, if the composition of the fuel is unknown, these values can be obtained experimentally with calorimetric experiments that use: • Mahler bomb calorimeter for solid or liquid fuels. It entails the total combustion of a measured amount of fuel in an airtight constant volume container immersed in water. The reaction is triggered in an atmosphere of pressurized oxygen. When the system reaches the balance, the water temperature is measured and the value of heat removed is obtained. • Junker gas calorimeter for gaseous fuels. It involves the combustion of a certain measured flow rate of fuel in an open system and counter-current heat exchange with a measured flow rate of water. These experiments obtain the LHV value since they are obtained with a large excess of fuel and, therefore, the water produced has an extremely small partial pressure that is less than the saturation pressure at temperature π0 . Therefore, there are no condensation phenomena. The HHV is obtained from the LHV analytically by adding the latent enthalpy of vaporization of the water. 3.3.5 Recuperative preheating It is now important to point out that the combustion heat transfer is not limited to the adiabatic flame temperature ππ΄πΉ,π0 achievable with ππ = π0 . Because this thermal energy, whether it is the LHV or HHV, can be made available at very high temperatures thanks to the recuperative preheating of the reagents. It is assumed that: • the sum of the heat capacities of the two reactants is equal to that of the products of combustion in the temperature range π0 ÷ ππ • the heat exchanger (recuperator) that preheats the reactants by cooling the combustion products has an infinite surface • the heat released from the combustion does not change with the increase of ππ The recuperative preheating of the reactants produces an increase in the adiabatic flame temperature, and increases both the terms that define the mean temperature of the heat made available to a power production plant, if any. So thanks to the possibility of the recuperative preheating, the energy of a fuel can be generated at a desired high temperature. Therefore, the efficiency limits with which it is converted into mechanical work are not tied to intrinsic irreversibilities of the combustion process, but to technological limitations of the power cycle. 31 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 Fig. 3.8 - Recuperative preheating of the reactants of an adiabatic combustion It is assumed there is a reversible cycle capable of producing work by exploiting the cooling of the exhaust gases from the temperature ππ΄πΉ,ππ up to ππ . The maximum work that can be produced by a source not in thermodynamic equilibrium with the reference state is the exergy of the flow, as previously shown. In the case where ππ = π0 the reversible cycle is a trilateral cycle. With the hypothesis of constant heat capacity and isobaric process, the reversible power that can be produced is therefore only a portion of the thermal power available: ππππ£(ππ =π0 ) = π»π»π (1 − π0 ππππ(ππ΄πΉ,π0 |π0 ) ) (3.40) Instead, in the case where the reactants are preheated, the producible power will be higher. Indeed, the HHV does not depend on ππ , whereas the efficiency of the trapezoidal cycle increases, increasing the mean log temperature of heat introduction ππππ£(ππ ) = π»π»π (1 − π0 ππππ(ππ΄πΉ,π π ) |ππ ) (3.41) For a preheating tending to infinity, a conversion efficiency equal to 100% is ideally obtained. In reality, this does not happen because in the combustion reaction chemical species, whose absolute entropies should be properly considered, were formed and destroyed, as will be explained in the next chapter. 32 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 Fig. 3.9 - Increase of the mean extraction temperature of the thermal power when the recuperative preheating of the reactants increases 3.3.6 Entropic balance for a reacting system Now an entropic balance of the general combustion reaction can be made considering a general production of entropy by irreversibility and the fact that the heat released by the system to the environment (negative) involves an increase in entropy of the same environment: βππππ + ππ + π − ππ = 0 π0 (3.42) By applying the entropic balance on a mass-basis to a combustion reaction, the following will be obtained: π + βππππ = ∑ ππ π ππ (π, π) − ∑ ππ π π (π, π) π0 π∈π (3.43) π∈π where the entropies must be calculated as π (π, π) = π 0 + Δπ (π,π) , taking into account the enthalpy of reference π 0 and Δπ (π,π) due to the deviation from the thermodynamic state of reference. What was said before referring to the enthalpy balance of reactant systems is also valid for the entropies since also in this case the choice of the reference value cannot be arbitrary and should be made according to well-defined considerations. The reference entropy for a compound involved in a reaction must be defined in accordance with the third law of thermodynamics. Based on this law, all substances that show a pure crystalline structure at a temperature of zero Kelvin have null entropy. On the contrary, the entropy will be greater than zero.The reference value for the entropy is called absolute entropy and is given by two contributions: the increase in entropy of heating and that of the phase change starting from zero Kelvin up to standard reference conditions π0 , π0 . 33 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 Fig. 3.10 - Qualitative representation of the definition of reference entropy at zero Kelvin So, for example, if the component can be considered an ideal gas at the reference state, it turns out: π π (π0 , π0 ) πππ’π ππ π 1 (π) ππ π 2 (π) ββππ’π ββπ‘π =∫ ππ + +∫ ππ + π ππ‘π π πππ’π 0 ππ‘π πππ£π π π (π) π0 π π (π) ββππ£π π0 π π +∫ ππ + +∫ ππ − π π ππ ( ) π πππ£π π ππ ππ‘(πππ£π) πππ’π πππ£π ππ‘π (3.44) where for greater clarity the specific heats of solid, liquid and gas, respectively, are indicated with ππ π 1 , ππ π 2 ,ππ π , ππ π . The absolute entropy of a general compound at state π, π is therefore equal to: π (π,π) = π π + Δπ (π,π) (3.45) Since both the reactants and the products are mixed and considering it an ideal mixture of ideal gases, it has to be considered that every species is not at pressure π, but at the partial pressure ππ . The corrective term in molar basis becomes Δπ (π,ππ ) , and will take the form of: π π π (π) π Δπ π,(π,ππ ) = ∫ π0 π π π π (π) ππ π ππ π ππ − π π ππ ( ) = ∫ ππ − π π ππ ( ) − π π ππ ( ) π0 π π0 π π0 (3.46) π where the term −π π ππ ( π ) = −π π ππ(π₯π ) is the mixing entropy. π By applying the entropic balance in mass form to a combustion reaction in which the products are brought back to the inlet condition of the reactants and with the hypothesis of ideal gases, the following will be obtained: π + βππππ = ∑ ππ π π π − ∑ ππ π ππ + π0 π∈π π∈π 34 of 124 (3.47) Second-law analysis of power cycles - Energy Conversion A – V7.0 ππ π π (π) π + ∑ ππ ∫ π π0 π∈π − [∑ ππ π∈π ππ π π (π) π ππ − ∑ ππ ∫ π0 π∈π π ππ − ππ π π’ ππ π π’ ππ ( ) − ∑ ππ ππ ( )] πππ π0 πππ π0 π∈π Moreover, with the assumption that the sum of the heat capacities is equal for the reactants and the products in the field π0 ÷ ππ , it is found that the entropy balance does not depend on the temperature of the reactants ππ . 3.3.7 Irreversibility generated in combustion When applying the entropy balance to an adiabatic reactor, the reaction products are obtained at the adiabatic flame temperature. The previous relation makes it possible to calculate the entropy variation generated by the combustion process: βπππππ = ∑ ππ π π π − ∑ ππ π ππ + π∈π + ∑ ππ ∫ π∈π ππ΄πΉ,ππ π0 − [∑ ππ π∈π π∈π ππ ππ π (π) ππ π (π) ππ − ∑ ππ ∫ ππ − π π π0 π∈π (3.48) ππ π π’ ππ π π’ ππ ( ) − ∑ ππ ππ ( )] πππ π0 πππ π0 π∈π That is to say, the difference in entropy is due to a term which depends only on the nature of the chemical species, regardless of pressure and temperature, one term that depends on the composition π₯ and on the temperature, and one that depends on composition and pressure: βπππππ = βππ₯π + βπ(π₯,π) + βπ(π₯,π) (3.49) Introducing the hypothesis of equal heat capacities, the term dependent on the temperature can be written as: ππ΄πΉ,ππ βπ(π₯,π) = ∑ ππ ∫ ππ π∈π ππ π (π) ππ π (3.50) which tends to zero for ππ → ππ΄πΉ,ππ , that is for very high recuperative preheating. 3.3.8 Reversible work and chemical exergy After combining the enthalpy and entropy balance applied to a system that cools the reaction products up to the temperature of the reactants ππ , it is found that: π − π = π»π − π»π = ∑ ππ βπ,(ππ ,π0 ) − ∑ ππ βπ,(ππ ,π0 ) π∈π π∈π { π + βππππ = ππ − ππ = ∑ ππ π π,(ππ ,π0 ) − ∑ ππ π π,(ππ ,π0 ) π0 π∈π π∈π 35 of 124 (3.51) Second-law analysis of power cycles - Energy Conversion A – V7.0 where in the case of a globally reversible process βππππ = 0 the following turns out: ππππ£ = (π»π − π»π ) − π0 (ππ − ππ ) (3.52) where the term π»π − π»π is the maximum thermal power that can be extracted by the combustion reaction and therefore it is equal to the π»π»π. The maximum work is therefore different from the π»π»π due to the change in entropy between reactants and products. Indeed, it depends on the absolute entropies and the terms of pressure, and represents the production of entropy due to the combustion process. Nevertheless, this correction is small and the numerical value of the ππππ£ is, for most of the fuels in question, similar to the π»π»π. This is because the reaction heat can be made available at very high temperature with a recuperative preheating. Thus, it can be ideally converted into work with yields tending to the unit. The reversible work is also equal to the difference of exergy between reactants and reaction products obtained by bringing the products back to the inlet thermodynamic conditions of the reactants, and without them changing their chemical composition or without them being reversibly remixed in the environment. In some texts, the term π» − π0 π takes the name of physical exergy, in the sense that it represents the exergy in a certain thermodynamic state without considering a chemical composition change. The reversible work can therefore be written as: πβ πβ ππππ£ = π΅π − π΅π (3.53) In literature, the common approach for reversible work calculation considers the contribution of each reactant and product as if they are at standard pressure and physically separated from each other. In this case, the difference in entropy due to mixing is null, and ππ − ππ can be calculated from the standard entropy of formation of the species sa . The results found up to this point are valid if the fact that both reactants and products are not in chemical equilibrium with the environment is neglected. For example, in the case of stoichiometric π combustion of methane with pure oxygen, the mole ratio 2 is equal to 2. Thus, the mole fraction of πΆπ»4 the oxygen in the reactants is equal to 66% versus 21% in the ambient air. It is thus feasible to use an ideal membrane to separate the oxygen from the environment at a pressure of 0.21π0 and then compress it with an isotherm compressor up to pressure 0.66π0 . The work executed is equal to the heat transferred to the environment that, considering the isothermal process, can be calculated on a mass-basis as: π = π = π0 βπ = −π0 ππ2 ππ ,π ππ ,0 π₯π ,π π π’ π π’ [ππ ( 2 ) − ππ ( 2 )] = −T0 ππ2 ππ ( 2 ) πππ2 π0 π0 πππ2 π₯π2 ,0 (3.54) which has a negative value and represents a spent work. Likewise, it is possible to separate the combustion products through ideal membranes selective for a single chemical species, and then using isothermal machinery (turbines and compressors) to take these compounds from the condition of partial pressure in burnt gases to the partial pressure they have in the atmosphere. In this way, it is possible to reversibly emit combustion products into the atmosphere. 36 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 By expressing the flow rate of each species as ππ =ππΉ (πΌ + 1)π¦π , the specific additional power given by the whole of the isothermal and reversible compressors and turbines obtained is: βππππ£,πππ₯ = ππ (πΌ + 1) β π0 [∑ π¦π π∈π ππ,π π π’ ππ,π π π’ ππ ( ) − ∑ π¦π ππ ( )] πππ ππ,0 πππ ππ,0 (3.55) π∈π where a positive work is achieved if ππ,π > ππ,0 (for the products) and if ππ,π < ππ,0 (for the reactants). The sum of the reversible work and of these contributions is called chemical exergy change in the fuel: ππππ£ + βππππ£,πππ₯ = βπ΅πβ = π΅π πβ − π΅ππβ (3.56) It represents the maximum useful work that can be extracted from the combustion of a unit of mass of fuel bringing components back into not only thermodynamic equilibrium (same temperature and pressure), but also chemical equilibrium with the environment (same composition). It can be noticed that the assumption of fuels and combustion products as perfectly mixed or as separated from each other at standard pressure changes the value of ππππ£ but not of βπ΅πβ . Fig. 3.11 - Reversible mixing of the combustion products in the ambient air achieved by adopting semipermeable membranes and reversible isothermal expanders/compressors. 3.3.9 Comparison of indexes The values of LHV, HHV, ππππ£ and βπ΅πβ are very similar to each other, and it can be noted that: π» • the difference between HHV and LHV is greater the higher the ratio of the fuel. The two πΆ indexes are equivalent for graphite. • the difference between ππππ£ and βπ΅πβ is small since it depends only on the modest contribution of the powers that can be obtained from the reversible mixing of the different compounds with the ambient air. This difference is less than 1% for methane. • ππππ£ falls between the LHV and the HHV for the hydrocarbons. This is because, as already stated, the combustion heat can be made available at as high a temperature as one likes through recuperative preheating, thus exploitable at yields tending to the unit. • Since the enthalpy of condensation of water is released at low temperatures, it has a low exergy content, and hence generally βπ΅πβ < π»π»π 37 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 • • The chemical exergy is hardly used since the technological solution required for taking the combustion products to chemical equilibrium with the ambient air is extremely costly: infinite surface membranes, infinite intercooled compression and infinite reheated expansion. Furthermore, the membranes available are not similar to ideal components since they show high pressure drops. For the analysis of fossil fuel-powered thermodynamic plants, reference will therefore be made to the HHV or LHV for a first law analysis, and to ππππ£ for a second law analysis. 38 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 4 IRREVERSIBLE PROCESSES AND LOSS OF USEFUL WORK In this chapter the theorem of lost work is explained. The relationship between entropy generation and loss of mechanical work is proved and analyzed. 4.1 IRREVERSIBILITIES AND USEFUL WORK LOSSES In the last chapter the reversible works extractable from various energy sources such as the following were defined: • sources with infinite heat capacity and therefore isothermal: the reversible work is calculated using the Carnot efficiency; • flows not in equilibrium with the environment: the reversible work is calculated as the difference of the source exergy between the inlet and outlet conditions without considering variations in its composition. In the case of constant heat capacities, the efficiency of a trilateral or trapezoidal cycle can be used to calculate the reversible work; • reacting systems: the reversible work can be assumed to be similar to the HHV or the LHV, depending on the aggregation state of water in the reaction products. This assumption passes through the observation that the heating value can be extracted from the system at temperatures ideally tending to infinity using the recuperative preheating method. These indexes represent the thermodynamic limit that can ideally be reached using only reversible processes. Thus, they are the term with which it is possible to compare when the performance of a plant in a second law perspective has to be assessed. In fact, this limit can never be reached due to the presence of non-ideal and non-reversible components that involve losses of useful work. When there is an irreversibility, there is a loss of ability to produce useful work. To clarify this concept, the following two examples should be examined. In both cases, a heat sink with infinite heat capacity with which heat can be exchanged (taken or released) freely without variation in the ability to produce heat must be defined. π0 is the temperature of this heat sink. Starting from the ideal and reversible cases seen previously, now a cause of irreversibility in each of them is introduced: • an irreversibility of heat transfer in the exploitation of sources with infinite capacity • a pressure drop in the exploitation of a source with finite heat capacity not in equilibrium with the environment 4.1.1 First example of irreversibility: heat transfer Consider the following sequence of transformations (all reversible): • the heat transfer π1 between a source with constant temperature ππππ₯ and a fluid that evolves into a Carnot engine cycle between ππππ₯ and ππππ . • the working fluid converts part of the heat π1 into work ππ’ and transfers heat π2 to a heat sinks at constant temperature ππππ . • The minimum temperature may be higher than that of the environment π0 . Hence, the work obtained is equal to: ππ’ = π1 (1 − ππππ ) ππππ₯ (4.1) Now an irreversibility represented by a heat transfer is introduced. The heat π1 is transferred under finite temperature difference between the heat source at ππππ₯ and the working fluid of the Carnot cycle that receives it at temperature ππ₯ . 39 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 The comparison between these two systems will have to be made with thermal powers transferred to the heat sinks being equal, with at the most the addition of another system consisting of a sink at ambient temperature π0 . In this case the work of the Carnot cycle is equal to: ππ’ ′ = π1 (1 − ππππ ) ππ₯ (4.2) while the work loss caused by the irreversibility is: βWπ’′ = ππ’ − ππ’ ′ = π1 ( ππππ ππππ − ) ππ₯ ππππ₯ (4.3) π1 π1 − ) ππ₯ ππππ₯ (4.4) which can be rewritten as: βππ’ ′ = ππ’ − ππ’ ′ = ππππ ( The term in parentheses represents the total increase of entropy βπ caused by the irreversible process considered. It is equal to the difference between the entropy increase of the fluid that receives heat π π βππ = 1 and the entropy decrease of the source that releases heat βπβ = 1 : ππ₯ ππππ₯ βπ = βππ + βπβ = ( π1 π1 − ) ππ₯ ππππ₯ (4.5) And therefore βWπ’ ′ = ππππ βπ (4.6) However, this is not the real loss of useful work associated with the presence of an irreversible process. In fact, compared to the reversible case, the heat sink at temperature ππππ now receives a heat π2 ′ > π2 . The difference is equal to the term βππ’′ because it is equal to: π2 ′ = π1 − ππ’ ′ = π1 − (ππ’ − Δππ’ ′ ) = π1 − ππ’ + Δππ’ ′ = π2 + Δππ’ ′ (4.7) If one wanted to make a correct comparison between the two systems, the additional heat discharged to the sink at ππππ must be removed and discharged reversibly to a sink at temperature π0 . If ππππ > π0 , the additional heat discharged has an energy potential and can be transformed into work with a further reversible cycle between ππππ and π0 , getting a certain useful work that goes in part to compensate for the loss of work due to the introduction of heat with an irreversible process: ππ’ ′′ = Δππ’ ′ (1 − π0 ) ππππ (4.8) Then, if the comparison is made properly, keeping unaltered both the heat π1 transferred by the source ππππ₯ and the heat π2 received from the sink at temperature ππππ , it is found that the actual useful work is equal to: 40 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 Δππ’ = ΔWπ’ ′ − ππ’ ′′ = Δππ’ ′ [1 − (1 − π0 π0 )] = ΔWπ’ ′ ππππ ππππ (4.9) Remembering that βππ’ ′ = ππππ βπ Δππ’ = π0 βπ (4.10) Thus, the product of the sink temperature π0 at which the heat has no energy content for the increase of entropy (in this case, the only one) of the universe βπ. Irreversible heat transfer MC MC a) = + MC b) Fig. 4.1 - Irreversible path caused by a heat transfer under finite temperature differences 4.1.2 Second example of irreversibility: Isenthalpic throttling Now an irreversibility is introduced in the reversible path used to draw the definition of exergy: suppose interposing a valve upstream of the isentropic expansion, in which a process called "isenthalpic throttling" is performed. 2 The total enthalpy, defined as the sum of the static enthalpy and the kinetic energy β π = βπ + π2 , is conserved in an adiabatic duct without work exchange (π = 0 and π = 0), neglecting the variations in potential energy linked to the altitude. 41 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 Fig. 4.2 - First law balance for an isenthalpic valve Depending on the thermodynamic state of the fluid that experiences the isenthalpic throttling, different phenomena can be observed: • For an ideal gas, the isenthalpic process is also isotherm; • For a real gas below the Joule-Thompson inversion temperature at the given pressure (e.g. superheated steam or supercritical fluids), the throttling cools down the fluid; • For real gas above the Joule-Thompson inversion temperature at the given pressure (in particular for liquids) the throttling heats up the fluid. The static temperature of the fluid can be higher or lower in the downstream section because of the variations in kinetic energy due to the different pipe section and fluid density. The temperature is different also in the smallest section of the valve (nozzle throat), typically lower. If the phenomenon is analyzed more in detail considering the physical processes that take place, it can be observed that when the flow, starting from thermodynamic equilibrium conditions (1), meets the narrowing caused by the insertion of the valve plug in the duct, it undergoes an expansion. Consequently, the kinetic energy increases and the static enthalpy decreases up to the minimum section of the duct, called contracted section (C). This is valid for subsonic flows upstream of the valve, a situation that, however, covers all cases in real energy plants. The expansion takes place generally without large irreversible effects, so with modest increases in entropy, since the acceleration of a fluid or its expansion is simple to achieve. Subsequently, the flow encounters an abrupt enlargement in section, which causes important irreversible phenomena (boundary layer separations, vortices, etc.), and large increases in entropy until a new condition of thermodynamic equilibrium (2) is reached. In this phase, the fluid slows down and is compressed. Fig. 4.3 - Representation of the irreversible phenomena due to turbulence in the transformation of isenthalpic throttling for an ideal gas By inserting a throttling in the example shown in point 3.2 upstream of the isentropic turbine, a process that can be divided into three transformations can be obtained, as shown in Fig. 4.4. 42 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 The transformations are: 1) the irreversible throttling from π₯ to π¦ 2) the isentropic and adiabatic, and therefore reversible, expansion from π¦ to π΅ 3) the isothermal and reversible compression from π΅ to 0 passing through point π΄ Fig. 4.4 - Irreversible path given by an isenthalpic throttling in the exploitation of a current of gas not in equilibrium with the dead state The powers that are obtained in the two reversible machines with the hypothesis of ideal gas are: • • expansion power: πΜππ π΅ = πΜ (βπ¦ − βπ΅ ) = πΜ (βπ₯ − β0 ) compression power: πΜπΆππ = πΜ π0 (π π΅ − π 0 ) = πΜ π0 (π π¦ − π 0 ) while the reversible work in the previous case was πΜπππ£ = πΜ [(βπ₯ − β0 ) − π0 (π π₯ − π 0 )]. The expansion work does not change compared to the previous case. The compression work increases because the compression starts with a lower pressure. The loss of useful work (power) between the two cases is given by the difference between the compression works: βπΜ = πΜπ0 (π π΅ − π 0 ) − πΜπ0 (π π΄ − π 0 ) = πΜπ0 (π π΅ − π π΄ ) = πΜπ0 (π π¦ − π π₯ ) (4.11) This is only due to the term of entropy increase of the universe given by the irreversible process: Μ βπΜ = π0 βππ₯→π¦ (4.12) The result caused by the fluid dynamic irreversibility described above in terms of loss of useful work (power) is thus identical to that caused by the irreversibility of heat transfer explained previously: the product of the ambient temperature for the increase in total entropy (of the universe) caused by the irreversibility. 43 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 4.2 GENERAL DEMONSTRATION With reference to Fig. 4.5, the system S (consisting of fluids, machines, heat exchangers, etc.) is considered. A single assumption is made in a demonstration that is entirely general in nature: the system is initially in a condition of thermodynamic equilibrium I (with its thermodynamic properties identified), and it evolves toward a new condition of thermodynamic equilibrium F with a real process (irreversible). It is assumed that the transformation is adiabatic. This does not limit the general validity of the demonstration. If there is a heat transfer, it suffices to incorporate the elements that exchange heat with each other in the system S and consider the system as adiabatic again. The same sign convention previously defined (outgoing work and outgoing heat positive) is adopted. Moreover, the existence of an environment (a heat sink with infinite thermal capacity) at temperature π0 is assumed. If the first law of thermodynamics is applied to the isolated system, the work exchanged with the outside is equal to the variation of internal energy of the system: ππΉ = ππΌ − π → π = ππΌ − ππΉ (4.13) Whereas the second law of thermodynamics states that every real transformation of an isolated system involves an increase in its entropy: ππΉ ≥ ππΌ (4.14) Now consider bringing the system S from state I to state F with a reversible path. To do so, it is inevitable to introduce a change in the properties of the system. The environment at temperature π0 is introduced into the system considered. To make the path reversible, it is necessary to decrease the entropy, assuming a heat transfer which involves the availability of the environment. A certain amount of heat π∗ will be taken from the environment, which will decrease the entropy of the environment. π∗ will be equal to the additional work extracted from the system βππ’ , since the initial I and final F instants are not changed. W * + Fig. 4.5 - Diagram used in the general demonstration for calculating the power ideally recoverable from replacing an irreversible process with a reversible process The first law of thermodynamics applied to S is: π + βππ’ = ππΌ − ππΉ + π ∗ 44 of 124 (4.15) Second-law analysis of power cycles - Energy Conversion A – V7.0 where π∗ is the heat taken for "free" from the environment at π0 . An additional work term exchanged by the system π appears to maintain the same initial and final conditions. Having assumed a reversible path, the second law of thermodynamics leads to the immutability of the entropy of the universe: π∗ =0 π0 (4.16) βππ’ = π∗ = π0 (ππΉ − ππΌ ) = π0 βππππ (4.17) ππΉ − ππΌ + βππππ = ππΉ − ππΌ − And therefore ππΉ − ππΌ = βππππ Thus, it turns out that: The physical meaning of the equation is that, by replacing an irreversible process with a reversible process, with the same initial and final states, an additional work βππ’ is obtained. equal to the product of the ambient temperature and the increase in entropy of the universe caused by the irreversibility present in the irreversible process. In other words, every time the entropy of the universe increases, because of an irreversible process, an ability to produce mechanical work equal to βππ’ is lost. Now this general demonstration will be applied to the two previous examples. In both cases, the irreversible process will be replaced by one or more reversible processes and the increase of producible work will be calculated. 4.2.1 Heat transfer: replacement with a reversible engine + heat pump The first process described will be reconsidered. System S consists of sources ππππ₯ and ππππ , and the Carnot engine that operates between temperature ππ₯ and temperature ππππ . Initial state I switches to final state F, where the source at ππππ₯ has released the heat π1 , the heat sink at ππππ has received the heat π2 , and the work π is obtained. The process contains an irreversibility caused by the heat transfer between the source at ππππ₯ and the engine that receives it at ππ₯ . A completely reversible path can be imagined by replacing the heat transfer between ππππ₯ and ππ₯ . with a reversible engine that receives the heat π1 and transforms one fraction of it into mechanical work: ππ’ ′ = π1 (1 − ππ₯ ) ππππ₯ (4.18) The heat discharged to the starting engine operating between ππ₯ and ππππ is less than that of the initial diagram of the term ππ’ ′ . To comply with the condition that the final state F is the same, it is necessary to re-establish the heat π1 that enters the starting engine altogether, and assume a heat intake from the environment at temperature π0 . To transfer this heat at temperature ππ₯ , a reversible heat pump is needed, which requires a mechanical work ππ’ ′′ equal to: ππ’ ′′ = ππ’ ′ (1 − π0 ) ππ₯ (4.19) The additional work achieved by replacing the irreversible process with a reversible process is equal to: 45 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 βππ’ = ππ’ ′ − ππ’ ′′ = ππ’ ′ − ππ’ ′ (1 − π0 π0 ) = ππ’ ′ ( ) ππ₯ ππ₯ (4.20) Which can also be rewritten as βππ’ = π1 (1 − ππ₯ π0 π1 π1 ) ( ) = π0 ( − ) ππππ₯ ππ₯ ππ₯ ππππ₯ (4.21) where the term in brackets is the difference of entropy variation of the heat sink at temperature ππππ₯ and that of the tank at temperature ππ₯ βππ’ = π0 ( π1 π1 − ) = π0 (βπππ₯ − βπππππ₯ ) = π0 βπ ππ₯ ππππ₯ (4.22) Which is what was set out to be proven. MC’ ’ W MC MC PdC Fig. 4.6 - Replacement of the irreversibility of heat transfer with a series of reversible processes that preserve the initial and final conditions of the sources 4.2.2 Throttling: replacement with reversible isothermal expansion + heat pump Now the irreversible process of isenthalpic throttling that occurs in the valve is considered. In this case the system S is represented by the supply duct to the valve, the valve itself and the exhaust pipe. The initial state I is with the fluid at a standstill, the final F state requires that a mass π of air be passed from the inlet section π₯ to the outlet section π¦. 46 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 To make the process from I to F reversible, it can be assumed to replace the irreversible adiabatic expansion with a reversible isotherm expansion at temperature ππ₯ , which occurs in a reversible machine producing a work equal to the absorbed heat, that is: ππ’ ′ = π∗ = ππ₯ βππ₯→π¦ (4.23) This work is equal to the “dissipated work due to fluid friction”, or simply “work of fluid friction” in the irreversible isenthalpic throttling process. It is equal to the work that would have been obtained with an isentropic expansion between the ππ₯ and the ππ¦ . π¦ π¦ ππ€ = ∫ ππ₯ ππ = ∫ πππ π₯ π₯ Reversible isothermal expander Fluid-dynamic irreversibility (w a ) HP a) b) + Fig. 4.7 - Replacement of the throttling irreversibility with a series of reversible processes that preserve the initial and final conditions of the sources The heat π∗ that must be supplied at temperature ππ₯ can be taken from the environment at temperature π0 by a heat pump requiring work equal to: ππ’ ′′ = π∗ (1 − π0 π0 ) = ππ’ ′ (1 − ) ππ₯ ππ₯ (4.24) In terms of work, the net result is given by: βππ’ = ππ’ ′ − ππ’ ′′ = ππ’ ′ [1 − (1 − π0 π0 )] = ππ₯ βππ₯→π¦ = π0 βππ₯→π¦ ππ₯ ππ₯ (4.25) It is noticed that βππ’ , obtainable by making the process reversible, is less than the ππ€ dissipated by π the isenthalpic throttling according to the ratio 0 . This is because not all the irreversibility produced ππ₯ by a valve can be recovered, and the recoverable portion depends on the temperature at which the irreversibility takes place. It follows that: 47 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 • • • • • • An isenthalpic throttling involves the dissipation of a mechanical work equal to ππ€ = ππ₯ βπ; This dissipated work remains within the fluid found at the end of throttling with a temperature higher than it would have had following an isentropic expansion and, therefore, with a greater possibility of providing work through downstream processes; In both cases, by bringing the source back to conditions of equilibrium with the environment through a sequence of reversible processes, the work lost forever is equal to π0 βπ, regardless of the temperature at which throttling takes place; A part of the work dissipated by friction can therefore ideally be recovered with a series of reversible processes. Its extent depends on the temperature at which the irreversibility was introduced, ππππ₯,πππ = ππ€ − βππ’ = βπ(ππ₯ − π0 ); It is always preferable to introduce irreversibility at high temperatures because the wasted work can then be exploited more efficiently by the downstream processes; ππππ₯,πππ can be fully recovered if the downstream processes are all reversible, otherwise only a part of it will be recovered. This result can be physically described according to two ways of thinking: a) According to the diagram illustrated above: the replacement of the irreversible phenomenon with one that is reversible requires an introduction of heat, which is not "free" because it is required at a temperature ππ₯ > π0 . The work required to restore it must be subtracted from the reversible expander. This work tends to zero the more ππ₯ tends to π0 , and it is larger the higher the temperature where irreversibility takes place is. b) According to the concept of “recovery”: the fluid "sees" the work dissipated due to friction as heat introduced at temperature ππ₯ . This heat has an energy value: if a reversible process is considered, a mechanical work can be recovered from it. This work is equal to a fraction of the work of fluid friction expressed by the efficiency of a Carnot cycle between ππ₯ and π0 . Obviously, the two ways have identical results: both clarify the physical meaning of the term βππ’ , which is the loss of ability to produce work caused by an irreversibility, imagining that the rest of the transformations (in this case: the heat pump or the Carnot recovery cycle) is reversible. In other terms, when a work of fluid friction ππ€ is dissipated, βππ’ is always and in any case lost, while the difference (ππ€ − βππ’ ) can be recovered totally or in part, depending on the subsequent transformations the fluid undergoes. 48 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 5 THE MOST COMMON CAUSES OF IRREVERSIBILITY In this chapter, a list of irreversibilities is presented. In power production plants, multiple forms of irreversibility can take place. The most important are: • heat transfer • irreversible fluid dynamic phenomena • combustion reactions 5.1 5.1.1 LOSSES IN HEAT TRANSFER General definitions Adiabatic efficiency The adiabatic efficiency of the heat exchanger is defined as the ratio between heat introduced into the cold flow and the heat transferred altogether from the hot flow. πππ = πΜπππ π πΜπ πΜβ − πΜπππ π = =1− πΜβ πΜβ πΜβ (5.1) This efficiency depends on the temperature difference between the hot fluid and the environment, and on the insulation of the component. High temperature heat exchangers have always an adiabatic efficiency lower than one, while condensers can be considered adiabatic since they work at temperatures similar to the one of the environment. Effectiveness The heat exchanger effectiveness is defined as the ratio between the actual heat exchanged and the heat ideally exchangeable with an infinite surface as ε= πΜ πΜS≡∞ (5.2) The effectiveness is also defined as the ratio between the actual temperature difference and that obtainable with an infinite surface heat exchanger, referred to the fluid with lower heat capacity. Excluding the cases in which the pinch point is inside the heat exchanger, if the heat capacity of the cold flow is lower than that of the hot flow, the pinch point will be at the hot side of the heat exchanger. If the surface of the heat exchanger is infinite, the cold flow will reach the inlet temperature of the hot flow. Assuming that the heat capacity of the cold flow doesn’t change considerably with the temperature, the effectiveness can be calculated as π= πΜπ ππ,π (ππ,ππ’π‘ − ππ,ππ ) (ππ,ππ’π‘ − ππ,ππ ) βππ = = πΜπ ππ,π (πβ,ππ − ππ,ππ ) (πβ,ππ − ππ,ππ ) βππ,π≡∞ whose representation is depicted in the following figure 49 of 124 (5.3) Second-law analysis of power cycles - Energy Conversion A – V7.0 Fig. 5.1 - Graphic representation of the effectiveness of heat exchangers with pinch point on the hot side (left) and on the cold side (right) If the pinch point is inside the heat exchanger, attention should be paid to the calculation of the exchangeable heat with ε=1. This is the case of a Once-Though heat exchanger for preheating and evaporation, or a supercritical heat exchanger, or those cases in which the mean heat capacities of the fluids change drastically while adopting an infinite surface heat exchanger. The case of an O-T heat exchanger is represented in the following figure Fig. 5.2 - T-Q Diagram of a O-T heat exchanger for a simple (left) and complex (right) molecule In the base case, only preheating and evaporation occurs. While in an infinite surface heat exchanger, also superheating is realized. In the case of simple molecule, the heat capacity of the vapor is much lower than that of the liquid phase. This makes the mean equivalent capacity of the cold side lower than that of the hot side. Thus, the βπ=0 condition is at the hot side. On the other hand, in the case of complex molecule, the heat capacities of liquid and vapor phase are similar, and the pinch point is placed at the beginning of the evaporation phase. Note that even with ε=1, it is impossible to obtain a reversible heat exchange. Heat exchange surface calculation Let’s consider a heat exchanger crossed countercurrent by two flows in which the heat is spontaneously transferred from the hot fluid to the cold fluid. The heat transfer surface π΄ of a heat exchanger can be calculated using the relation: πΜ = ππ΄βππππ 50 of 124 (5.4) Second-law analysis of power cycles - Energy Conversion A – V7.0 where π is the global heat transfer coefficient obtained from the transfer coefficients of the hot and cold flow, and the thermal resistance of the metal wall and the fouling. Reference is generally made to the internal or external surface of heat transfer, and the ratio between them must be taken into account, especially in cases of finned tubes. π πππ’π,ππ₯π‘ 1 1 1 | = + π πππ’π,πππ‘ + π π€ + π΄ππ₯π‘ + π΄ π πππ‘ βπππ‘ βππ₯π‘ ππ₯π‘ π΄πππ‘ (5.5) π΄πππ‘ Instead, the mean log temperature difference in the case of two currents with constant heat capacity can be calculated as: βππππ = βπ1 − βπ2 βπ1 ln ( βπ2 (5.6) ) which tends to zero if the temperature difference is canceled out inside or at one of the two ends of the exchanger. If the heat capacities are not constant, the parameter βππππ cannot be calculated this way, but it is necessary to divide the exchanger into a number of subsystems in order to be able to consider the heat capacities of the constant flows in each one. This approach is fundamental in the case in which the pinch-point is inside the heat exchanger and not in one of the two ends. For each subsystem, the surface required for heat transfer can be calculated and the total value of βππππ can then be calculated as: βππππ = πΜ ∑π(ππ π΄π ) (5.7) One simple case is the primary heat exchanger of a non-regenerative saturated Rankine cycle, in which the heat exchanger must be divided into an economizer and an evaporator. It must be considered that in the two components the transfer coefficients may be different. A more complex case is represented by the exchanger of heat introduction of a supercritical cycle. In this case, as shown in Fig. 5.3, the division of the heat exchanger into a high number of sections is required to properly locate the pinchpoint and calculate the surface. T T Q Q Fig. 5.3 - Division of a heat exchanger into several sections for the correct estimation of the transfer surface 51 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 5.1.2 Entropy generated in an irreversible heat exchange Whenever there is a heat transfer, a negative entropy variation of the hot current and a positive entropy variation of the cold current take place. Considering the infinitesimal process of introducing heat into the cold source at a certain temperature and assuming the process isobaric, it is found that: ππΜ = πΜππ (π) ππΜ ππ = π π (5.8) this represents an irreversibility, which gives rise to an increase in entropy. Once the TQ diagram is plotted for adiabatic heat exchangers crossed by two fluids, the entropy production can be calculated as: ππ,ππ’π‘ βπΜ = βππΜ + βπβΜ = + ∫ ππ,ππ πβ,ππ’π‘ ππΜ ππΜ +∫ π π πβ,ππ (5.9) where βππΜ is positive and βπβΜ negative. Fig. 5.4 - General notation for modeling an adiabatic countercurrent heat exchanger From which it is possible to extract the integral mean values 1/ππ,π and 1/ππ,β by applying the mean value theorem: βπΜ = πΜ πΜ (ππ,β − ππ,π ) − = πΜ ππ,π ππ,β (ππ,β ππ,π ) (5.10) The equation shows that the entropy production and the loss of useful work: • are proportional to the transferred heat πΜ (if one refers to the power instead of to the energy, the thermal power transferred) • are proportional to the mean difference in temperature between the two currents. The smaller the difference at each point of the TQ diagram for the two currents is, the lower the production of irreversibility is. At the limit for temperatures coinciding across the heat transfer, the integral mean values are coincident ππ,β = ππ,π and βπΜ = 0, but infinite transfer surfaces are necessary (this would also be true if the temperatures of the two currents were equal at one point only). • are inversely proportional to the square of the mean temperature at which the heat is transferred. To understand the physical meaning of this effect, one can think of a way to achieve the heat 52 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 transfer in a reversible manner by introducing an infinite number of reversible cycles interposed between the two currents that exchange heat with a certain temperature difference. Now, imagine to raise the mean temperature of the process, but keeping the same precise temperature difference. Since the efficiency of each infinitesimal cycle is equal to that of Carnot, its efficiency is inversely proportional to the mean temperature at which it receives heat. π =1− ππ πβ − Δπ Δπ =1− = πβ πβ πβ (5.11) For πβ → ∞, π → 0 and therefore it is impossible to recover reversible work from the heat transfer process, so the process is already irreversible. In practice, this aspect is little used since generally the heat source is at a certain temperature and it is not possible to raise it at will. An exception is the case of preheating in the combustion of fossil fuels. If the heat exchanger is not adiabatic, a further entropy production, linked to the heat exchange to the environment that receives it at a constant temperature π0 , is taken into account. In this case the loss of useful work is broken down into two terms: (i) one similar to what has been seen due to the heat transfer between the hot flow and the cold flow, (ii) the other linked to the thermal loss. The total increase in entropy is given by: Μ βπΜ = βππΜ + βπβΜ + βππππ = +∫ ππ,ππ’π‘ ππ,ππ πβ,ππ’π‘ ππΜ ππΜ πΜπππ π +∫ + π π π0 πβ,ππ (5.12) Μ where βππΜ and βππππ are positive and βπβΜ is negative. The heat exchanger is fictitiously divided into two different components by dividing the flow of hot fluid: in the first a flow of hot fluid equal to πΜ′β = πΜβ πππ that transfers heat πΜπ to the flow of cool fluid circulates, in the second a flow πΜ′′β = πΜβ (1 − πππ ) that transfers heat πΜπππ π to the environment circulates. Fig. 5.5 - General notation for modeling a non-adiabatic countercurrent heat exchanger 53 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 The increase of entropy of the universe of the two sub-components can be calculated as: (ππ,β − ππ,π ) (ππ,β ππ,π ) (ππ,β − π0 ) βπΜ′′ = πΜπππ π (ππ,β π0 ) βπΜ′ = πΜπ 5.1.3 (5.13) (5.14) Considerations on the design of the components These general considerations have a direct effect on the design sizing choices of heat exchangers of a power production plant. A heat exchanger is sized following a technical-economic optimization process that sees, on one hand, the desire to maximize the efficiency of the plant and, on the other, that of curbing the investment costs. To reduce the irreversibility of heat transfer, and thus raise the efficiency, it is necessary to minimize the temperature differences between the two currents. However, this leads to an increase in the transfer surfaces or an increase in the number of components in which the heat transfer process is divided (consider the regeneration of a saturated Rankine cycle).However, larger surfaces involve higher costs. Thus, the optimal solution is given by the trade-off between the increase in remuneration obtainable from greater efficiencies and the increase in the investment cost. For example, the analysis of the cost of electricity produced (LCOE – Levelized Cost of Electricity) is considered, and two plants based on the same technology with the same number of equivalent hours are compared. As the surface of a heat exchanger increases, on the one hand there will be an increase $ of the intercept (investment cost by cooling capacity factor (CCF)), and on the other, a ππβπ¦πππ $ reduction in the slope of the variable costs (increase of plant efficiency). The minimum LCOE ππβ can be obtained with one plant or another, depending on the proportion between the costs/benefits of the intervention. It should also be kept in mind that in many cases the increase in surface and cost reduces the irreversibility of heat transfer only to a minimum degree. For example, for two currents with a large difference in heat capacity, the reduction of the temperature difference on the cold side of the heat exchanger reduces the irreversibilities only minimally, and the advantages in terms of efficiency in the face of a very high increase in cost. In other cases, instead, the reduction of loss of heat transfer can also lead to penalizing the power produced. For example, consider the condenser of an air-cooled steam power plant. On the one hand, reducing the condensation pressure is expedient in terms of efficiency. On the other hand, however, the heat exchangers are more expensive with a consumption of the fan motors that can grow considerably, ending up reducing the actual benefit in terms of efficiency. 54 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 2 1 0 h Fig. 5.6 - Comparison in terms of LCOE of heat exchangers with larger surfaces. In case 1, the operation is not economically advantageous. It is, instead, in case 2. 5.2 LOSSES IN EXPANSION AND COMPRESSION Turbomachines will be handled differently depending on whether they process a fluid whose specific volume depends on the temperature and pressure (gas and vapor) conditions, or an incompressible fluid (pumps). 5.2.1 Gas/vapor turbomachines: turbines and compressors The operation of a turbine or an expander is defined by an adiabatic efficiency πππ on the adiabatic expansion as a whole, or by a polytropic efficiency π∞ or ππ¦ that takes into account the fluid dynamic quality of the infinitesimal process. It defines the ratio between the real and the ideal work for a whole expansion or infinitesimal process. Consider an adiabatic infinitesimal expansion between ππ₯ and ππ₯ − ππ. With the definition of polytropic efficiency ππ¦ , it is seen that the real work produced will be given by the difference of that ideally producible work along an isentropic expansion and that dissipated due to friction phenomena, which cause a production of irreversibility and an increase in entropy of the fluid. ππ¦ = ππΜπππ ; ππΜπππ£ ππ¦ = ππΜπππ£ − ππΜπ€ ππΜπππ£ (5.15) where the ideal work is equal to ππΜπππ£ = −πΜπ£ππ (considering that the ππ < 0) and friction work is equal to ππΜπ€ = πππΜ ππΜπ€ = πππΜ = ππΜπππ£ (1 − ππ¦ ) = −πΜπ£ππ(1 − ππ¦ ) (5.16) For polytropic efficiencies equal to zero (isenthalpic throttling), the ideal work and the friction work are equal and there is the maximum production of irreversibility. By integrating the entire expansion, it turns out that: πππ’π‘ −πΜπ£ππ(1 βπΜ = ∫ πππ − ππ¦ ) π 55 of 124 (5.17) Second-law analysis of power cycles - Energy Conversion A – V7.0 The specific volume can now be expressed in terms of the compressibility factor like π£ = π βπΜ = − π π’ πππ’π‘ ππ πΜ(1 − ππ¦ ) πΜπ π’ πππ ∫ ππ = πΜ (Μ Μ Μ Μ Μ Μ Μ Μ Μ 1 − ππ¦ )ππ ( ) ππ πππ π π ππ πππ’π‘ π π’ π ππ π (5.18) Therefore, the loss of efficiency depends on the "fluid dynamic quality" (defined by ππ¦ ) of the turbine, and grows as the expansion ratio increases. It is independent from the temperature at which the expansion takes place since the infinitesimal increase of entropy ππ is proportional to the infinitesimal work, which in turn is proportional to π£ and π, but it is divided by π, which therefore cancels out. Another quantity called adiabatic efficiency πππ is often preferred to the polytropic efficiency. It considers only the inlet and outlet conditions of the fluid from the machine, but requires that the flow rate does not vary between inlet and outlet. The gas that is ideal and at specific constant heat is considered, and it is assumed that there is the condition for calculating the expansion temperature change given by a certain polytropic efficiency. If the specific heat is constant, then it can be written that ππ¦ = ππ πππππ πππππ πβπππ = = πβπππ£ ππ πππππ£ πππππ£ (5.19) If the isentropic πβ linked to each infinitesimal process is considered, the ideal work is found for ππ = 0 and it is equal to πβπππ£ = π£ππ Then, it can be written for an ideal gas that πβπππ ππ πππππ = πβπππ£ π£ππ π π ππ ππ | = ππ¦ π πππ ππ π ππ¦ = (5.20) (5.21) Which integrated between the inlet and outlet conditions gives: π π πππ’π‘ πππ’π‘ )| = ππ¦ ππ ( ) (5.22) πππ πππ ππ πππ πππ’π‘ | = π½−ππ¦ π (5.23) πππ πππ Then, the adiabatic efficiency can be defined as the ratio between the work produced and the isentropic enthalpy change from the inlet conditions. Again, considering the specific constant heats, this means that: ππ ( πππ = ππ π₯ππππ πππ (1 − π½ −ππ¦ π ) ββπππ = = ββπππ£ ππ π₯ππππ£ πππ (1 − π½ −π ) (5.24) For an expansion, the value of πππ is always greater than that ππ¦ due to the recovery phenomenon. This phenomenon can be explained with a discretization of the expansion process. Consider carrying out an expansion in very small pressure changes such that each one has an efficiency equal to the polytropic efficiency. The presence of an irreversibility in the expansion process leads to 56 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 an increase in entropy and to the dissipation of a certain work of fluid friction which remains within the fluid. Thus, at the end of each expansion the enthalpy is greater than the corresponding isentropic one, with ββ1 > ββ1′′ and π1 > π1′′ . Due to the divergence of the isobaric lines, it is evident that ββπππ£,1→2′ > ββπππ£,1′′ →2′′ in the second expansion , ββπππ£,2→3′ > ββπππ£,2′′ →3′′ in the third, and so on. 0 T 1 1’’=1’ 2 2’’ 2’ 3’’ 3 3’ 4 4’ 4’’ 5 5’ 5’’ s Fig. 5.7 - Graphic representation of the recovery work for a gas turbine When extending this observation up to the discharge pressures, the isentropic enthalpy change obtained from the sum of enthalpy changes along the real expansion ββπππ£,π¦ is greater than that calculated simply between inlet and outlet conditions ββπππ£,ππ ββπππ£,π¦ = ∑ ββπππ£,π→(π+1)′′ (5.25) ββπππ£ = ∑ ββπππ£,π ′′→(π+1)′′ = β0 − β5′′ (5.26) Since the specific work produced is: π€ = ππ¦ ββπππ£,π¦ (5.27) An adiabatic efficiency higher than the polytropic efficiency is obtained. πππ = where the term ββπππ£,π¦ ββπππ£,ππ π€ ββπππ£,ππ = ππ¦ ββπππ£,π¦ ββπππ£,ππ (5.28) > 1. In compressors, the effect is exactly the opposite. Recovery plays against the process since the fluid heats up, and it requires greater power to be further compressed due to the divergence of the isobaric lines. The real work is the sum of the ideal work and the wasted work due to irreversible fluid-dynamic phenomena. For this example, the sign notation for the work is changed: now it is positive if entering the system. 57 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 ππΜπππ ππΜπππ£ + ππΜπ€ 1 − ππ¦ ππΜπ€ = ππΜπππ£ ( ) = πππΜ ππ¦ ππ¦ = ππΜπππ ; ππΜπππ£ ππ¦ = (5.29) (5.30) The increase in entropy given by the compression process is therefore equal to: βπΜ = ∫ πππ’π‘ πππ 5.2.2 ( πππ’π‘ 1 − ππ¦ πΜπ£ππ 1 − ππ¦ πΜπ π’ π ) =∫ ( )π ππ ππ¦ π ππ¦ ππππ πππ Μ Μ Μ Μ Μ Μ Μ Μ Μ 1 − ππ¦ πΜπ π’ πππ’π‘ = πΜ ( ) ππ ( ) ππ ππ¦ Μ Μ Μ πππ (5.31) Hydraulic machines: Pumps The operation of a pump is defined by a polytropic efficiency π∞ or ππ¦ , which defines the ratio between the ideal work and the real work for an infinitesimal compression. In the case of incompressible fluid, the adiabatic efficiency πππ is equal to the polytropic efficiency. Considering that the temperature increase is negligible, the work of fluid friction is equal to: πΜπ€ = ∫ πππΜ = ππ₯ βπΜ (5.32) The ideal work is equal to πΜπππππ = πΜπ£βπ. It is positive (βπ > 0) and entering the system. Therefore, βπ can be calculated as: βπΜ = πΜπ€ πΜπππππ 1 − ππ¦ πΜπ£βπ 1 − ππ¦ = ( )= ( ) ππ₯ ππ₯ ππ¦ ππ₯ ππ¦ (5.33) Concerning the efficiency loss related to the fluid dynamic losses of the pumps, it can be highlighted that: • It is preferable to raise the mean temperature ππ₯ at which the pumping takes place. The physical meaning of this result is clear: since a work of fluid friction πΜπ€ has to be introduced (which does not depend on the temperature), it is preferable that the introduction take place at the highest possible temperature in order to have a lower βπ and to subsequently be able to recover a larger fraction of it. • Even if the effects of real liquid start to become noticeable as the temperature rises, the conclusions do not change since • The greater the increase in pressure is, the greater the irreversibility production. • As always when considering the fluid dynamic irreversibilities in machines, the losses are linked to the fluid dynamic quality of the components. • It is generally small since the specific volume is very small for all fluids 5.2.3 Considerations on the design of the components Regarding the turbomachinery, the close relationship between design and execution costs and efficiency of the machine is clear. For a turbine, high volumetric expansion ratios lead to supersonic flows which in certain conditions may create shock waves that penalize the fluid dynamic efficiency 58 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 of the machine. Dividing the expansion into many stages certainly allows these effects to be reduced and to obtain a greater polytropic efficiency on the face of a larger investment and higher machine inertia. A similar argument can be made for distributing the enthalpy change over multiple stages in order to limit the peripheral mean speed and stresses on the blades. The pumps instead should be positioned at the high temperatures in order to be able to recover a larger fraction of the work of fluid friction, but this entails pumps with a higher specific cost and an excellent compromise solution. 5.3 LOSSES DUE TO PRESSURE DROPS They involve a reduction in pressure in a specific component or in piping, and they may be localized (filters, valves, abrupt sections and direction variations, etc.) and distributed (ducts) pressure drops. Considering the isenthalpic throttling, it suffices that πβ = πππ + π£ππ: ππ = − π£ππ π (5.34) Since ππ < 0, the pressure drop involves an increase in entropy. It is more intuitive to refer to an absolute pressure drop and use the relation: ππ = π£ππ π (5.35) These drops will be treated differently for liquids and gases. 5.3.1 Pressure drops of a liquid A pressure drop in liquid phase causes an infinitesimal change in temperature. Considering the specific constant volume, it can be written that: βπΜ = πΜ π£βπ π (5.36) The equation shows that the pressure drops in liquid phase are paid more at low temperature and are, of course, proportional to the flow rate which sustains the drop and to the absolute pressure drops (regardless of pressure at which it takes place). Since the specific volume π£ is small, the influence on the efficiency of the cycle is small. In fact, large pressure drops (many dozens of bars) on the liquid side are accepted. In designing thermodynamic plants, the absolute pressure drops on the liquid side are generally kept constant. 5.3.2 Pressure drops of a gas A pressure drop in the vapor phase instead causes an increase in entropy equal to: π π’ πππ ππ ππ ∫ ππ πππ’π‘ π π π π’ πππ βπΜ = πΜ πΜ ππ ( ) ππ πππ’π‘ π π’ πππ’π‘ + βπ π π’ βπ βπΜ = πΜ πΜ ππ ( πΜ ππ (1 + ) = πΜ ) ππ πππ’π‘ ππ πππ’π‘ βπΜ = πΜ where for small pressure drops ππ (1 + βπ πππ’π‘ )~ βπ ππ 59 of 124 (5.37) (5.38) (5.39) Second-law analysis of power cycles - Energy Conversion A – V7.0 The equation shows that the pressure drops in the gas phase are more critical at low pressure since the βπ relative pressure drops count and are, of course, proportional to the flow rate which sustains the ππ drop and to the coefficient of compressibility πΜ . Since the specific volume is high, the influence on the cycle efficiency is greater than in the case of liquid. Careful attention must therefore be paid to the pressure drops, especially at low pressure. In designing thermodynamic plants, the relative pressure drops on the gas side are generally kept constant in order to not penalize low pressure configurations too much. 5.3.3 Considerations on the design of the components The pressure drops in a heat exchanger are another typical example of a technical-economic optimization. A shell and tube heat exchanger is considered. The analysis is done with the same number of tubes. When the diameter of the tubes is reduced, the metal mass is decreased due to two effects: (i) the first is that lower thicknesses can be used keeping constant the pressure and allowable ππ· stress difference as defined by the law of Mariotte = π; (ii) the second is that for decreasing cross2π‘ sections the fluid flows faster with higher transfer coefficients, thus, the surface required is reduced with the same differences in temperature and transferred heat. So, if on the one hand reducing the cost is possible, on the other, however, greater pressure drops are obtained. Indeed, they are related to the square of the speed with a consequent effect of loss of useful work due to production of irreversibility and increased consumption of the pump. 5.4 LOSSES DUE TO MIXING Mixings are considered isobaric and are reversible only if the flows that are mixed have the same temperature and equal chemical composition. One typical example is the mixing of liquid and vapor, both at saturation conditions. Three types can be outlined (Fig. 5.8): a) Two currents with different pressure and at the same temperature and composition: the irreversibility is given by the isenthalpic throttling sustained by the current at a higher pressure before being mixed; b) Two currents at different temperature: the increase of entropy of the universe is comparable to that of a heat transfer with a surface of infinite transfer; c) Different chemical composition: the process is irreversible even if it takes place at the same temperature and pressure since the partial pressures of the various fluids are considered. The increase of entropy of fluids that are mixed together at the same temperature and pressure is caused by the increase in specific entropy of both liquids, which is caused by a decrease of their partial pressure. In order for the mixing to take place reversibly, therefore without a total increase of entropy, it would be necessary to imagine semipermeable membranes, through which the fluids can mix without decreasing their partial pressure, as already seen for the definition of chemical exergy of fuels. A classic example is the desuperheating of the vapor that is performed with sprayed water in order to reduce the temperature of the live steam produced by the steam generator. From the entropic analysis point of view, it is better to carry out this process with hot water instead of with cold water. 60 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 T ππ¦ ππ₯ π¦ πΜπ₯ , ππ₯ , ππ₯ πΜπ₯ + πΜπ¦ , ππ₯ , ππ₯ ππ₯ < ππ¦ π₯ πΜπ¦ , ππ¦ , ππ₯ βπΜ = πΜπ¦ ππ₯ − ππ¦ s T ππ₯ πΜπ₯ , ππ₯ , ππ₯ πΜπ₯ + πΜπ¦ , ππ₯ , ππ§ ππ§ πΜπ¦ , ππ₯ , ππ¦ πΜπ₯ + πΜπ¦ ππ§ > πΜπ₯ ππ₯ Μ + ππ¦ ππ¦ ππ¦ Q Fluido A πΜπ₯ , ππ₯ , ππ₯ πΜπ₯ + πΜπ¦ , ππ₯ , ππ₯ , ππ§ Fluido B πΜπ¦ , ππ₯ , ππ₯ πΜπ₯ + πΜπ¦ ππ§ > πΜπ₯ ππ₯ Μ + ππ¦ ππ¦ Fig. 5.8 - Different types of irreversible mixing of two currents 5.5 LOSSES DUE TO CHEMICAL REACTIONS The chemical reactions of combustion bring with them a certain production of irreversibility depending on how the combustion itself takes place. Specifically, reference is made to the reaction of a unit of fuel mass starting from reactants at temperature ππ . The increase in entropy is due to the formation and destruction of chemical species and absolute entropies associated with them, to the partial pressures of the mixed components, and to the temperature. The possibility to minimize the increase of entropy of the universe connected with the sole combustion reaction with the recuperative preheating method has already been shown. 5.6 OTHER LOSSES: SELF-CONSUMPTIONS, AUXILIARY SYSTEMS These losses are simply electrical or mechanical powers that are spent for the operation of the auxiliary systems, such as indoor air conditioning, electrical consumption, safety and fire systems. These powers are simply considered as lost works βπΜπ . 61 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 6 APPLYING THE GENERAL FORMULA TO ANALYSIS OF POWER PLANTS In this chapter, the main framework for second law analysis in power systems is defined. This framework is identified in terms of parameters needed to perform a second law analysis. If a production plant is considered, and it is seen as broken down into an N series of components (or subsystems), the following formula can be written: π π Μ πΜπππ = πΜπππ£ − ∑ βπΜπ = πΜπππ£ − π0 ∑ βππππ,π 1 where • • • • πΜπππ : πΜπππ£ : βπΜπ Μ : βππππ,π (6.1) 1 is the real work; is the work that would be obtained with a series of reversible processes; are the N useful works lost due to irreversibility in the N components; are the N productions of entropy that occur in the N components considered. A series of elements has to be defined in order to properly apply this equation: 1. Definition of the time interval; 2. Definition of the physical (or conceptual) boundaries of the system; 3. Definition of the energy source; 4. Definition of the dead environment; 5. Definition of the efficiency of the power cycle; 6. Choice of the N processes producing irreversibility in which the system is schematized. 6.1 DEFINITION OF THE TIME INTERVAL The equation can be applied to: • instantaneous quantities (powers): useful for defining the specific rated powers of the single components and of the plant. This analysis is used to set the performance to be tested, performance to which important penalties/bonuses are tied when reached; • quantities integrated in the time interval (energies): useful for energy, economic and environmental balances, for technical-economic feasibility studies, and for final balance sheets; the most common time interval is the year, but often monthly, weekly, daily and hourly balances are carried out; • average quantities during the time interval (powers). 6.2 DEFINITION OF THE BOUNDARIES OF THE SYSTEM CONSIDERED For example, the following are identified: • flanges: real or virtual inlet and outlet flanges of all fluids involved in the process; • electricity inlets and outlets (gross power, pump and fan consumption); • auxiliary systems; • thermal loss to the environment (insulation, isolation). 6.3 DEFINITION OF THE ENERGY SOURCE AND OF MATERIAL FLOWS Flows of matter can enter and exit the borders of the system, and thermal, mechanical and electrical powers can be exchanged. The main flow to define is that of the energy source. Added to this one, there are the flows of matter entering and exiting the system, and the flows of electrical or mechanical power that are introduced or 62 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 generated by the system. One example is the consumption of the auxiliary systems, when it is not considered a plant self-consumption itself. As for the flows of matter, the work potentially obtainable between the inlet and outlet conditions must instead be considered. Therefore, the following shall be defined: • flow rate on a mass-basis; • chemical composition; • thermodynamic conditions at the inlet: πx ,πx ; • restrictions on the discharge conditions: ππππ may coincide with the dead state π0 . ππππ may also be higher, as often occurs for gases containing sulfur oxides or nitrogen, for which there is no desire to reach the dew point, or for geothermal fluids, for which there is a desire to prevent the deposit of compounds that can obstruct the flow in the re-injection sink; • Kinetic energy, π 2 /2 (the "total" quantities of the flow are usually considered, and in the case of wind energy it is the prevailing term); • Gravitational energy, ππ§ (important only in hydroelectric applications, of little importance for gases and vapors). 6.4 THE "DEAD" STATE The most common assumptions are: • ambient air at π0 , π0 : it is the most useful solution for open cycles that use the ambient air as the working fluid or for air-condensed closed cycles. • sea water, or gravitational water at π0 , π0 : it is a solution often adopted for closed cycles that have to transfer heat to the environment through a coolant. It can be assumed at infinite heat capacity or at finite heat capacity. As will be seen afterwards, these two assumptions are equivalent, and an additional mixing loss will have to be introduced if a coolant at finite heat capacity is used. • minimum cycle temperature: this is also a solution often adopted for closed cycles that have to transfer heat to the environment through a coolant; it coincides with the first for open cycles. • a suitable mean temperature of the coolant: this is the case of cogeneration plants in which the enthalpy of condensation is released to a thermal utility through a heat transfer fluid. 6.5 DEFINITION OF THE EFFICIENCY OF THE POWER CYCLE An efficiency is defined for plants that generate mechanical energy or electricity. It is obtained by dividing πΜπππ by an FE term that defines the energy input of the energy source. This FE parameter can be arbitrarily chosen in accordance with the type of energy source considered and the type of analysis one wants to make, and actually it does not affect the generality of the entropic analysis. The FE parameter can be chosen, for example, equal to the thermal power available at the plant inlet πππ , getting the expression of the entropic analysis applied to a thermodynamic efficiency of first law. π πΜπππ πΜπππ£ βπΜπ = −∑ πΜππ πΜππ πΜππ (6.2) ππΌ = ππππ£ − ∑ ΔππΌ,π (6.3) π=1 π i=1 One alternative is, on the other hand, to assume FE equal to πΜπππ£ of the system, getting an expression that instead refers to the second law. 63 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 π πΜπππ πΜπππ£ βπΜπ = −∑ πΜπππ£ πΜπππ£ πΜπππ£ π (6.4) i=1 ππΌπΌ = 1 − ∑ ΔππΌπΌ,π (6.5) i=1 In this case the efficiency ππΌπΌ represents the thermodynamic quality of the system with respect to the reversible thermodynamic limit, and it is an extremely useful index for comparing systems that exploit very different energy sources. 6.6 NUMBER N OF MODELED PROCESSES The choice of the N processes into which the system is divided is arbitrary and must be made based on technical and theoretical considerations. In the general approach, the system is divided into components, each of which is a source of an overall irreversibility that characterizes its operation. Nothing, however, prohibits unifying multiple components and analyzing them as a single process. One example is the steam generator of a combined cycle that is physically made up of different banks of heat transfer tubes (economizer, evaporator, superheater) but that can be handled as a single heat transfer process in which superheated vapor is produced by cooling a certain current of hot burnt gases discharged by the gas turbine. The opposite approach is instead that of dividing the processes as much as possible with the aim of giving each cause of loss the right weight. Therefore, a heat exchanger can be broken down into four different irreversible processes: (i) the actual heat exchange to transfer heat between the hot and cold currents under non-infinitesimal temperature changes, (ii) the pressure drops on the hot side, (iii) the pressure drops on the cold side modelled as isenthalpic throttling and (iv) the thermal loss to the environment. The recommended approach is the latter, the only one able to point out on which causes of irreversibility to focus in order to improve the efficiency of a thermodynamic cycle. 64 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 7 SECOND LAW ANALYSIS OF A STEAM RANKINE CYCLE In this chapter, the list of losses in Rankine cycles is shown. The framework presented in the previous chapter is first outlined, then the losses are evaluated and commented. 7.1 INITIAL DEFINITIONS Boundaries of the system considered Fluid inlet and outlet flanges, electricity inlets and outlets, auxiliary systems, etc. The analysis is applied only to the vapor cycle considering that all the auxiliary systems are powered inside the cycle, so the only electricity outlet is the interface with the net downstream of all the electrical losses. The system receives heat from the steam generator where combustion takes place. Dead state Sea water (or ambient air) of infinite heat capacity at T0 is assumed. The choice is discussed by illustrating the efficiency losses of the condenser. Reference to the rated instantaneous quantities The analysis is conducted referring to the instantaneous quantities, hence to the powers. Energy source Combustion of a given fossil fuel, whether it is solid (coal) or gaseous (natural gas), is considered as the energy source. The heat ideally extractable from the combustion is the HHV or the LHV, which can be converted into work with efficiencies ideally close to the unit thanks to regenerative preheating. The reversible work is very similar to the LHV (or HHV), so the entropic analysis applied to the first law efficiency is equivalent to that applied to the second law. Definition of efficiency The formula to be applied is: π πΜπ’ πΜπππ£ βπΜπ = −∑ πΜπΏπ»π πΜπΏπ»π πΜπΏπ»π π (7.1) 1 ππΌ = ~1 − π0 ∑ π=1 π₯ππΜ πΜπΏπ»π Number of modeled processes The following ten processes are identified: o βπ1 irreversibility of heat transfer in the condenser o βπ2 fluid dynamic irreversibility in the pumps o βπ3 irreversibility of heat transfer in the preheating line o βπ4 irreversibility of heat transfer in introducing heat into the cycle o βπ5 fluid dynamic irreversibility in the turbine o βπ6 pressure drops in the liquid phase o βπ7 pressure drops in the vapor phase o βπ8 thermal losses o βπ9 mechanical/electrical losses o βπ10 losses tied to consumptions of the auxiliary systems The Ts diagram of the general steam Rankine cycle is shown in Figure 7.1. 65 of 124 (7.2) Second-law analysis of power cycles - Energy Conversion A – V7.0 ππππ₯ 5 7 6 3 4 1≡2 8 π0 Fig. 7.1 – General Ts diagram for a Steam Rankine Cycle 7.2 βπΌπ IRREVERSIBILITY OF HEAT TRANSFER IN THE CONDENSER The condenser of a Rankine cycle with water vapor can use water or ambient air, or can transfer heat to a current of heat transfer fluid for thermal heating systems or thermal users in a backpressure cogeneration arrangement. In this latter case, the cold fluid is water at ambient pressure or pressurized if temperatures close to or higher than 100°C have to be reached. Only the first case with a cooling current in equilibrium with the environment is considered here. The current is remixed with the environment itself after having been heated in the condenser. It is assumed that the condenser is adiabatic. This assumption is reasonable since it operates at temperatures near ambient temperature. Thus, the losses, which depend on the difference in temperature between the hot fluid and the environment, are in fact negligible even without insulation. In the condenser, the fluid discharged by the turbine is condensed. A discharge from the turbine in two-phase or in superheated vapor can be achieved, depending on the type of plant, the evaporation pressure, the superheating temperature and the reheating, if any. In this latter case, the first part of the condenser desuperheats the fluid to the point of saturated vapor. The condenser is built following different architectures, based on the type of coolant: • water-based: it is built as a Shell&Tube heat exchanger (Figure 7.2), in which the cold water flows in the tubes and the vapor in the shell. The vapor comes into contact with the cold tubes, condensates and subcools. The drops of liquid fall onto the tubes underneath and meet with the vapor, performing a partial heat and mass transfer. Sufficient volume is guaranteed under the bank of tubes, in which the subcooled liquid has a residence time sufficient to return to equilibrium with the vapor and reach a saturated liquid condition. This volume is called heat sink of the condenser and its level is one of the control parameters of a Rankine cycle. There is usually a normal operation zone defined by a minimum level that has to always be guaranteed in order to prevent pump cavitation, and a maximum level that serves to prevent some tubes 66 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 from remaining submerged, which would result in an actual subcooling of the condensate. The water discharge temperature is limited by regulations protecting sea7 and river ecosystems. Fig. 7.2 - S&T condensers for vapor systems • Air: the vapor is distributed in different tubes finned on the outside to increase the heat transfer coefficient of the ambient air. The air is usually moved by forced convection with fans (suction or forced draft), while natural circulation towers can be used for large power plants. In the first case, depending on the tube class, condensate conditions that range from subcooled liquid (first class) to two-phase fluid (last classes) can be obtained at the outlet. In this case as well it is necessary to build a heat sink to re-balance the vapor liquid equilibrium condition after mixing the various flows. These condensers are commonly used in modern combined cycle plants and are shown in the following Figure (7.3). Fig. 7.3 - Air condensers for vapor systems • Evaporation tower: a water circuit is used as coolant and it is cooled in an evaporation tower, that removes heat through heat and mass exchange between the ambient air and the water sprayed by the nozzles. The evaporation towers can be open cycle (if the cooling water itself is that sprayed) or closed cycle. In both cases the cooling water is taken to the wet bulb temperature of the ambient air and thus it is possible to reduce the condensation pressure with less water consumption (Figure 7.4). 7 For the sea, the discharge must never be higher than 35°C and must not involve changes in temperature higher than 3°C at 1000 m from the outlet. For lakes, the discharge must never be higher than 30°C and must not involve changes in temperature higher than 3°C at 50 m from the outlet. For rivers, the mean temperature difference between any section upstream and downstream must be not higher than 3°C (1°C on the half-section). For canals, the maximum temperature at the outlet is 35°C. 67 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 Fig. 7.4 - Schematics of closed cycle (left) and open cycle (right) evaporation towers Subcooling is harmful for the efficiency of a power production plant. Indeed, it involves a lower inlet temperature in the preheating and, therefore, demands a higher flow rate bled from the turbine with reduced obtainable power. It always has to be limited with a proper sizing of the heat sink. The condensate extraction pump installed in the pit under the condenser controls the heat sink level. Air and water-based condensation will be discussed, while the evaporation tower will not be analyzed in detail since it considers not only heat transfer, but mass transfer as well The condensation phase mainly involves two irreversibilities: that of heat transfer and that of mixing the coolant discharged by the condenser into the environment. Fig. 7.5 - TQ diagram of the heat transfer process to the environment and diagram of the irreversibilities considered • the first is related to the heat exchange between the vapor that condenses and the coolant Μ =∫ βπ1π ππ,ππ’π‘ π0 πππ’π‘ ππΜ ππΜ −∫ π π πππ (7.3) This formula is always valid (also in case of desuperheating), but if the fluid discharged by the turbine is saturated vapor or in two-phase conditions (the most common condition in a vapor cycle), then it is possible to use a simplified formula, that will be used in the following calculations: ππ,ππ’π‘ Μ =∫ βπ1π π0 ππΜ πΜππππ − π πππππ 68 of 124 (7.4) Second-law analysis of power cycles - Energy Conversion A – V7.0 • the second is linked to mixing the coolant with the heat sink Μ =+ βπ1π π0 πΜππππ ππΜ +∫ π0 ππ,ππ’π‘ π (7.5) If there is no desuperheating, the total entropy increase is given by: βπ1Μ = + πΜππππ πΜππππ πππππ − π0 − = πΜππππ π0 πππππ πππππ π0 (7.6) And therefore: βπ1 = πΜππππ πππππ − π0 πΜπΏπ»π πππππ (7.7) It can be noted that the temperature ππ,ππ’π‘ has no influence and so the share between βππππ and βππ is not important, but rather their sum. The choice of ππ,ππ’π‘ has, however, obviously an effect on the flow rate of cold fluid, the consumption of the pumps and the surface and cost of the exchanger. The formula shows that the efficiency loss: a) is proportional to the πΜππππ πΜπΏπ»π = 1 − π relation, so it is more important the less the cycle efficiency is; grows with the increase in the terms βππππ and βππ ; both terms are set based on technicaleconomic considerations (cost of the exchanger, investments and consumption for the circulation of the coolant, equivalent plant operation hours, etc.). Ecological and permit limitations may also arise to limit βππ . In the case of very low π0 values, also considerations related to the higher cost of the LP vapor turbine and the difficulties in maintaining a too high vacuum level affect the choice of πππππ . In the most common cases, the πππππ − π0 difference takes on the following values: • 12/15°C for cooling with flowing water (river, sea inlet) • 15/20°C for cooling with air (reference: dry bulb temperature of the ambient air) • 10/15°C for cooling with evaporation towers (reference: wet bulb temperature of the ambient air) Once set the condensation temperature, the division between βππππ and βππ is decided based on technical-economic assessments. As βππ rises, there is a reduction of pump and fan consumption, but there is an increase in the heat transfer surface and condenser costs. The condenser of a Rankine cycle cooled with ambient air or water works at sub-atmospheric pressures (ππ ππ‘(35°C) = 0.056 bar) and is, hence, subject to leakages. The ambient air that penetrates the condenser is partially mixed in the liquid phase and is then drawn by the pump. It is important to remove these gases, typically called “incondensable gases” or better said “supercritical species”, from the working fluid before they reach the boiler, where they would create corrosion phenomena. Their accumulation in the condenser also leads to increased pressure in the component and reduced performance. The deaerator removes them. A vacuum pump is also connected to the condenser to remove the air it contains during start-up following a shut-down. 69 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 In addition to the irreversibilities of heat transfer, it is then necessary to consider those due to the pressure drops and the condensation auxiliary systems: − The pressure drops are certainly on the cold side in the case of both water and air condensers. These losses involve the use of pumps and fans that lead to power absorption, which in the case of the air condensers, is indicatively 1% of the discharged heat, and therefore a nonnegligible amount of the gross power produced. − On the condensate side, on the other hand, the pressure drops are significant only in the case of air condensers in which the working fluid flows in the tubes, where concentrated (distribution and collection) and distributed pressure drops occur. In this case, the condensation is not isobaric, and therefore not at constant temperature. A pressure drop equivalent to 0.31°C of difference between the saturation temperatures is considered in the large plants. − In the case of large water condensate plants, the pressure drops are instead negligible, and the process is actually isothermal. 7.3 βπΌπ FLUID DYNAMIC IRREVERSIBILITY IN THE PUMPS Pumps are installed in two points of the plant: • condensate extraction pumps: the first group of pumps is located under the condenser and its job is to extract the saturated liquid and to raise its pressure in order to overcome the pressure drops in the low pressure preheaters up to the deaerator. The pump works at subatmospheric pressure and the seals have to be designed to cut down the leakages of incondensable gases. These pumps have to work under a proper head because the fluid is saturated and cavitation phenomena are more likely. • feed pumps: these are located under the deaerator and carry the working fluid up to the maximum plant pressure, which in the case of ultra-supercritical plants has to be higher than 300 bar. This process is usually executed with two pumps in series: the first, called booster, which raises the fluid pressure (5-7 bar) in order to guarantee an adequate head for the actual feed pump that has to achieve high compression ratios. This pump can be driven by an electric motor under an inverter to perform the adjustment, or be one of a group of turbine pumps in which the pump is driven by a turbine that expands medium pressure vapor that is then discharged to the condenser. This solution is preferable for the large plants in which the efficiency of the turbine pump group may exceed that of the electric motor. The pumps are generally multi-stage centrifugal pumps, and are never single-stage; for safety and reliability reasons 2x100 or 3x50 configurations are recommended, as shown in Figure 7.6. 3 x 50 2 x 100 Fig. 7.6 - Arrangement of the pumps in a power plant to guarantee high reliability The efficiency loss linked to the fluid dynamic irreversibilities that take place inside the pumps is given by the truly lost work, which is not that of friction but π0 βπ: 70 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 βπ2 = π0 πΜπΏπ»π ∑ βππΜ = π π0 πΜπΏπ»π ∑[ π πΜπ π£π βππ 1 − πππ,π ( )] ππ₯,π πππ,π (7.8) As for the efficiency loss regarding the fluid dynamic losses of the pumps, the following notations can be put forward: a) It is preferable to raise the mean temperature πΜ at which pumping takes place. If one wants to force the plant layout to the utmost to enhance this effect, it would be necessary to adopt a preheating line set up with a series of direct (or surface) heat exchangers, each of which preceded by a pump, moving the feed pump to the end of the preheating process as shown in Fig. 7.7. b) The relative importance of the losses in the pumps rises linearly with the maximum cycle pressure (ππππ₯ − ππππ ~ ππππ₯ 8) , while the term πΜπΏπ»π varies little with the maximum cycle pressure. c) As always, when the fluid dynamic irreversibilities in the machines are considered, the losses are related to the fluid dynamic performances of the components. In view of the small specific volumes, the efficiency losses associated with the pumping phase are small compared to those of other components, so economic efforts and plant complications to increase the polytropic efficiency or raise the temperature at which the pumps work are not usually rewarded. An exception to the rule are the ultra-supercritical plants, where in view of the high maximum pressures of the cycle, the term of useful work loss (basically proportional to ππππ₯ ) may be substantial. It is generally always expedient to raise the maximum pressure of a vapor cycle since the increase in specific work spent in the pumping is orders of magnitude lower than what can be obtained in expansion (π£ π βͺ π£ π£ ). Nevertheless, there is an optimum cycle pressure beyond which the increase in power absorbed by the pump is higher than that obtainable by the turbine since the specific volume in the gas phase is linked both to temperature and pressure. This pressure is found at many hundreds of bar, for which technical-economic restrictions related to the materials used in the boiler intervene. This is shown in Figures 7.7 and 7.8. bleedings from steam turbine 8 deaerator 6 deaerator 4 7 10 9 4 8 3 2 5 high-pressure preheaters low-pressure preheater 1 3 T 2 low-pressure prehaters 5 feedwater pump 11 7 6 1 direct heat exchangers s Fig. 7.7 - Configuration of a preheating line that minimizes the irreversibilities of the pumping process 8 The minimum pressure of the vapor cycles is usually a few hundredths of a bar, four orders of magnitude lower than the maximum pressures. 71 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 Δπ2 ππ¦ ππππ = ππ ππππ₯ Fig. 7.8 - Change in the term βπ2 in relation to the maximum pressure of the cycle 7.4 βπΌπ IRREVERSIBILITY OF HEAT TRANSFER IN THE PREHEATING LINE The preheating line of a modern steam power plant has a high number of preheaters, all of the surface type except for the deaerator in which the liquid mixes with the vapor. By convention, the heat exchangers arranged between the condenser and the deaerator are low pressure preheaters, and those installed between the deaerator and the heat generator are high pressure preheaters. The aim is to raise the mean heat introduction temperature by preheating the liquid going into the boiler up to a temperature ππ₯ . 7.4.1 Indirect preheaters Every preheater is built as a Shell&Tube exchanger, in which the liquid to be heated flows in the tubes and the vapor condensates outside. Based on the bled vapor conditions, one may or may not get a desuperheating section (typical of the high pressure preheaters), while the fluid is usually two-phase for the low pressure preheaters. The condensing vapor flow is mixed with the flow coming from the next previously throttled preheater, which is generally found in two-phase conditions with a small vapor title. The condensate on the hot side is also sub-cooled in order to use less flow rate of fluid bled from the turbine, reduce heat transfer βππππ and finally increase the efficiency of the plant. The only substantial difference between low and high pressure preheaters is the pressure at which they work. The two exchangers are recognizable from the shape of the headers for distribution and collection of water in the tubes. The headers are flat for the low pressure exchangers and are domeshaped for the high pressure exchangers in order to better withstand the high pressures reducing the thicknesses of the material used. 72 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 Fig. 7.9 - TQ diagram of the cycle preheating phase. Fig. 7.10 - Indirect preheater with dual passage on the water side Therefore, these heat exchangers operate with three flows, as seen in Figure 7.10: two that transfer heat (the vapor and the condensate recycled by the next preheater) and a fluid receiving it (the pressurized water current). The two hot currents can also be thought as mixed at the unit inlet. The loss of useful work associated with the production of irreversibility is therefore due to three phenomena: (i) the heat transfer, (ii) the pressure drops and (iii) mixing of bled vapor and recycled condensate. The last two losses will be discussed separately in the following chapters, while that of heat transfer can be written as: βπ3,π = βWΜπ π0 π0 π0 ππΜπππ ππΜπππ = βππΜ = (βπβΜ + βππΜ ) = (∫ −∫ ) π π πΜπΏπ»π πΜπΏπ»π πΜπΏπ»π πΜπΏπ»π β π (7.9) where by introducing appropriate average quantities the following is achieved βπ3,π = π0 πΜπππ ππ,β − ππ,π πΜπΏπ»π ππ,β ππ,π 73 of 124 (7.10) Second-law analysis of power cycles - Energy Conversion A – V7.0 From which one notes that the loss of efficiency grows: • With the increase in weight of regenerated heat compared to introduced heat, which rises in turn as temperature ππ₯ of final preheating increases. Actually, the efficiency of the cycle benefits from an increase in ππ₯ because the loss of efficiency related to the introduction of heat in the cycle decreases, which certainly takes place under greater differences in temperature. • With the increase in the mean temperature change between vapor and condensate. This change in turn depends on two parameters: − βππππ 9, the temperature difference at the pinch-point that in turn is the result of a technical-economic optimization in sizing the heat exchanger: (i) the smaller βππππ , the larger the transfer surfaces and therefore the dimensions, weights and costs of the heat exchangers; (ii) the larger βππππ , the larger the productions of entropy, which turns into a smaller plant efficiency. Since these are heat exchangers that operate with extremely high heat transfer coefficients on both sides of the transfer surfaces (water in forced convection and condensing vapor), the economic optimum leads to very small βππππ , in the order of 2-3°C. − βπππ which in turn depends on the sub-cooling of the condensate: it is expedient to reduce this difference in temperature to get a lower βππππ . Sub-cooling is achieved by positioning a certain number of tubes under the level of the liquid and by checking this level with the valve on the regenerative bleeding side; if the level drops, the valve is choked and less vapor flow passes through. • Number of heat exchangers ππ : it is the result of a technical-economic optimization. The greater the ππ , the more complex and expensive the plant becomes, but the temperature differences that can be adopted are fewer. Three low pressure and four high pressure preheaters are generally used. Δπ3 Δππππ ππ ππ ππ₯ Fig. 7.11 - Trend of βπ3 as the final preheating temperature and number of preheaters change 7.4.2 Deaerator Since the condenser and condensate extraction pump, and the final stages of the turbine, work at subatmospheric pressures, there are some leakages of ambient air that partially mix with the liquid. These incondensable gases must be eliminated by the deaerator in order to prevent them from accumulating in the condenser (with increased condensation pressure and reduced performance), and to prevent corrosion phenomena at high temperatures. The deaerator consists of a column containing drilled plates at the top of which the pressurized liquid is distributed. The liquid falls downward, breaking up into drops, and meets the superheated vapor injected at the base of the tower. A heat and mass transfer occurs between vapor and liquid, in which the incondensable gases are stripped and 9 βππππ and βπππ are used without distinction in the lecture notes. 74 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 removed from the liquid current due to the change in solubility of the incondensable gases (Henry's law). The vapor and incondensable gases are drawn from the top of the tower. In some models, the stripped gases are conveyed in a series of paths created with sheets so that the incondensable gases can be stratified at the base of the tower as they are heavier than the vapor. The discharge into the atmosphere is usually seen as a plume of condensed vapor. Demineralized water is added to restore this loss and to restore the nominal salts content. There is a plenum chamber under the tower to recreate a condition of liquid and vapor equilibrium. The deaerator is always placed high up to allow the feed pump to have the correct head and not run into cavitation, as can be seen in Figure 7.12. incondensable gases + steam deaerated water bled steam pressurized liquid Fig. 7.12 - Diagram of the deaerator and its operation The deaerator is modelled as a direct heat exchanger that takes the pressurized liquid to the saturation temperature. The βππππ is therefore null and the irreversibility loss is calculated like in the previous exchangers. 7.5 βπΌπ IRREVERSIBILITY OF HEAT TRANSFER IN INTRODUCING HEAT INTO THE CYCLE Combustion takes place in a boiler in which the power released by the reaction is transferred to the working fluid in different heat transfer sections. The zone where there is the flame is surrounded by membrane water tubes that mainly receive heat by radiative and partly convective heat transfer, where the liquid evaporates to form saturated vapor. The maximum temperatures in this zone are very high, but lower than the adiabatic flame temperature, and gradually decrease as they move toward the top of the combustion chamber. It is impossible to define a real βππππ in this section since the heat transfer coefficient is not only convective, but radiative as well. Moreover, the definition of the βπ is not univocal because in this section there are heat transfer irreversibilities and combustion irreversibilities at the same time. The fluid enters with a certain degree of sub-cooling in order to be certain that there is no beginning of vaporization in the economizer, that would lead to an appreciable increase in pressure drops. The evaporation takes place in boilers having cylindrical bodies, positioned at the top of the tube bundle where there is physical separation of vapor and liquid in equilibrium conditions. The saturated liquid is taken from an insulated downcomer tube outside the boiler to the base of the boiler in the lower cylindrical body. The fluid is then distributed in numerous membrane water tubes surrounding the combustion chamber. Circulation is natural and guaranteed by differences in density between the two columns of fluid. The liquid rises up toward the upper cylindrical body while partially evaporating. The vapor title is always kept low with an appropriate circulation flow rate to prevent burn-outs and reduced transfer coefficients. Vapor is 75 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 extracted from the upper cylindrical body and there is usually a demister installed to block any drops of liquid. Following the evaporation section are the superheating and reheating banks that cool the flue gas produced in the combustion chamber. In this case it is instead legitimate to consider a well-defined βπ of heat transfer once the conditions upstream and downstream of each component are known. The maximum temperatures that the vapor can reach are about 600-620°C due to the technological limitations on the materials. The economizer is found at the lower temperatures, which takes the compressed liquid from final preheating ππ₯ up to temperatures close to those of evaporation. Lastly, there is the regenerative preheating of the reactants, or of the oxidant air, through a Ljungström exchanger. Preheating the air to higher temperatures is not always advantageous since on the one hand a higher adiabatic flame temperature, hence lower irreversibility of combustion, would be obtained. On the other hand, a greater irreversibility of heat transfer would be obtained since heat at temperatures higher than the maximum temperatures of the materials cannot be introduced. Moreover, the differences in heat capacity do not allow high preheating and curbed stack temperature to be obtained at the same time. A complete schematic of a coal dust boiler is seen in Figure 7.13. ECO out (to EVA) SH2 out (to turbine) SH1 out SH2 RH SH1 EVA ECO SH1 inlet Ljungstrom Boiler feedwater inlet Fig. 7.13 - Diagram of a coal dust boiler for a large steam power plant 76 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 Fig. 7.14 - Temperature-thermal power diagram of a highly advanced ultra-supercritical (USC) steam generator: the vapor temperatures (700/720°C) are much higher than those corresponding to today's state of the art. Note that the introduction of heat into the vapor cycle corresponds to a temperature Tx equal to about 340°C, to which a combustion product outlet temperature of around 380°C corresponds. Also note the regenerative preheating of the combustion air at very high temperatures (about 360°C), which allows the low temperature combustion products to be discharged, compatible with a high steam generator efficiency The loss of useful work of a steam generator can be conceptually divided into: 1. Non-reversible preheating of the reactants 2. Combustion that takes the flue gas to the adiabatic flame temperature 3. Cooling of the flue gas and introduction of heat in the working fluid in the evaporator tubes 4. Further cooling of the flue gas and introduction of heat in the SH, RH and ECO banks 5. Mixing the flue gas with the environment This choice does not represent the physical meaning of the problem because, as already stated, the second and third losses occur at the same time. The combustion is not adiabatic since part of the heat is transferred to the evaporative tube bundle and therefore the combustion products do not reach the adiabatic flame temperature. Due to this difficulty, it is decided to consider the irreversibility of the steam generator in only two contributions, referring to the thermal power entering the cycle in agreement with the definition of efficiency provided in equation (6.2): 1. 2. Derating of the exergy of the fuel to a heat introduction on an isotherm at the maximum cycle temperature; Heat transfer from the maximum cycle temperature to the working fluid. 7.5.1 βπΌππ Derating the thermal energy from π»∞ to π»πππ Now the physical meaning of the choice of an intermediate sink at a constant temperature equal to the maximum temperature of the cycle will be clarified. Today's state of the art of heat generators allows maximum vapor temperatures of 600-620°C to be reached, as shown in Fig. 7.15. Also bear in mind 77 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 that the maximum temperature to which metal is taken is at least 30°C higher than the temperature of the working fluid. external convection conduction internal convection T Fig. 7.15 - Temperature trend in the SH or RH banks The many attempts made to overcome these limitations over the decades have been unsuccessful because the sizing of the steam generator requires adopting: • Enormous heat transfer surfaces: because of the not high global heat transfer coefficients, especially in the convective part of the generator where on the one hand there are burnt gases that have low exchange coefficients, like all low-density gases; • These enormous surfaces not only transfer heat, but must also withstand high internal pressures. This is the reason why the need to adopt large thicknesses arises, meaning large volumes and metal material weights able to retain good mechanical properties at a high temperature; • Therefore, it is not acceptable to use metal alloys with high specific costs (€/kg), which certainly exist and are adopted in other plants (e.g. gas turbines), which would involve prohibitive costs. The example shown in Table 7.1 and the following figures show these concepts. Fig. 7.16 - Breakdown in percentage of the materials used in the steam power plants according to technological advancement in terms of pressures and maximum temperatures of the working fluid 78 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 Fig. 7.17 - Maximum temperatures (SH and RH) adopted in the coal plants, based on the metal materials used Table 7.1 - Comparison between unit costs of the superheater tubes of a steam generator operating at the current state of the art (600°C) and those (prohibitive) required to raise the temperature by 100°C. In the solution with the highest temperature, the ratios (thickness to diameter) increase both because the pressure rises and because the allowable stress of the material decreases. Dimensions, mm (internal diameter – thickness) Material Material cost per kg Material cost per meter Cost per meter ratio, with the same passage area ππππ₯ = 600°πΆ 221 – 32 P91 € 5.5 € 1100 1 ππππ₯ = 700°πΆ 175 - 60 Alloy A617 A130 € 48.0 € 16,600 24 This loss considers the derating of the fuel exergy modelled as a heat available at infinite temperature in heat at the maximum cycle temperature. It should be emphasized that not even this loss has an actual physical meaning since the flue gases are discharged at a lower temperature than ππππ₯ and, therefore, the flue gas cooling curve inevitably crosses the isothermal source. However, the approach is valid considering that it is possible to raise the heat transfer temperature at will through recuperative preheating. The meaning of this term is, on the other hand, that of representing the loss given by technological limitations that do not allow the live steam conditions at the turbine inlet to rise beyond a certain temperature. βπ4π = π0 πΜπΏπ»π ( πΜπΏπ»π π0 )= ππππ₯ ππππ₯ That is, the complement at one of the Carnot efficiency between π0 and ππππ₯ . 79 of 124 (7.11) Second-law analysis of power cycles - Energy Conversion A – V7.0 7.5.2 βπΌππ Introducing heat into the cycle starting from π»πππ of the working fluid While examining the quality of a vapor cycle, there is no particular interest in penalizing an introduction of heat starting from temperatures higher than the assumed Tmax (e.g. 620°C), which would in any case not be acceptable for the reasons explained above. Another point to clarify is the assumption that the source has a constant temperature. Actually, the combustion products in a heat generator are a source at variable temperature, with heat transfer temperatures even much lower than the assumed 620°C. The introduction of heat is not penalized by the need to cool the combustion products, not even for cycles in which the temperature at which the heat starts to transfer from the source ventures to extremely high values. The efficiency loss can be expressed as: βπ4π = π0 πΜπΏπ»π (∫ ππΜ πΜπΏπ»π ππππ₯ − ππ − ) = π0 π ππππ₯ ππππ₯ ππ (7.12) where ππ is the mean temperature of introduction of heat into the cycle (falling between ππππ₯ and ππ₯ ) . The efficiency loss naturally decreases as ππ rises. This can be obtained by: • Increasing ππ₯ and, therefore, with greater liquid preheatings; • Increasing the evaporation pressure (and hence the temperature), by switching to supercritical pressure; the mean heat introduction temperature continues to rise, even if the variation of ππ is relatively low with the pressure increase (Fig. 7.18 ) since a significant part of the heat in any case tends to enter at temperatures close to the critical temperature of water (Fig. 7.19). In the gradual transition between the liquid and vapor states, the supercritical isobaric line presents very high specific heat values in proximity of the critical temperature (which result in a low isobaric slope in the TS plane). For very high pressures, the benefits in terms of βπ4π can be canceled out by the increase in the pump losses, which linearly grow with the maximum pressure. 600 450 400 Mean temperature, °C Temperature, °C 550 500 500 400 300 450 400 220 200 350 350 300 250 100 subcritical 300 supercritical 200 0 0.2 0.4 0.6 0.8 1 Q, % 100 200 300 400 500 Pressure, bar Fig. 7.18 - Variation of Tm (°C) as the pressure (bar) changes, for Tx =350 °C and Tmax = 600 °C. A pressure increase from 300 to 500 bar leads to an increase in the mean heat introduction temperature of just 26°C. 80 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 80 250 bar 70 cp, KJ/kgK 60 275 50 40 300 350 30 20 400 500 10 0 300 350 400 450 500 550 600 T, C Fig. 7.19 - Variation of the specific heat of water with the temperature for different supercritical pressures. The area under the dome ∫ ππ ππ represents the heat introduced for the gradual transition from the liquid state to the vapor state • Introducing one or more reheatings. The pressure at which reheating is carried out must be optimized considering thermodynamic and technical-economic aspects. From the thermodynamic viewpoint, when the reheating pressure decreases, (i) the incoming thermal power increases (beneficial effect since πΜπΏπ»π is the denominator of all the βππ ), but (ii) the efficiency loss linked to the heat transfer in the reheater increases. The optimum RH pressure is that for which the efficiency loss introduced by the reheater is equal to the sum of the efficiency loss reductions of the other processes obtained due to an increase in πΜπΏπ»π with the reduction of ππ π» as can be seen in Fig. 7.20. Fig. 7.20 - Increase in the thermal power introduced into the vapor cycle as the RH pressure decreases (left), and definition of the optimum RH pressure from the point of view of the second law analysis (right) From the technical and economic viewpoint, the RH is necessary in order to limit the liquid title in the expansion, which would be too high for subcritical cycles at high pressures or, at most, USC cycles with just SH. From the economic point of view, introducing a superheating is very costly since: • It is necessary to interrupt the expansion; • It is necessary to transport vapor at high temperatures to the steam generator (and back) with relatively high thermal losses and large passage sections; • High cost of the heat transfer section materials; • Reduced plant reliability. 81 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 The most recent plants tend to limit the number of reheatings to one, due to the extremely high costs that adopting a second reheating would entail. Even if the pressure is pushed to the maximum and one/two reheatings are adopted, the term βπ4π is nevertheless high. This is the inevitable consequence of adopting a fluid with a critical temperature very remote from the maximum heat introduction temperatures that the current "state of the art" allows. 7.5.3 Binary cycles One alternative proposed to overcome this situation, is replacing water vapor with another working fluid with higher critical temperatures. Liquid metals (Hg, Na, K) meet this requirement. Furthermore, they are thermally stable at high temperature since they are monatomic. In theory, it is therefore possible to hypothesize a saturated vapor cycle with evaporation temperature of about 600°C and condensation temperature of about 30-40°C, as shown in Fig. 7.21a. Actually, when a fluid has a high critical temperature, it inevitably has very low saturation pressures, unacceptable for a number of reasons in a large power cycle10. In other words, every working fluid is suitable for carrying out a saturated vapor cycle (the best thermodynamic cycle) in a relatively limited temperature range (200300°C), much lower than the range (about 600°C) that the state of the art of the materials and the ambient temperatures allow. This is why "binary" cycles are often proposed. They are formed by two overlapping cycles in which the enthalpy of condensation of the top cycle is introduced into the evaporator of the bottom cycle. Although another inevitable irreversibility is introduced, represented by the heat transfer between the two cycles, the total efficiency can be higher than a supercritical vapor cycle. A series of technological problems linked to the adoption of liquid metals (Hg incompatible with the environment, Na explosive with water) has, in fact, thwarted the commercial success of these solutions. An example is seen in Fig. 7.21. Another proposal studied, though never carried out commercially, is a water (topping) and ammonia (bottoming) binary cycle to exploit very low condensation temperatures, smaller seal problems and a smaller size of the turbomachinery. Fig. 7.21 - a) Saturated vapor cycle with mercury: the condensation pressure at ambient temperature is unacceptable (Torricellian vacuum!) b) Hg/H2O binary cycle, where the fluids work at technically acceptable pressures. Assuming regenerative limit cycles, a total efficiency of the two cycles equal to the Carnot efficiency between Tmax and T0 would be obtained. 10 Actually, pressures higher than 400 bar or lower than 0.01 bar are not adopted in power cycles. The first limitation is related to the mechanical stresses in the primary heat exchanger, the second to the volumetric flow rates of the turbine and the seal problems (air re-entering the cycle). 82 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 7.6 βπΌπ FLUID DYNAMIC IRREVERSIBILITY IN THE TURBINE The expansion of a large steam power plant is divided into a high number of turbine stages and cylinders.First of all, there is usually a very high-pressure stage followed by a high pressure cylinder. This is generally followed by reheating and then one or more medium-pressure cylinders in parallel. The flow at low pressures is divided into multiple cylinders (two or four, depending on the size of the plant) through a crossover to reduce the size of the final turbine stages. Holes are drilled in the turbine cylinders to draw regenerative bleeds; the only bleed controlled under pressure is the one of the deaerator that has to always be at a pressure higher than the ambient pressure in order to guarantee venting of the incondensable gases. The following results from the relation previously obtained for the irreversibility produced due to expansion in the turbine: βπ5 = π0 πΜπΏπ»π [ πΜπ π’ πππ Μ Μ Μ Μ Μ Μ Μ Μ Μ πΜ (1 − ππ¦ )ππ ( )] ππ πππ’π‘ (7.13) This loss for a steam power plant is not easy to calculate because: • the flow rate is not constant along the expansion due to the presence of regenerative bleeds. The useful work loss must be calculated for every turbine section between two bleeds and the inlet and outlet conditions; • The polytropic efficiency is not constant along the machine and depends on the design and construction constraints of each stage. The high pressure and low pressure cylinders generally have an efficiency lower than the medium pressure casings because the first are penalized by high profile losses and the others by high losses due to 3D effects; • The presence of drops of liquid further penalizes the efficiency for turbines that expand a twophase fluid. 7.7 βπΌπ PRESSURE DROPS IN THE LIQUID PHASE The total efficiency loss is equal to: βπ6 = π0 πΜπΏπ»π ∑( π π£π πΜπ βππ ) πΜ π (7.14) π π’ βππ ππΜ ) ππ ππ (7.15) 7.8 βπΌπ PRESSURE DROPS IN THE VAPOR PHASE The total efficiency loss is equal to: βπ7 = π0 πΜπΏπ»π ∑ (πΜπ π The equation demonstrates that the pressure drops in the vapor stage are mostly paid at low pressure βπ (the relative pressure drops π count, and are obviously proportional to the flow rate that sustains the ππ drop and to the compressibility coefficient πΜ ). Since the specific volume is high, the influence on the cycle efficiency is greater than in the case of liquid. Careful attention must therefore be paid to the pressure drops, especially at low pressure. As a result, if a valve has to be placed inside a Rankine cycle, it is preferable to do so on the liquid side and at the highest possible temperature: the best point is therefore at the outlet of the economizer 83 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 (maximum temperature of the liquid), while the worst is downstream of the turbine where the pressure βπ is minimum and the π is maximum ππ 7.9 βπΌπ THERMAL LOSSES The thermal losses, related to the inefficient insulation of the hot components, cause an efficiency loss given by: βπ8 = π0 πΜπΏπ»π ∑ [πΜπππ π,π ( π 1 1 − )] π0 ππ (7.16) The loss is higher the greater the dissipated thermal power is and the higher the mean temperature of the fluid releasing heat is. There is a significant thermal loss in the pipeline connecting the steam generator to the turbine. This section is particularly critical because the distances are far and the temperatures are the highest in the plant. There is generally a temperature reduction of about 2-3°C in these sections. 7.10 βπΌπ MECHANICAL/ELECTRICAL LOSSES The mechanical and electrical losses represent a dissipation of totally lost "valuable" energy. In general, it is impossible to recover anything. They degrade into heat at ambient temperature ∑π πΜ ππ,π (7.17) βπ9 = πΜπΏπ»π 7.11 βπΌππ AUXILIARY LOSSES If it is assumed that the auxiliary electrical consumption is drawn upstream of the measurement of useful power πΜπ’ , it represents a cycle efficiency loss equal to: ∑π πΜππ’π₯,π πΜπΏπ»π A schematic of the last losses presented is shown in Fig. 7.22. βπ10 = (7.18) Fig. 7.22 - Flows of power assumed for the auxiliary consumption 7.12 COMPARISON OF LOSSES The result of the analysis described above regarding an USC coal dust power plant taken as reference from the European Benchmark Task Force (ππππ₯ = 620 °C) is shown in Fig. 7.23. The benchmark for 84 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 the heat sink is 15°C. The power plant has a net electrical efficiency of 45.5%, which corresponds to a cycle efficiency of 47.36%11 Fig. 7.23 - Example of a second law analysis applied to a supercritical vapor cycle (EBTF reference). Note the importance of the losses linked to the introduction of heat into the cycle. The most critical processes are also those on which it is preferable to intervene: • By introducing heat: when raising ππππ₯ the mean heat introduction temperature is raised, and hence also the cycle efficiency; • By improving the efficiency of the turbine; • By increasing the preheating efficiency; • By reducing the condensation temperature; • Auxiliary systems (fans, circulation pumps, safety systems and self-consumptions). A final consideration is that very often when comparing steam power plants, the loss due to combustion and the transfer of heat at temperature ππππ₯ is not considered, and the analysis of the losses starts from a reversible work given by πΜπππ£ = πΜπΏπ»π (1 − π0 ) ππππ₯ (7.19) The analysis of the efficiency losses therefore becomes equal to: π πΜπ’ πΜπππ£ βπΜπ = −∑ πΜπΏπ»π πΜπΏπ»π πΜπΏπ»π (7.20) ππΌ = ππππ£ − ∑ βππ (7.21) 1 π 1 11 The analysis does not consider the losses of conversion from coal to heat entering the cycle, so the resulting efficiency refers to the heat entering the cycle. 85 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 7.13 OTHER NUMERICAL EXAMPLES The student has to supplement all the qualitative considerations put forth in this chapter with the numerical results obtained during Exercise 4 dedicated to the second law analysis for USC steam power plants. 86 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 8 EFFECT OF THE THERMODYNAMIC PROPERTIES OF THE WORKING FLUID ON RANKINE CYCLE PERFORMANCE In this chapter the impact of the properties of fluids is discussed. The thermodynamic properties of the working fluid have a significant impact on the maximum performance that can be obtained from a Rankine cycle. In particular, the critical properties, the molecular complexity and the molar mass of the fluid can significantly modify the Ts diagram, the TQ diagram and influence the design of the turbine, making a vapor cycle more or less suitable to exploit certain types of heat sources. First of all, the effect of molecular complexity will be analyzed because it is the most influent on the thermodynamics of the cycle, while the impact of molecular mass and of critical properties will be discussed at the end of the chapter. 8.1 SOURCES AT CONSTANT TEMPERATURE - NON-REGENERATIVE CYCLE In order to exploit a source with infinite heat capacity, the limit thermodynamic cycle of reference is the saturated Rankine cycle, operating between the source temperature ππππ₯ and the cold sink at temperature π0 with unitary machinery efficiency and without pressure drops. In this case it is considered to not adopt regeneration. Two fluids with different molecular complexity are compared: the first is water representing a simple molecule fluid, the other is MD4M, a siloxane with high complexity (C14H42O5Si6). The two fluids differ in complexity and molecular weight (18 kg/kmol versus 459 kg/kmol) while they have a very similar critical temperature (374°C versus 380°C). The choice of using two fluids with the same critical temperature is justified by the need to simplify the analysis by using the principle of corresponding states. Evaporation and condensation temperatures being equal, it is therefore possible to consider the residual enthalpy and that of the specific heat on a molar-basis equal for the two fluids. In addition, the evaporation temperature is 180°C, which corresponds to a reduced temperature of 0.7, at which it is possible to consider the volumetric and thermodynamic behavior of the saturated vapor equal to that of an ideal gas without introducing excessive approximations. The comparison will be performed on ideal cycles and therefore with adiabatic and isentropic machines, negligible pressure losses and heat exchangers with infinite surfaces. Fig. 8.1 shows the Ts diagrams for the two fluids for non-regenerative saturated cycles. Fig. 8.1 - Ts diagrams for two saturated Rankine cycles operating with water (left) and with MD4M (right) 87 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 8.1.1 Effect of the molecular complexity As it is well known, the molecular complexity affects the shape of the saturation dome and, in particular, the slope of the saturated vapor line. For the siloxane MD4M, the expansion takes place in the superheated vapor field and with small temperature changes. For both fluids the introduction of heat into the economizer, πΜπΈπΆπ = πΜ πΜ π (ππππ₯ − π0 ), is an irreversible process, while for the complex fluid also the transfer of heat to the environment is irreversible in the desuperheating section, πΜπ·πΈπ = πΜ πΜπ 0 (ππ·πΈπ − π0 ). In the case of the simple fluid, this drop is not present since the expansion takes place in the two-phase field. By applying the second law analysis to the first law efficiency, what is obtained in the most general case is: Μ Μ πΜπ’ πΜπππ£ π0 βππΈπΆπ π0 βππ·πΈπ = − − πΜππ πΜππ πΜππ πΜππ ππΌ = ππππ£ − βππΈπΆπ − βππ·πΈπ π0 ππΌ = (1 − ) − βππΈπΆπ − βππ·πΈπ ππππ₯ (8.1) (8.2) (8.3) where the reversible efficiency is clearly equal to the efficiency of the Carnot cycle operating between the evaporation and condensation temperatures. Considering the liquid having a constant specific heat, the loss of efficiency in the economizer is: Μ π0 βππΈπΆπ π0 Μ + βππ π Μ ) = (βππππ πΈπΆπ Μ πππ πΜππ π0 ππππ₯ πΜ πΜ π (ππππ₯ − π0 ) = (πΜ πΜ π ππ ( ) )− π0 ππππ₯ πΜππ βππΈπΆπ = βππΈπΆπ (8.4) (8.5) That when collecting πΜ πΜ π (ππππ₯ − π0 ) is: βππΈπΆπ πΜ π (ππππ₯ − π0 ) π0 π0 = π ( − ) πΜ (ππππ₯ − π0 ) + ββΜππ£π ππππ,πΈπΆπ ππππ₯ (8.6) if mole quantities are considered, then the ββΜππ£π and the correction of βπΜπ between liquid and saturated vapor are independent from the properties of the fluid (complexity and molecular mass) given the same reduced temperature. Therefore, the sole variable affecting the βππΈπΆπ is the πΜπ 0 of ideal gas since πΜ π = πΜπ 0 + βπΜπ . It follows that for simple fluids with small πΜπ 0 , the heat introduced in evaporation is greater than the heat introduced in economizing. These fluids are therefore able to reduce the weight of the loss of efficiency due to the introduction of heat and maximize efficiency. Complex fluids instead have a high value of πΜπΏ and an introduction of heat with a greater production of irreversibility. Moreover, the loss in the desuperheating has to added for fluids with high complexity. It is equal to: Μ π0 βππ·πΈπ π0 Μ + βππππ Μ = (βππππ )π·πΈπ πΜππ πΜππ πΜ πΜπ 0 (ππ·πΈπ − π0 ) π0 ππ·πΈπ = (−πΜ πΜπ 0 ππ ( ) )+ π0 π0 πΜππ βππ·πΈπ = βππ·πΈπ And therefore 88 of 124 (8.7) (8.8) Second-law analysis of power cycles - Energy Conversion A – V7.0 βππ·πΈπ πΜπ 0 (ππ·πΈπ − π0 ) π0 = π (1 − ) Μ ππππ,π·πΈπ πΜ (ππππ₯ − π0 ) + ββππ£π (8.9) The turbine discharge temperature ππ·πΈπ is the higher the greater the complexity of the fluid: if the complexity increases, πΎ → 1 and π → 0, and the isentropic expansion tends to also be isothermal. The loss of efficiency due to the transfer of heat into the environment is therefore dependent on the molecular complexity of the fluid, both due to the effect of greater heat transferred and to the increase of ππππ,π·πΈπ . Considering vapor an ideal gas throughout the expansion, it turns out that: 0 πΜπ (ππππ₯ (1 − π½ βππ·πΈπ = πΜ π (ππππ₯ π ππ 0 − ) − π0 ) − π0 ) + ββΜππ£π (1 − π0 ππππ,π·πΈπ ) (8.10) The table shows the efficiency losses and the efficiency that can be achieve with the two fluids. It is possible to see that in this case the water allows a cycle to be carried out with an efficiency greater than 10 percentage points compared to the complex fluid MD4M. Table 8.1 - Comparison of efficiency losses for two saturated Rankine cycles operating with water and with MD4M water MD4M ππππ£ 30.89% 27.84% 17.12% ππΌ βππΈπΆπ 3.06% 8.64% 5.14% βππ·πΈπ The TQ diagram for the two cycles is shown in Fig. 8.2 to highlight what is obtained analytically. Fig. 8.2 - TQ diagrams for two saturated Rankine cycles operating with water and with MD4M to exploit a source at constant temperature 89 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 8.2 SOURCES AT CONSTANT TEMPERATURE - REGENERATIVE CYCLE Now a regenerative cycle is considered: as already mentioned, a saturated cycle with simple molecule can reach the Carnot efficiency through a reversible regenerative preheating obtained with infinite regenerative bleeds from the turbine. In this case the introduction of heat is totally reversible, since it takes place at the temperature of the source. In practice, this is not possible, so a limited number of preheaters (surface or direct) is used. Each of these involves a loss of useful work due to the heat transfer under finite temperature differences, but it allows the loss due to heat introduction into the cycle to be limited, with beneficial effects on the efficiency. Their number is defined by a technicaleconomic optimization, given the greater plant engineering complexity and higher plant cost of a regenerative solution. For a complex fluid, on the other hand, it can be noted that the regeneration can take place not only through regenerative bleeds, but also (and above all) by using the hot vapor discharged by the turbine to preheat the compressed liquid. To do this, it is used a surface recuperator that cools the vapor up to temperatures close to that of condensation, and that reduces both the loss of introduction of heat into the cycle in the economizing section and the loss of transfer of heat to the environment in the desuperheating section. This heat exchanger is generally built as a finned coil placed at the turbine outlet: the liquid flows in the tubes while vapor flows outside, where the transfer surface is larger to compensate in part the low heat transfer coefficients characteristic of a gas in rarefaction conditions. While significantly increasing the efficiency of the cycle by reducing the two efficiency losses that characterize it, this component obviously results in losses of useful work due to: • pressure drops on both sides of the heat exchanger, which involve an increase in consumption of the pump and reduction of the work of the turbine; • heat transfer under finite temperature differences. Even imagining a heat exchanger with infinite surface, the different heat capacity of the vapor and liquid inevitably involves a change in temperature on the hot side greater than that obtainable on the cold side, and thus a cancellation of the temperature difference only on the cold end of the heat exchanger. A reversible recuperation is obtainable only with an infinitely complex fluid, since the specific heat of liquid and vapor in this case are equal, and the isentropic expansion is also isothermal. See the TQ diagram in Fig. 8.3: in the first case recuperation is not adopted and the cycle efficiency is about 17%; by adopting an ideal recuperation (recuperator with infinite exchange surface), the energy input is reduced with a considerable increase in plant efficiency. Table 8.2 - Comparison of efficiency losses for two saturated Rankine cycles operating with MD4M: non-recuperative and recuperative for the exploitation of a source at a constant temperature ππππ£ ππΌ βππΈπΆπ βππ·πΈπ βππππ non-REC REC 30.9% 17.1% 27.5% 8.6% 2.1% 5.1% 0.0% 1.3% 90 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 Fig. 8.3 - TQ diagrams for two saturated Rankine cycles operating with MD4M: non-recuperative (right) and recuperative (left) for the exploitation of a source at a constant temperature 8.3 SOURCES AT VARIABLE TEMPERATURE In the case of sources at variable temperature, the reversible cycle of reference is the trilateral cycle (if the source can be cooled down to ambient temperature) or the trapezoidal cycle (if the minimum temperature of the source is higher than the ambient temperature). These cycles allow heat to be introduced reversibly and the source to be totally exploited. In real cycles, the critical temperature of the fluid plays an important role in addition to the complexity. The reversible work obtainable with a reversible trapezoidal cycle is: πΜπππ£ = πΜππ,πππ₯ (1 − π0 ππππ(ππππ₯2|ππππ₯1) ) (8.11) where πΜππ,πππ₯ = πΆ(ππππ₯2 − ππππ₯1 ) is the thermal power available that completely cools the hot source. ππππ₯1 can be equal to the π0 when there are no limits in cooling the source. A non-reversible cycle will have a lower useful work because of losses linked to the introduction of heat into the evaporator and the economizer, the transfer of heat in the desuperheating and the total lack of exploitation of the source. The second law analysis applied to the second law balance gives: πΜπ’ πΜπππ£ π0 Μ Μ Μ Μ ) = − (βππΈππ΄ + βππΈπΆπ + βππ·πΈπ + βππππ₯ πΜπππ£ πΜπππ£ πΜπππ£ ππΌπΌ = 1 − βππΈππ΄ − βππΈπΆπ − βππ·πΈπ − βππππ₯ (8.12) (8.13) In this numerical example, the source has a maximum temperature of 230°C and may be cooled up to ambient temperature. 8.3.1 Simple molecules A simple molecule fluid is considered, and a saturated cycle is used. Most of the heat is introduced at constant temperature equal to that of evaporation. The efficiency is penalized due to the introduction of heat into the cycle under finite temperature differences, and to the total lack of exploitation of the source that will lead to a certain loss due to mixing with the environment. If a saturated limit cycle 91 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 with a reversible preheating is considered, it has the efficiency of the Carnot cycle. The optimum evaporation temperature is identified considering that the producible power is given by: πΜ = πΜππ (1 − π0 ) πππ£π (8.14) where the incoming power is only a fraction of the maximum power extractable from the source. Considering the infinite surface of the heat exchanger, it is found that πΜππ = πΆπ (ππππ₯,2 − πππ£π ). By differentiating from the variable πππ£π , it is possible to find the maximum producible power: π0 π [πΆπ (ππππ₯,2 − πππ£π ) (1 − )] ππΜ πππ£π = =0 ππππ£π ππππ£π ππΜ π0 π0 = −πΆπ (1 − =0 ) + πΆπ (ππππ₯,2 − πππ£π ) ππππ£π πππ£π πππ£π 2 ππΜ π0 = −πΆπ + πΆπ ππππ₯,2 =0 ππππ£π πππ£π 2 (8.15) (8.16) (8.17) and therefore πππ£π,πππ‘ = √π0 ππππ₯,2 (8.18) The value of the optimum evaporation temperature is therefore not dependent on a limit on the cooling temperature of the source until it is less than the πππ£π,πππ‘ πππ£π,πππ‘ = max (√π0 ππππ₯,2 , ππππ,π ) (8.19) Fig. 8.4 - Representation of the optimum temperature of a Carnot cycle for the exploitation of sources at variable temperature Fig. 8.5 shows the TQ diagram optimized for the case with water and non-regenerative saturated cycle. Note that the optimum evaporation temperature is equal to 129°C versus the 124°C expressed in the theoretical formula. The available heat is not fully exploited. The loss due to the irreversible mixing 92 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 of the source with the environment must also be added to the losses of useful work in the process of heat introduction (both into the economizer and the evaporator). A superheating, which increases the heat introduction mean temperature, improves the saturated cycle. While a substantial increase in efficiency is obtainable using one or more evaporation levels, a common practice in combined cycle plants. In this way, it is possible to minimize the βπ of heat transfer and introduce more heat into the cycle. T T ππΈπΆπ 250 ππΈππ΄ 200 150 100 ππ’ 50 ππππ₯ ππΆπππ· 0 -100% -50% 0% 50% 100% Q Fig. 8.5 - TQ diagram for a water recovery cycle with optimized evaporation temperature (left), and possibility of reducing the temperature difference in the heat introduction process with a cycle having two levels of evaporation and superheating (right) Adopting a simple fluid but with lower critical temperature, instead, the evaporation section compared to the heat introduced always weighs less, and for critical temperatures lower than ππππ₯,π , it is feasible to implement a supercritical cycle with the objective of reducing the irreversibilities due to heat introduction process. But in this case the large difference of specific heat between liquid and vapor results in a highly marked point of inflection at the critical point and an limited increase in mean temperature of the heat introduction, as seen previously for the water vapor cycle. 8.3.2 Complex molecules Substantial advantages can instead be obtained with a complex molecule that shows a greater liquid specific heat on a molar-bases. It is possible to raise the evaporation temperature for a subcritical saturated cycle while keeping the same pinch-point, and at the limit for an infinitely complex cycle there would be an introduction of heat only into the economizer with the possibility of minimizing the temperature differences in the primary heat exchanger. Fig. 8.6 shows the TQ diagram optimized for water and for MD4M. Note that a greater optimum evaporation temperature is obtained for the fluid with greater complexity, and the different share between heat introduced into the economizer and evaporator practically allows the βπ of heat transfer to be canceled out in the economizer. The losses due to introducing heat are still present in the evaporator, while those due to mixing with the environment are severely limited. The available heat is almost fully exploited and, despite the heavy loss due to the desuperheating process, the power produced increases by 23% compared to the case with water. 93 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 The loss due to the transfer of heat into the environment (desuperheating) can be limited by using a recuperator according to the minimum temperature of the source. Fig. 8.6 - Comparison between the TQ diagrams for the exploitation of a source at variable temperature with water (left) and with MD4M (right) Table 8.3 - Comparison of the efficiency losses for two saturated Rankine cycles operating with water and MD4Mfor the exploitation of a source at variable temperature ππππ£ ππΌπΌ βππΈπΆπ βππΈππ΄ βππππ₯ βππ·πΈπ water MD4M 100.0% 58.8% 72.5% 3.6% 0.4% 20.4% 10.2% 17.2% 0.1% 16.9% For the complex fluids, the switch to supercritical cycles is instead highly expedient since the slope of the isobaric lines is similar in the liquid and vapor fields with a supercritical transition characterized by a less pronounced point of inflection. This is shown in Fig. 8.6. 94 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 T T simple complex s s Fig. 8.7 - Comparison between the TQ diagrams for the exploitation of a source at variable temperature with supercritical cycles using a simple molecule (left) and a complex molecule (right) 8.4 GENERAL CONCLUSIONS From this discussion, we can gather that water is not the only option, despite it is the most used working fluid in Rankine cycles applications. Moreover, it is not the best solution when the available heat source is characterized by low thermal power and/or low maximum temperature. For large fossil-fired power plants, it is out of the question that water vapor cycles are the best option (apart from binary cycles with I group metals as working fluid). Any other thermal source (geothermal, heat rejection, biomass, solar plants) the optimal working fluid can be chosen among a large number of candidates, considering thermodynamics, but also economics and technical issues. The cycles that uses organic fluids (hydrocarbons, refrigerants, siloxane) as working fluid are called ORC (Organic Rankine Cycles). These cycles are applied in a large number of energy systems related with renewable energy and energy efficiency improvement. 8.4.1 Economics The cost of a plant mainly consists of the cost of the heat exchangers and the turbine. The cost distribution between heat exchangers and turbine is certainly influenced by the efficiency of the cycle: the lower this parameter is, the greater the cost of the heat exchangers is (and especially the condenser) compared to the cost of the turbine. The other aspect that affects the cost of the two components are the different economies of scale that can be exploited. Heat exchangers, Shell&Tube for example, are components that have a certain maximum size in both diameter and length of the Shell (TEMA sheets), and consequently a modular solution is used for large plants by placing multiple components in series and/or in parallel with a specific cost for the approximately constant transfer surface and small economies of scale. The turbine on the other hand (if the need to have a cross-over in the very large plants is ruled out) is a scalable component and has high reductions in the specific cost per kW produced as the size increases. Using the same working fluid for the same inlet and outlet conditions and same speeds, the number of stages and the optimal geometry of the machine does not change. Therefore, its cost is proportional to the diameter of the machine itself, which in turn is linked to the square root of the flow rate on a mass-basis with obvious economic improvement for machines of greater power. These considerations are not always valid and in particular do not apply if: • the machines are very small and there are technological limitations in the machines themselves; • for very large machines in which the crossover in multiple turbine cylinders is required; • for expansions that do not work under the same inlet and outlet conditions; 95 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 • for fluids with a different complexity and molecular weight. With the same expansion ratio π½, a greater molecular complexity involves greater volumetric expansion ratios and an expansion that must be divided into a greater number of stages in order to ensure a high efficiency. As already stated, the molecular weight affects the enthalpy change, and once again the number of stages of the turbomachine that cannot be excessively filled. Even if all these exceptions for plants with comparable efficiency are considered, it is correct to ponder that the cost of the heat exchangers has more weight for large-scale plants, whereas the turbine is the main cost for the small ones. 8.4.2 Considerations concerning the choice of the working fluid The choice of the working fluid depends strongly on the size of the plant, on the maximum temperature of the heat source and on its heat capacity (if it can be considered at constant or variable temperature), on technical limitations on the components, and on the chemical stability of the working fluid itself. Once again remember that the effects of the molecular weight and molecular complexity on the thermodynamic properties of the fluids are separated. Although very often highly complex fluids are also heavy fluids, it is recommended to not confuse these two properties and their effects on the thermodynamic cycles. Effect of the molecular complexity If the heat source is at constant temperature and there are no technical limitations that force a minimum vapor quality at the turbine outlet, it can be stated that is always theoretically possible to carry out a subcritical saturated reversible cycle both with simple fluids and complex fluids. It is necessary to use infinite regenerative bleeds in the first case and an infinitely complex fluid in the other case. In case of real cycles, the efficiency will be reduced by the losses due to fluid-dynamics and heat exchange. From the technological point of view, regeneration is quite different according to the case: • in a single molecule cycle, it is more complex and costly, it requires regenerative bleeds and a high number of additional components; • in the case of complex molecule, a surface recuperator that does not considerably change the architecture of the cycle and that attains high benefits in terms of efficiency is adopted. The turbine design is also different when the type of fluid changes: • For simple molecules, the expansion is in the two-phase field with problems related to the presence of drops of liquid hammering of the surfaces, and reduction of the performance and useful life of the component. The volumetric ratio of expansion, however, is slightly less with the same π½ compared to the complex fluids. • Expansion is in the field of superheated vapor for complex molecules. The design of the turbine, however, can be more critical due to the higher volumetric expansion ratios. When the minimum vapor quality at the turbine outlet is a technical issue, the advantage in using complex fluids is evident. In case of simple fluids, superheated cycles are required, reducing the evaporation temperature and reducing the performances of the cycle because of the large temperature differences and the consequent losses. When the heat source has a finite heat capacity (variable temperature), complex working fluids exhibit advantages because of the possibility to reduce the temperature differences in the heat introduction process. In addition, it is possible to raise the maximum temperature of the cycle. In presence of limitations on the minimum temperature of the heat source, a recuperator downstream the turbine can 96 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 be included, increasing the efficiency without reducing the heat source exploitation. When the heat source changes its temperature, supercritical cycles with complex molecules are even more convenient because of the better coupling between the fluid and the heat source thermal profiles. In conclusion, a complex fluid can be convenient both in case of constant and variable temperature heat sources, when the maximum temperature is not sufficient to design an high-efficiency water steam cycle. Effect of the molecular weight The molecular weight of the fluid instead has no influence on thermodynamic considerations and on the ideal cycles; on the other hand, it has a strong impact on the design of the components, in particular on the turbine. An increase in molecular weight has the following results : • the flow rate on a mass-basis circulating in the cycle increases with the same thermal power introduced because both evaporation heat and specific heat capacity decrease. • the turbine enthalpy drop reduces: this has important consequences on the design of the turbomachine, which can be made at the limit with only one slightly loaded stage. As will be later seen, this is not entirely true since other limitations related to the maximum volumetric ratio of expansion per stage are considered in the choice of the number of stages. • for an ideal gas, sound speed decreases when molecular mass increases, leading to a difficult design of the turbine blades, increasing the risk of shock waves and reducing offdesign performances. The choice of a heavy fluid can be useful in order to improve the design of the turbomachinery that are more economic than the corresponding ones in water vapor cycles. Effect of the critical properties As the critical temperature of the two fluids increases (or as the evaporation and condensation temperatures decrease), the reduced temperatures within which the cycle evolves decrease, but the results reported do not change significantly. It is however pointed out that, independently from the complexity and the weight of the fluid: • given the same evaporation temperature, ββππ£π increases as the critical temperature increases, and so both the fluids witness a decrease in efficiency losses associated with the introduction and transfer of heat at constant temperature. In case of variable temperature heat sources, it can be convenient to choose a fluid with a low critical temperature to reduce the relative importance of the evaporation phase, or to adopt a supercritical cycle. • The higher the critical temperature, the smaller the pressure values of the cycle. It results in higher expansion ratios of the turbine and lower condensation pressure. The design of the turbine is more difficult (higher number of stages and higher volumetric flow rates at the outlet), and the issue of leakages of incondensable gases increases. The effect of the critical pressure is certainly present, but has a lower impact because of the exponential relation between the saturation pressure and temperature. • On the contrary, given the same evaporation and condensation temperatures, as well as the same critical pressure, the lower is the critical temperature, the higher is the pressure of the vapor at the turbine inlet and the lower the volumetric flow rate. Moreover, once fixed the maximum temperature of the heat source, the choice of a fluid with an optimal critical temperature can give some advantages: • It is always possible to realize a supercritical cycle in presence of variable temperature heat sources. 97 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 • • For large plants with low heat source temperature, it is convenient to use fluids with low critical temperature, reducing issues related with the vacuum in the condenser and the volumetric flow at turbine outlet, as well as the number of stages consequently. This is the typical application for geothermal. Water as working fluid would lead to low efficiency cycles, and expensive and large turbines. For small plants with high temperatures (biomass, heat recovery, small concentrating solar plants, domestic cogeneration), it is convenient to choose a fluid with high critical temperature, but low critical pressure, resulting in low density vapor and high volumetric flow rates. Water as working fluid (that has a high critical pressure) would lead to very small blades in the turbine, with a strong decrease in expansion performances. This plants are adopted usually in cogeneration architecture aiming at the reduction of expansion ratios in the turbine. Other comments These observations must necessarily be associated with others related to the safety and cost of the working fluid. Certainly, water is an optimal fluid on many levels: it is cheap, non toxic, corrosive, flammable, it has no impact on the ozone layer (ODP) and on the global warming (GWP), it is stable up to very high temperatures, and has very good heat exchange properties. On the contrary, the organic fluids are often flammable (hydrocarbons), toxic (methanol, benzene, toluene) or they present an high ODP (chlorinated compounds now forbidden), or GWP (in particular, the refrigerants). In addition, they have lower heat exchange coefficients in comparison with water, in particular in case of mixtures. It is also mandatory to take into account the thermal stability of the fluids that beyond a certain threshold tend to show dissociation phenomena with consequent modification of the thermodynamic properties and formation of sludge which can damage the components. Specifically, the fluids that can be used in power production plants are water (a simple and light fluid) and organic fluids (certainly more complex and generally much heavier than water). Primarily for safety reasons, liquid metals that have high ππ and low complexity, and relatively simple molecules but with a molecular weight higher than that of water, such as ammonia or methanol, are excluded from the comparison. For large power plants with high maximum temperature sources the advantage of using water as working fluid is clear due to the lower costs, the higher maximum temperature of the cycle and the lower safety risks. The reference examples are: • the vapor section of a large combined cycle in which the choice to increase the plant complexity with HRSGs having three evaporation levels and RH is economically advantageous. • Fossil-fired power plants, in which the turbine has a high number of stages in view of the high enthalpy change, and the plant engineering complication created by the regenerative bleeds relatively impacts on the overall plant cost in which the heat exchangers, and in particular the steam generator and condenser, weigh more. • Nuclear plants in which the cost of nuclear reactor is the largest cost. For small or low temperature plants, it is instead expedient to use an appropriate organic fluid (among tens of candidate fluids) in saturated or supercritical cycles that can better connect with the heat source with high plant engineering simplicity. Added to this, the advantages linked to a simpler and less expensive turbine should then be considered. Despite having a high cost due to the poor transfer 98 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 coefficients, the recuperator is recommended always since it generally leads to a real reduction in the cost of the electricity produced. In general, the choice must be made based on thermodynamic, technical and economic considerations, calculating the benefits in terms of potential increase in efficiency and cost. Fig. 8.8 shows the field of application of the organic fluid engines and the applications where they can be used. STEAM THERMAL STABILITY LIMIT 400 grey zone Heat source temperature, C 340 3 7 260 5 2 1 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 8 9 180 10 4 6 100 11 20 Domestic cogeneration Remote applications Automotive engines WHR Rural solar energy Small solar plants Big solar plants Biofuels engines WHR Biomass combustion WHR from industrial plants High T. geothermal brines Low T. geothermal brines OTEC 12 1 kW 10 kW 100 kW 1MW 10 MW power output Fig. 8.8 - Field of application for the organic fluid engines 8.4.3 Mixtures of fluids The discussion up to now has been focused on pure fluids, but the use of mixtures of fluids can be interesting both for exploiting sources at variable temperature and for improving the properties of the resulting fluid. Use of a mixture in case of variable temperature heat sources Mixtures can be advantageous when both the heat source and the heat sink are at variable temperature. This is due to the fact that the phase change, evaporation for example, is not isothermal and takes place with a rise in temperature. A mixture formed by two miscible fluids that at the same temperature have two different saturation temperatures is considered. The fluid F1 has a higher saturation temperature, a lower saturation pressure, less volatility and is called high-boiling; the other F2 is more volatile and is called, on the contrary, low-boiling. As the mixture composition varies for a certain pressure, the bubble and dew points for the mixture, i.e. the temperature at which the first bubble is formed due to an evaporation process or the first drop due to a condensation process, can be defined. Except for azeotropic compositions, in which the two points coincide, the temperature increases as evaporation progresses for all the other compositions. 99 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 Fig. 8.9 - Notation used for defining the temperature glide for a phase transition of a zeotropic mixture Starting from a mixture with composition of liquid π₯ π1 , the first bubble forms at the temperature π1 , has the composition π₯ π£1 , and is obviously richer in the more volatile component. The liquid phase is now enriched in the less volatile component and will have π₯ π2 composition. The evaporation temperature is π2 > π1 and the vapor formed has π₯ π£2 composition. The process continues until the liquid has π₯ π3 composition corresponding at a temperature π3 . The evaporation process therefore takes place with a temperature variation called glide. To implement this process, once-through heat exchangers in which there is no free open surface evaporation like in the pool boiling evaporators are required. Then by choosing an appropriate mixture of fluids it is possible to get an evaporation that follows the cooling of the source well, with a consequent reduction of the heat introduction into the cycle losses. The same conclusions apply, however, also to the condensation of the fluid and therefore very often the advantage during the introduction of heat phase is balanced by an increase in the loss relating to the transfer of heat to the environment. Therefore, fluid mixtures are particularly interesting when both the hot sink and cold sink are at variable temperature. This is the case of a cogeneration plant in which the cold sink has a certain current of water (possibly pressurized), which heats up to the conditions required by the thermal load. In this case, the reversible cycle of reference is the mixtilinear cycle, in which both heat sources change their temperatures in the process. The use of mixtures makes it possible to approximate this cycle as shown in the TQ diagrams in Fig. 8.10, where the diagram of heat transfer of the saturated cycle obtainable with a pure fluid and that obtainable with the optimum mixture is shown. In this specific case the increase in efficiency is higher than 8.8 percentage points. 100 of 124 Tempe 150 100 150 transition1500 lead differences in th 1000 temperature diff pushed to 500 the lo 0 surface. 100 Second-law analysis of power cycles - Energy Conversion A – V7.0 50 50 0 5 10 15 0 Power, MW 300 500 1000 300 60004000 Entropy, J/(kg K) 300 PU Temperature, °C Pure fluid 250 250 200 200 150 150 100 100 50 300 300 55 10 10 15 15 00 300 300 Power, MW 250 250 250 250 200 200 200 200 150 150 150 150 100 100 100 100 50 50 500 500 10001000 1500 300 300 6000 6000 Entropy, J/(kg K) En Optimal mixture 5000 5000TOLUE TOLUEN A higher temp 4000 4000 Ahighlighted relatively lim s 3000 involves releva the 3000 economiz phase transition enhance the cy 2000 larger than the the 2000 reductio colder temperatu temperature di 1000 1000 50 50 00 55 10 10 results obtainable 15 for15 pure 0 0 and 500 500 mixture 1000 1500 1500 with 1000 optimal toluene Fig. 8.10 - Numerical Toluene, a Entropy, J/(kg K) 300 300 Power, MW MW 300 300 Power, Entropy, J/(kg K) ethanol composition for biomass-fired cogeneration applications Temperature, °C Temperature, °C inlet. Both evap 30002000 involves releva characterized b phase transition transition lead 1500 larger2000 thaninthe differences th 1000 colder1000 temperatu temperature diff 500 pushed to the lo surface. 0 0 50 00 Temperature, °C Temperature, °C 3500 5000 The optimal cyc TOLUEN 3000 an almost 4000 satu 2500 s A relatively 250 250 and 0 0 300 300 6000 En TOLUE 5000 TOLUE 250 250 Mixtures, however, are used very little since in the case of leaks there is a change in the composition A higher incre temp of the fluid and proceeding with the make up. Another aspectAisfurther4000 200the need for chemical analyses before200 200 200 highlighted lim to small temp related to the coefficient of heat transfer in the evaporation phase: for pure fluids it is very high, while 3000 the economiz with an incr for mixtures it150 may be considerably lower. This fact combined with the reduction of the temperature 150 150 150 enhance the cy temperature di difference in the heat exchanger involves a considerable increase in the heat transfer surface and the 2000 the reductio cost of the heat exchanger. higher second 100 100 100 100 temperature 1000di Use of a mixture to change the critical properties of the50fluid 50 50 50 00 55 10 10 15 15 0 0 5 5 10 10 15 15 00 500 500 1000 1000 1500 1500 2000 0 5000 300 Temperature, °C Temperature,°C 300 300 The typical application is thePower, closed CO cycles are developed solar Power, MW2 cycle. These300 Entropy, for J/(kg K) power plants and 300 MW Entropy, J/(kg K) for improving flexibility of coal power plants. Increasing the maximum pressure of the cycle, the 4000 250 250 distance from250 the saturation line is reduced yielding 250 a decrease of the compression work due to the TOLUE TOLUE lower compressibility factor. A condensation cycle would be even more advantageous, because the A further incre 3000 200 200 compression 200 phase could be managed with a pump. 200 Unfortunately, it is not possible to realize this The critical te to small temp cycle for solar applications because the availability of low-temperature (15-20°C) heat sinks is usually cycle adopts with an 2000 incr 150 150 limited in typical solar plant locations. This temperature requirement is fixed by the critical 150 150 point tempe temperature of the CO2 that is 31°C. Nevertheless, it is possible to mix the CO2 with small amountstemperature of di recuperator is other substances 100 with a higher critical temperature, aiming 100 at increasing the critical temperature of the higher second 100 100 economizer1000 on mixture; then, condensation cycles become possible with the mere availability of cooling air at 3035°C. 50 50 50 50 mperature, °C 300 250 200 150 Power, MW Power, MW 300 0 0 250 101 of 200 124 150 500 500 1000 1000 1500 1500 Entropy, J/(kg K) Entropy, J/(kg K) 2000 2000 0 2 5000 4000 TOLUE 3000 The critical te cycle adopts 2000 point tempe Second-law analysis of power cycles - Energy Conversion A – V7.0 9 SECOND LAW ANALYSIS OF AN OPEN-LOOP BRAYTON GAS CYCLE In this chapter, the list of losses in open-loop cycles is shown. The framework presented in Chapter 6 is first outlined, then the losses are evaluated and commented. Physical (or conceptual) boundaries of the system considered The boundaries of the system are: (i) the section where the air enters the filter and fuel chamber and (ii) the section where the combustion products exit to the turbine discharge (or of the recuperator, if there is one). All the auxiliary systems are powered inside the system, from which the output power then exits. Choosing the dead state In the case of internal combustion engines, such as the open cycle gas turbines, choosing the dead state is mandatory, and the atmospheric air is chosen. The nominal conditions can refer to the conditions: • “ISO”, meaning 15 C, 1.01325 bar, standard composition, U.R: 60%, or • “site”, those of the site where the gas turbine is installed • Real. Choosing the energy source For static applications, the most widely used energy source is natural gas, which is defined through: • Chemical composition • Temperature • Pressure Of the various energy indexes used for the fuel, the preferable one for the second law analysis is ππππ£ . It is not tied only to the combustion reactions, but takes into account the temperature and pressure, in addition to the composition. The temperature is often different from that of the gas pipeline due to preheating of the fuel. The fuel pressure is considerably higher (often 10 or more bar) than that of the combustor due to the presence of control valves upstream of the combustor. Reference to the rated instantaneous quantities Reference will be made to the powers under nominal conditions in the following analysis. 9.1 CRITERIA FOR CHOOSING THE COMPRESSION RATIO There are two performance indexes for an open cycle gas turbine: efficiency and specific work. π= πΜπ’ πΜπππ£,ππ’ππ πΜπ’ π€π’ = πΜπππ (9.1) (9.2) The first accounts for how the fuel is used and may change numerically from one country to the next depending on the convention taken for the πΜπππ£,ππ’ππ , or if it refers to the LHV, the HHV, whether or not it takes into account the gas pressure in the gas pipeline upstream of the combustor. The second, on the other hand, accounts for the power of the plant in connection with its physical size. It is not calculated as a difference of the specific turbine and compressor works since the flow rates in a gas turbine cycle are different and change both along the compression (drawing air to cool the blades) and along the expansion (mixing with cold air). The cost of the electricity produced depends on both the investment cost and the variable costs. The latter term depends on the fuel cost and the efficiency of the plant. The former term, on the other hand, 102 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 depends on the machine cost related to the installed power and, therefore, the inverse of the specific work. In fact, it can be considered that the machine cost depends on its mass, and therefore on its volume. In considering the machine as a cylinder of size π΄ (frontal cross section) and πΏ (length) (see Fig. 9.1), it mainly depends on the mass of aspirated air and on the pressure ratio. It is considered that the intake sections at the compressor and the discharge sections of turbine π΄ depend on the volumetric flow rate that crosses them ππΜ and on the axial speed of the flow ππ π΄= πΜπππ πΜπππ = ππ πππ (9.3) where the axial speed cannot be changed at will since the pressure drops increase as the speed rises and the velocity triangles have to be well proportioned. Therefore, this parameter is independent from the compression ratio and the power of the machine. Moreover, in open gas cycles it is not possible to pressurize the cycle and change the fluid density. Thus, also the term ππ have any influence. As a result, the passage area depends on only the flow rate on a mass-basis and, therefore, on the circular crown defined by the blade height β and the average diameter π·π of the stage: 2 πΜπππ ∝ π΄ = ππ·π β = ππ·π β π·π (9.4) As will be seen further on, not even the blade height can be chosen regardless of the need to have high efficiency stages. In particular, it is decided to set a certain optimum β/π·π . The machine length can be considered a function of the number of compressor stages. This number depends on the compression ratio and, considering a π½πππ value per stage, it will be equal to ππππΊ = next whole [ ππ (π½) ] ππ(π½πππ ) (9.5) The volume of the gas turbine is therefore proportional to the flow rate of aspirated air πΜπ ∝ π·π 2 and the logarithm of the pressure ratio. Of the two effects, the one linked to the front dimension predominates, which is therefore the parameter of interest for calculating the specific cost of a gas turbine. It is therefore correct to consider the specific work as an index of the specific power produced. A L Fig. 9.1 - Diagram of the overall dimensions of a gas turbine in simple cycle configuration. 103 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 Reasoning with TIT (Turbine Inlet Temperature) being equal in simple cycle and increasing the compression ratio, both indexes initially rise up until the maximum π€π’ is reached. For higher π½, the specific work decreases, and the maximum efficiency is reached. The trend on the π − π€π’ plane is shown in Fig. 9.2 for two different TITs. Fig. 9.2 - Change in efficiency and specific useful work as the compression ratio changes for a simple cycle gas turbine Choosing the optimum compression ratio is therefore based on the need on the one hand to maximize efficiency and hence reduce the variable operating costs, and on the other that of having small machines and hence with a lower investment cost. There are two philosophies: • Heavy Duty Machine or industrial machinery. These are machines that work very few hours a year (certainly less than 1000 hours a year, sometimes less than a hundred) to cover the peaks, and so it is necessary that they have the lowest possible investment costs. Work is carried out in maximum π€π’ conditions. The optimum compression ratios are low, 12-15, and the efficiencies are between 35% and 38-40%. They are recommended in cogeneration applications, where the discharged heat can be used. • Jet Machines or aeroderivative machinery. They aim at maximizing efficiency and therefore at minimizing the use of fuel. They are designed to work many hours a year, and the optimum compression ratio is higher than the previous case (about 30). Special attention have to be paid to the polytropic efficiency of the turbine and compressor since the efficiency loss linearly depends on the ππ(π½) Now, focus will be placed on defining the optimum compression ratio in the perspective of maximizing the cycle efficiency, based on the second law analysis and applying the methodology to a simple cycle turbine, a turbine in recuperated cycle, and providing some information on the cooled turbines usually used in industrial practice. 9.2 SIMPLE CYCLE WITH NON-COOLED TURBINE This is a not very realistic example because today all modern open cycle gas turbines are equipped with turbine blade cooling, but it is helpful for teaching purposes in order to present the characteristic pattern of the typical losses of a gas cycle. If the blades are not cooled, it means that flue gases have a maximum temperature of about 800-900 °C and the combustion is carried out with a large excess of air. 104 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 Fig. 9.3 - Section plane of a modern gas turbine in simple cycle The cycle is outlined with the following seven irreversible processes: • Δπ1 irreversibility due to pressure drops (in the air filtration system and in every other part of the cycle) • Δπ2 fluid dynamic irreversibilities in the compressor • Δπ3 combustion irreversibility • Δπ4 fluid dynamic irreversibilities of the turbine (adiabatic supposition) • Δπ5 losses due to the discharge of exhaust gas into the atmosphere • Δπ6 thermal losses • Δπ7 mechanical/electrical/auxiliary losses mf 3 2 4 3 5 1 4 2 0 0 5 1 Fig. 9.4 - Simple cycle of a gas turbine in the T-S plane. Adiabatic turbomachinery are hypothesized and the thermal losses (present in various parts of the plant) are not represented Every one of them involves an efficiency loss whose dependency on the compression ratio π½ is examined. The analysis is conducted keeping the maximum cycle temperature value constant. Consistent with the hypotheses adopted above, the general formula is the following: 105 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 πΜπ’ πΜπππ£,ππ’ππ 7 7 πΜπππ£,ππ’ππ βπΜπ π0 βπΜ π = −∑ =1−∑ πΜππ’ππ π€πππ£,ππ’ππ πΜπππ£,ππ’ππ πΜπππ£ 1 7 π = 1 − ∑ βππ (9.6) 1 (9.7) 1 where πΜππ’ππ is the fuel flow rate. The analysis will be presented with the maximum cycle temperature being equal, while changing the compression ratio: as the π½ rises, the temperature of the air discharged by the compressor increases, so the flow rate of required fuel decreases, which can be calculated by the enthalpy balance of the combustor. The parameter πΜπππ ⁄πΜππ’ππ increases. 9.2.1 π«πΌπ irreversibility due to pressure drops (in the air filtration system and in every other part of the cycle) The ambient air in a gas cycle is aspirated by the compressor and through a filter chamber in which all the impurities suspended in it are removed in order to protect the compressor blades and limit fouling phenomena. The filters used are so efficient that the GT machines outlet air can contain less particulate matter than that contained in the aspirated air. The intake process is adiabatic and preserves the total enthalpy and total temperature. Thus, acceleration of the air up to the inlet condition of the compressor entails a temperature reduction. To prevent the formation of frost, it is necessary to keep the temperature at an appropriate value by using anti-icing systems. These systems heat the air before the filter chamber to prevent ice crystals from depositing on the filtering surfaces (and consequently increasing pressure drops) or on the compressor blades. The heat can come from the following sources: • heat exchanger with hot water at 30-40°C heated with the gases discharged by the turbine; • by mixing the aspirated air with the hot air of the turbine enclosure. In these cases the process is not adiabatic and requires the fluid to be heated; this is always penalizing for the cycle performance since the mass flow rate aspirated by the compressor is always reduced because the compressor works at constant volumetric flow rate, in absence of VGV as it will be discussed later. The anti-icing system has to be enabled only if strictly necessary, or for temperatures lower than 6-7°C and relative humidity of about 50%. In other cases, the opposite process is carried out, meaning cooling the aspirated air to increase the power produced by the cycle. Cooling can be achieved with: • battery with cold water chilled by a refrigeration cycle, that however demands added consumption; • a spray of cold water in a thermal and mass exchange process. From the energy viewpoint, it is interesting, but if the water is not perfectly demineralized, it can foul the surfaces. A measure of this type is usually poorly tolerated by the GT manufacturers that provide very stringent criteria on the flows entering the machinery. In the case of GT cycles, it is absolutely legitimate to consider the aspirated air an ideal gas since the ambient conditions are very far away from the saturation curve. The specific heat of the air does not depend then on the pressure, but only on the temperature since the air is made up of diatomic and triatomic molecules. 106 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 π π0 π π’ βππ βπ1 = ∑ πΜπ πΜππ’ππ π€πππ£,ππ’ππ ππ ππ π=1 (9.8) where π is the number of pressure drops (finite, but relatively small in relative terms, otherwise the βπ term π is to be replaced with ∫ πππ /ππ ). βπ1 is directly proportional to πΜπππ ⁄πΜππ’ππ and, therefore, ππ increases as β increases. The following important elements in the equation that defines the loss of efficiency has to be noted: • There is no dependency on the temperature at which throttling takes place; even if the dissipated infinitesimal work πππ€,π = π£ππ is proportionate to the specific volume (and hence to the temperature), part of this might be recovered in other processes, so the loss of useful work calculates only the amount lost forever. • The absolute pressure drop does not count, only the relative does. In other words, the same Δπ is paid to a different extent depending on the pressure at which it takes place (the higher it is, the less is paid). • The drop is tied to the ratio between the flow rate on a mass-basis that the throttling sustains and the flow rate of fuel: flow rates being equal, the greater the flow rate of the fuel introduced into the cycle, the less the efficiency loss is. 9.2.2 π«πΌπ fluid dynamic irreversibilities in the compressor The compressor is supposed as an adiabatic machine, distinguished by a polytropic efficiency ππ¦,πΆ . The efficiency loss associated with this component is: βπ2 = 1 − Μ Μ Μ Μ Μ Μ Μ Μ ππ¦,πΆππ π0 πΜπππ π π’ πππ’π‘ πΜ ( ) ππ ( ) ππ ππ¦,πΆππ Μ Μ Μ Μ Μ Μ Μ Μ πππ (9.9) One can note that βπ2 grows as the compression ratio increases and that it does not depend on the temperature at which the compression takes place. The increase of the term πΜπππ ⁄πΜππ’ππ further contributes to the rise in βπ2 with π½. The importance of the efficiency loss is naturally correlated with the "fluid dynamic quality" of the component defined by ππ¦,πΆππ . 9.2.3 π«πΌπ combustion irreversibility The combustion reaction takes place in the combustor, in which the fuel (usually natural gas) reacts with the oxidant (compressed ambient air) to form combustion products at a temperature that is based on the preheating of the reactants and the dilution of the combustion. The irreversibility of combustion is based on: • the temperature of the reactants, meaning the air and fuel. Since the heat capacity of the combustion air is much greater than that of the fuel, the benefits obtainable as the air temperature increases are much greater than those obtainable by increasing the fuel temperature. • the temperature of the combustion products, which is equal to the adiabatic flame temperature, coinciding with the turbine inlet temperature. 107 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 ππ ππ π0 π0 Μ βπ3 = βπππππ = [∑(πΜπ π π )π − ∑(πΜπ π π )π ] πΜππ’ππ π€πππ£,ππ’ππ πΜππ’ππ π€πππ£,ππ’ππ π=1 (9.10) π=1 To calculate this formula, it is necessary to use thermodynamic tables to calculate the absolute entropies and the trend of specific heat of ideal gas with the temperature. As the compression ratio Μ increases, the combustor inlet air temperature increases: βπππππ and also πΜππ’ππ decrease. Finally, the βπ3 loss decreases. The same result is obtained by associating πΜππ’ππ π€πππ£,ππ’ππ at the thermal power produced in the combustion, potentially at infinite temperature. The equation can be expressed in approximated terms (the heat capacity of the fuel is neglected) as: βπ3 = πΜππ’ππ π€πππ£,ππ’ππ πΜππ’ππ π€πππ£,ππ’ππ π0 π0 ( − )= πΜππ’ππ π€πππ£,ππ’ππ ππππ|ππππ₯,π2 π∞ ππππ|ππΌπ,π2 (9.11) where π2 is the temperature of the reactant made up main of combustor inlet air. The equation highlights that the efficiency loss decreases as the mean temperature between the reactants and products increases. With ππΌπ the same, the temperature of the combustion products is fixed, so the loss depends only on the combustor inlet temperature, and therefore decreases as the compression ratio increases, always and in any case keeping itself at high values. 9.2.4 π«πΌπ fluid dynamic irreversibilities of the turbine (adiabatic assumption) Consistently with the hypothesis of adiabatic expansion, the turbine is assumed a machine distinguished by a polytropic efficiency ππ¦,ππ π΅ . The efficiency loss associated with this component is: βπ4 = π0 πΜππ₯β π π’ πππ πΜ (1 − Μ Μ Μ Μ Μ Μ Μ Μ )ππ ππ¦,ππ π΅ ( ) πΜππ’ππ π€πππ£,ππ’ππ ππ πππ’π‘ (9.12) One can note that as for the compressor, as the compression ratio rises, βπ4 increases both due to greater π½ and because of the increase of the term πΜππ₯β ⁄πΜππ’ππ . As always in the case of losses linked to fluid dynamics in the gas phase, it does not depend on the temperature at which the expansion itself takes place. The importance of the efficiency loss is naturally correlated with the "fluid dynamic quality" of the component defined by ππ¦,ππ π΅ . In the case of non-cooled machinery, the turbine is a less critical component of the compressor since with polytropic efficiency being the same, the term (1 − Μ Μ Μ Μ Μ Μ Μ Μ ) ππ¦,ππ π΅ will be lower than the term ( Μ Μ Μ Μ Μ Μ Μ Μ Μ Μ 1−π π¦,πΆππ Μ Μ Μ Μ Μ Μ Μ Μ Μ Μ π π¦,πΆππ ). It should also be noted that the turbine will have a lower expansion ratio than that of the compressor due to the pressure drops. 9.2.5 π«πΌπ losses due to the discharge of exhaust gas into the atmosphere (stack losses) The discharge of the combustion products downstream of the turbine (at temperature TOT) is assimilable to a heat transfer (with infinite transfer surface) with the environment at temperature π0 . π0 Μ Μ (βπππ₯β + βππππ ) πΜππ’ππ π€πππ£,ππ’ππ π0 1 1 βπ5 = πΜππ₯β,π ππ (πππ − π0 ) ( − ) πΜππ’ππ π€πππ£,ππ’ππ π0 ππππ|πππ,π0 βπ5 = 108 of 124 (9.13) (9.14) Second-law analysis of power cycles - Energy Conversion A – V7.0 βπ5 = πΜππ₯β,π ππ (πππ − π0 ) π0 (1 − ) πΜππ’ππ π€πππ£,ππ’ππ ππππ|πππ,π0 (9.15) As the compression ratio increases, the discharge temperature decreases on the one hand, and the term πΜππ₯β,π /πΜππ’ππ increases on the other. Of the two effects, the first is predominant and this efficiency loss decreases as π½ increases. It is also seen that the term in brackets at the bottom of the equation expresses the efficiency of a trilateral cycle operating with the hot gases as the heat source and with the environment. 9.2.6 π«πΌπ thermal losses Assuming that the system, in a certain point of the plant, loses an infinitesimal thermal power πππππ π,π = πΜπππ πβπππ π,π originating starting from a temperature ππ₯ to the environment, the resulting efficiency loss is equal to: π ππΜπππ π,π ππΜπππ π,π π0 βπ6 = ∑ [∫ −∫ ] πΜππ’ππ π€πππ£,ππ’ππ π0 ππ₯,π π=1 (9.16) π πΜπππ π0 βπ6 = ∑ πβπππ π,π (1 − ) πΜππ’ππ π€πππ£,ππ’ππ ππππ,π (9.17) π=1 With specific thermal losses being the same as the fluid flow rate, the term increases with the ratio πΜπππ ⁄πΜππ’ππ and therefore with the rise in π½, with the same law already found for Δπ1 . 9.2.7 π«πΌπ mechanical/electrical/auxiliary losses The mechanical and electrical losses depend on the gross power of the compressor and turbine, the electrical losses on the mechanical power at the generator inlet. The consumption of the auxiliary systems is usually defined as a fixed portion of the electrical power produced. All these quantities are therefore proportional to the power produced. βπ7 = 9.2.8 Μ ∑π π=1 βπππ,ππ’π₯ πΜππ’ππ π€πππ£,ππ’ππ (9.18) Observations The trends represented in Fig. 9.5 are based on π½, and are obtained for π0 =15°C, ππΌπ=900°C, polytropic turbine and compressor efficiency of 90%, relative pressure drops equal to 10%, concentrated thermal losses at the turbine outlet (πππ) equal to 10°C, mech/el/aux losses equal to 3% of the useful power. 109 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 0.6 1 0.9 0.5 0.8 0.7 0.4 0.6 0.3 0.5 0.4 0.2 0.3 0.2 0.1 0.1 0 0 0 10 20 30 40 50 0 10 20 30 40 50 Fig. 9.5 - Trend of the efficiency losses (left) and cumulative diagram (right) of a single gas cycle as the compression ratio changes. One can note: • that for the losses linked to the pressure drops (1) and to the thermal losses (6), they rise (although weakly) as the compression ratio increases. • that the losses in the turbomachinery (2 and 4) rise (a lot) as the compression ratio increases. • that the combustion losses (3) and discharge losses (5) are very significant and decrease (the discharge losses very much, the combustion losses less) as the compression ratio increases. • since some losses (above all those in the turbomachinery) increase as the compression ratio increases and others (combustion and discharge of the combustion products) decrease, there is an optimal compression ratio between 15 and 20, for which the sum of the losses is minimum and the efficiency is maximum. • The optimum compression ratio for the useful work is lower, as already noted previously. 200 40% 35% 150 25% 100 20% 15% 50 10% efficiency 30% 5% 0 0% 0 10 20 30 40 50 Fig. 9.6 - Trend of the specific work and efficiency as the compression ratio varies. The maximum values are at different π½ 9.3 RECUPERATED CYCLE WITH NON-COOLED TURBINE Unlike the previous case, recuperated cycles with non-cooled turbine are used in practice, in particular for industrial or domestic cogeneration applications, or for the exploitation of renewable energy sources (in particular, concentrating solar power towers). They are usually small gas turbines (300 kW 110 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 - 5 MW) because the heat exchanger is a critical component due to the low heat exchange coefficients and the high operating temperatures. It could be very expensive for sizes larger than 5 MW. Small sizes and high plant engineering simplicity requirements lead to a system without blades cooling and therefore with maximum temperatures (TIT) of about 800-900 °C. Fig. 9.7 - Section plane of a modern gas turbine in recuperated cycle Now a recuperated cycle will be considered. A schematic of such systems is shown in Fig. 9.7, where a highlight on the recuperator is placed, while the Ts and architecture are shown in Fig. 9.8. 3 mf 4 2’ 2’ 3 4’ 2 1 5 4 2 4’ 0 0 5 1 Fig. 9.8 - Recuperated cycle of a gas turbine in the T-S plane. Adiabatic turbomachinery are hypothesized and the thermal losses (present in various parts of the plant) are not represented. In addition to the seven losses considered in the previous case, an eighth one is added to them, tied to the thermal transfer irreversibilities in the heat exchanger. 9.3.1 π«πΌπ losses relating to the heat transfer in the recuperator The flue gas discharged by the turbine cool down in the recuperator to preheat the compressed air. 111 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 This component is a site of irreversibility since the heat transfer takes place under finite temperature differences. The heat capacities on the hot side and cold side are different since the specific heats of ideal gas are different and the flow rate of exhaust gas is higher than that of air. In the case of a noncooled turbine, the heat exchanger can theoretically work with very small temperature differences. However, the cost of the heat exchanger rises sharply as the βππππ drops because of the low heat transfer coefficients. In this example a pinch-point of 50°C on the hot end of the exchanger was assumed. The TQ diagram of the recuperator is shown in Fig. 9.9. Fig. 9.9 - TQ diagram of the recuperator of a non-cooled gas turbine 4′ 2′ π0 ππΜ ππΜ βπ8 = (∫ +∫ ) πΜππ’ππ π€πππ£,ππ’ππ 4 π π 2 π0 πΜπππ ππ,β − ππ,π βπ8 = ( ) πΜππ’ππ π€πππ£,ππ’ππ ππ,β ππ,π (9.19) (9.20) The formula points out that the loss is the more significant the higher the ratio between recuperated power and that introduced into the cycle is. This ratio decreases (up to the point of being canceled out) as the compression cycle of the cycle increases. Above this value of pressure ratio, it is not possible to build a recuperated cycle. As it is obvious, the loss is heavily influenced by the mean temperature difference between the two currents in the heat transfer. This difference cannot be small because the heat transfer coefficients of the two currents are low (low density gas), and also because high speeds cannot be adopted to limit the pressure drops, which in the gas cycles are strongly penalizing. 9.3.2 Observations The effects of the recuperator reflect on all seven terms βπ considered previously. Unlike what has been seen for the simple cycle, the heat that can be recovered is gradually lower as the compression ratio rises, more fuel is needed and the ratio πΜπππ ⁄πΜππ’ππ decreases. At the opposite limit for π½ = 1 there is no introduction of fuel and πΜπππ ⁄πΜππ’ππ → ∞. On the one hand, the denominator in the formula increases, so βπ tend to decrease, but on the other the irreversibilities in the combustion (the air is less preheated) and in the transfer of heat to the environment (the combustion products exit warmer) increase. The new trend of the various βπ is shown in the figure. The calculations are connected with a hypothetical cycle characterized by: π0 =15°C, ππΌπ=900°C, polytropic turbine and compressor efficiency 90%, relative pressure drops equal to 10%, concentrated thermal losses at the turbine outlet (πππ) equal to 10°C, mech/el/aux losses equal to 3% of the useful power, βππππ in the recuperator equal to 50 °C. 112 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 0.6 1 0.9 0.5 0.8 0.7 0.4 0.6 0.3 0.5 0.4 0.2 0.3 0.2 0.1 0.1 0 0 10 20 30 40 50 0 0 10 20 30 40 50 Fig. 9.10 - Trend of the efficiency losses (left) and accumulative diagram (right) of a recuperated gas cycle as the compression ratio changes. As a result, the following is obtained (as shown in Fig. 9.10): • the losses concerning pressure drops (1) and thermal losses (6) now decrease as the compression ratio rises as they are linked only to πΜπππ ⁄πΜππ’ππ and weigh more than in the previous case. • The losses concerning the turbomachinery (2 and 4) are rising since the effect of increased π½ is greater than that linked to the reduced πΜπππ ⁄πΜππ’ππ . They are less important than the previous case. • The combustion losses (3) continue to be highly significant, and increase (little) as the compression ratio increases. • That the losses at the discharge (5) rise (a lot) as the compression ratio increases. • The losses in the recuperator (8) are very large at the low compression ratios since the recoverable heat grows. They decrease as the compression ratio increases up to being canceled out when recuperation is no longer possible, or when the difference between TOT and π2 at the end of compression becomes lower than βππππ : a situation that in this case occurs due to π½~15. • Since some losses (combustion and discharge) increase as the compression ratio increases and others (above all those of the recuperator) decrease, there is an optimal relatively modest compression ratio (around 3), for which the sum of the losses is minimum and the efficiency is maximum. Recuperated gas turbines are designed with pressure ratios that are lower than for large simple cycle gas turbines. Then, it is possible to use single stage radial turbomachinery for both compression and expansion, yielding a strong reduction in costs. • The resulting optimal efficiency is about 10 points higher than for the simple cycle. 9.4 GAS CYCLE WITH COOLED EXPANSION As previously stated, it is difficult from the economic viewpoint to reach fluid temperatures higher than 600-620°C in steam power plants in which combustion is external. The temperatures felt by the materials are always higher than that of the fluid, and in view of the large transfer surfaces, the high thicknesses of the tubes is not advantageous in economic terms. It is best to use superalloys to build the SH and RH banks. Similar observations apply to closed gas cycle plants. The use of expensive materials able to withstand higher temperatures is instead common in the gas turbines, where the increased efficiency and specific work is strictly tied to the start of expansion 113 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 temperature. Investing in more expensive materials is economically advantageous in the case of a gas turbine since the combustion is internal (metal temperature lower than the fluid temperature) and the masses are small. In the last 50 years the maximum temperature has constantly increased. This result was attained with (i) the use of materials able to withstand higher temperatures, (ii) single crystal blade manufacturing methods, (iii) improved cooling techniques and (iv) use of protective ceramic coatings. The last two points are of fundamental importance since even the best materials from this point of view (NiCrMo superalloys and blades formed by growing a single crystal) are unable to tolerate temperatures over 900°C. Therefore, cooling flows are necessary, because they allow the maximum temperature of the metal to be limited adopting different technological solutions (internal convection, impingement, film cooling or transpiration). Fig. 9.11 - Technological solutions for cooling high temperature blading of a modern gas turbine If the technological development of gas turbines is examined in terms of TIT (Turbine Inlet Temperature), one notes the trend illustrated in Fig. 9.12: a constant increase over the years by about 12-13°C a year, due to improvement of the cooling technologies for about two-thirds, and to improvement of the materials (4-5°C) for about one-third. Fig. 9.12 - Left: historical trend of the turbine inlet temperatures over the years, with an increase of about 12-13°C/year. Right: historical trend of the maximum allowable temperatures of the materials used in the hot parts of the gas turbines. The increase is about 4-5°C/year for the industrial turbines. 114 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 The use of cooling flows on the one hand allows higher temperatures to be reached and to improve efficiency. On the other hand, added losses are inevitable due to (i) heat transfer, (ii) pressure drops of the cold air in the convective channels, (iii) the subsequent mixing with the exhaust gas, and (v) the modification of the flow in the channels and thus worsening of the fluid dynamics of the machine. The technological limit is transpiration cooling, which requires the use of porous materials and the formation of a low speed flow that marginally influences the main flow undergoing expansion. The qualitative trend of the increased efficiency attainable with a technological improvement, whether tied to the materials or to the cooling techniques, is shown in Fig. 9.13. The continuous lines represent a conventional technology: as the TIT grows, the cycle efficiency increases, but also a greater cooling air flow rate is necessary. The effect is positive up until when the benefits given by the increased temperature are canceled out by the increased losses due to the cooling of the blades and of the combustor surfaces. The optimum efficiency (π0 ) for a certain turbine inlet temperature is identified. cooling simple cycle If one wants to further increase the efficiency, it is necessary to use more efficient cooling methods or materials able to withstand higher temperatures. In both cases a smaller air flow rate is required and a switch is made to the efficiency π1 . Technological progress however makes it possible to go to higher TITs with coolant flow rates being equal and therefore to greater efficiencies (π2 ), until the loss due to the cooling becomes once again dominant and a new maximum efficiency (π3 ) at higher TITs is identified. The figure explains two important concepts: • There is a TIT value that optimizes the efficiency of a gas turbine, beyond which the irreversibilities tied to the cooling cancel the positive effects (in particular, but not only, the combustion irreversibilities) linked to the increase in TIT. • This value increases as the cooling technology improves. In addition to increasing TIT, also the flow rate of the cooling flows increases, with an overall positive effect on the efficiency. At the same time, the gap between the TIT and the temperatures of the cooled materials increases. Advanced technology Conventional technology Fig. 9.13 - Maximum efficiency and new TITs obtainable with a technological improvement 115 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 The hot parts of the machine are cooled with flows of air drawn from the compressor, some at intermediate pressures and some at the compressor delivery. The flows are classified in two broad categories, as can be seen in Fig. 9.14: • non-chargeable flows: are the compressed air flows (all drawn from the compressor delivery) that are injected upstream of the first turbine rotor, thus to cool the combustor and/or the first stator array. They contribute to the energy balance that defines the total temperature upstream of the first rotor array (hence from where the extraction of work from the cycle begins). This temperature will be called TIT (Turbine Inlet Temperature, sometimes called TRIT, Turbine Rotor Inlet Temperature). • chargeable flows: are the compressed air flows (independent from the compressor drawing points) that are used to cool the parts downstream of the first stator array (therefore downstream of TIT). They represent a thermodynamic penalization as they bypass the combustion. The chargeable flows are drawn at the appropriate pressure to be mixed at a certain point of the expansion and prevent further losses given by isenthalpic throttling processes. 2i 2ii “chargeable” cooling bleedings 2iii 3i 3 compressor 3ii 3iii 3iv “NON chargeable” cooling bleedings 1 Fig. 9.14 - Representation of the cooling flows drawn from the compressor of a gas turbine The temperatures that characterize a gas turbine are: • πΆππ (Combustor Outlet Temperature): is the total temperature downstream of the combustor, therefore downstream of the duct, called transition piece, which connects the combustor outlet with the turbine inlet. This component reduces the speed up to the value required at the turbine inlet, and makes the flow uniform. Naturally, πΆππ > ππΌπ, because the cooling flows (nonchargeable) mix with the main current in the first stator array. It is not measured but obtained from a mass and energy balance that considers all the non-chargeable flows used to cool the liner and transition piece. The configuration of the combustors and the solutions adopted have evolved over the years, changing from single combustors to silos, to many ring combustors able to guarantee a much more uniform flow at the turbine inlet and better control of the combustion process in order to limit the formation of pollutants. 116 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 Fig. 9.15 - Evolution of the combustors of gas turbine plants combustor walls transition piece 4 3 premixed combustion cooling bleedings for the combustor 4T COT~1600°C Combustor Outlet Temperature 3 3T TAF=COT 4 Fig. 9.16 - Diagram showing the distribution of the cooling flows in a gas turbine combustor and in the transition piece. First law balance for calculating the COT 117 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 The value of πΆππ has an influence on the emissions of nitrogen oxides and, generally speaking, the greater πΆππ is, the greater the NOx emissions are. They are limited by using premixed flames that give lower adiabatic flame temperatures and guarantee NOx values of 8-9ppm, an extremely lower value than just 60 years ago when diffusion flames were used that went over 1000 ppm. This technique however tends to be inefficient at the low loads where it is difficult to keep a premixed flame stable and switching to diffusion flames is required, with the resulting increase in nitrogen oxide emission. This is why every power plant is connected with an environmental protection agency EMS (Emission Monitoring System) that continuously analyzes the machine operation. • ππΌπ (Turbine Inlet Temperature). The value of ππΌπ is significant because it is tied to the flow temperature that strikes the high temperature part of the gas turbine most highly stressed mechanically, i.e. the first rotor. The ππΌπ is not a measurable quantity due to the very high temperature and the need to implement multiple gauge wells between the first stator and the first rotor array, with penalization of the fluid dynamics of the machine. It can be obtained with a mass and energy balance, but requires that all the cooling flow rates be known. 4T 4 stator 5 3T 5T Turbine (Rotor) Inlet Temperature 4 T5T=TIT 5S 5 Fig. 9.17 - Mass and energy balance astride the first stator array and definition of TIT • ππΌππΌππ : is the temperature resulting from a mass and energy balance that considers all the cooling flows, regardless of whether they are chargeable or non-chargeable. Obviously, ππΌππΌππ < ππΌπ, ππΌππΌππ ~ππΌπ − 200. It is not representative of a real temperature, but is useful when testing the turbine because it allows the operating conditions to be defined from a balance of measurable quantities. In particular, knowing the mechanical power of the turbine, it is possible to apply the relation: ππΌππΌππ πΜππ π΅ = (πΜπππ + πΜππ’ππ ) ∫ ππ (π)ππ (9.21) πππ that corresponds to a fictitious adiabatic expansion that starting from ππΌππΌππ arrives at the turbine outlet temperature πππ of the entire exhaust gas flow. The πππ is instead measured with many measurements along the circular crown at the outlet. 118 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 This parameter is important for calculating the machine performance during testing. Machine data such as compression ratio, mass of aspirated air and πππ are usually available. While the data regarding πΆππ and ππΌπ are never available because it is covered by industrial secrecy obligations. During testing, the supplier might misrepresent the performance test in his favor by reducing the cooling flow rates, raising the ππΌπ and therefore the efficiency, but showing the metal surfaces at excessive temperatures that if maintained would damage the reliability of the turbine. This is why the ππΌππΌππ is used as the variable for checking the operating conditions during testing. Cooled turbine 5 6 Adiabatic turbine 4T 7 3T 5T a 6 “chargeable” b 7 TITiso 4 iso 3 a) cooled expansion b) non-cooled expansion Fig. 9.18 - Simplified representation of the expansion and introduction of TITISO This concludes with a few observations on the influence of the cooling flows on the calculation of the irreversibilities in the cycle: • The flows circulating in the plant vary significantly both in the compression phase and in the expansion phase. The difference between the air flow rates depends on the materials used and the cooling techniques, in addition to the compression ratio of the cycle and the maximum temperature of the gases. • The expansion in the portion of cooled turbine (usually the first 4-5 stages) takes place at decreasing entropy, if reference is made to the main flow that expands. Therefore, the equation π€ = ββ does not apply, which is applicable only to adiabatic processes. • In addition to the fluid dynamic irreversibilities common to all the expansion processes in real machines (identified by the polytropic efficiency due to an adiabatic expansion), many other irreversibilities take place during the cooled expansion of a gas turbine, which are linked to: (i) the heat transfer between the main flow and cooling flows, (ii) the mixing of the main flow and cooling flows (which are introduced into the main flow at temperatures lower than the flow itself for heat transfer purposes), and (iii) the fluid dynamic losses that occur due to the introduction of the cooling flows (that have lower speeds) into the main current (at high speed). 119 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 The image in Fig. 9.19 illustrates a refined modeling of cooled expansion in a modern gas turbine to emphasize the complexity of the treatment necessary for proper simulation of these machines. The expansion refers to the turbine OL-AC and takes place in four stages: the first two cooled, the last two considered adiabatic. Every line connects the static conditions to the inlet and outlet of each stage, and is divided considering first the heat transfer effect and then the mixing with the cooling flow. The broken lines describe transformations that take place as the flow rate (cooled arrays) changes, while the continuous lines refer to the adiabatic expansion. The last line with increasing pressure represents the diffuser. For greater detail, the expansion of every single array is divided into another 10 intervals. Fig. 9.19 - Detailed representation of the numerical model that can be used to calculate a cooled expansion Note: consult the two scientific articles for an in-depth study of the topics discussed in this chapter • Predicting the ultimate performance of advanced power cycles based on very high temperature gas turbine engines, P. Chiesa, S. Consonni, G. Lozza, E. Macchi • A thermodynamic analysis of different options to break 60% electric efficiency in combined cycle power plants, P. Chiesa, E. Macchi or the book “Turbine a gas e cicli combinati” by G. Lozza, Progetto Leonardo, Bologna. 9.5 INTEGRATING COOLED GAS TURBINES IN COMBINED CYCLES The high temperature of the exhaust gases released by a gas turbine and the large thermal powers available make the execution of a steam recovery cycle interesting. The plant takes the name of combined cycle and is made up of a gas turbine and a steam recovery cycle. Having already separately 120 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 considered all the components that make up a gas cycle and a vapor cycle, now focus should be placed on the element of junction between the two plants, i.e. the HRSG (Heat Recovery Steam Generator), where the water is preheated, evaporated and superheated in one or more pressure levels. The component will be discussed from the viewpoint of the second law analysis, while the student is referred to other texts for an exhaustive description of these plants and their operation in only an electrical or cogenerative arrangement. Fig. 9.20 - Evolution of the efficiency of power production plants. Note the impact of the improved performance of the gas cycles compared to the vapor cycles in improving the efficiency of combined cycles. The major difficulty in a combined cycle plant is matching the cooling curve of the flue gas well with the heating curve of the working fluid. In view of the size of these plants and the temperature levels, water is used. Since it is a single molecule fluid, it is not the most advisable for exploiting a source with variable temperature like the exhaust gases. The use of organic fluids is not advisable in these application since their use is restricted by their thermal stability to temperatures of about 400°C. As already seen previously, the use of a single evaporation level does not allow the available heat to be fully exploited and involves high efficiency losses due to a process of introducing irreversible heat. This is mainly due to the greater weight of the evaporation section compared to the economizer or the superheating section. Generally, the process works at fixed βπππ and this condition usually occurs at the point of saturated liquid. An exception is the case of post-combustion, in which the πππ is so high that the pinch-point is moved to the economizer inlet. Bear in mind that the recuperative preheating by bleeds from the turbine for recovery plants is never expedient because it might on the one hand lead to a reduction in the flow rate in the turbine and hence a lower produced power, and on the other an incomplete exploitation of the thermal power available. When using a single-level cycle, there are mainly two causes of irreversibility: • Heat transfer: it results very high even when greatly reducing the βπππ the βππππ ; • Mixing the flue gas with the environment. Both of these losses can be limited by using recovery cycles with multiple evaporation levels. It is common to find HRSGs with three evaporation levels plus superheating in conventional plants. 121 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 The interaction between vapor cycle and gas cycle depends on the architecture of the plant. The three main ones are called Frame F, Frame G and Frame H. Three levels are used in Frame F. That with the lower temperature has a pressure adequate for achieving deaeration, so the deaerator is built into the top cylindrical casing. The arrangement of the heat transfer banks is selected to best follow the cooling curve of the flue gas. For example, the lower pressure SH and part of the high pressure ECO are placed in parallel with the medium pressure ECO. The fuel is preheated by drawing a water flow from the low pressure ECO. Frame G is similar to Frame F, but the fuel is preheated with hot water drawn from the medium pressure cylindrical casing at about 200°C. In both configurations there is little interaction between vapor cycle and gas turbine, and the blades are cooled by open cycle compressed air. Fig. 9.21 - Configuration with Frame F and Frame G for combined cycles, and OL-AC cooling diagram for the gas turbine blades GE proposed integrated cycles in the '90s called Frame H, in which the blades were cooled with vapor superheated in closed cycle with the aim of reducing the convective heat transfer losses, limiting the mixing losses and recovering more thermal energy. Using vapor to cool the rotor blades however introduced enormous difficulties in limiting vapor leakage along the turbine shaft. The Mitsubishi model ATS was placed on the same line. Its vapor cooling was restricted to just the stators in order to prevent problems linked to vapor leakage. These latter solutions proved unsuccessful owing because the current energy market demands that these plants operate in intermediate load (3000-4000h and 50100 start-ups a year), following that does not go well with the need to have long start-up periods for the vapor portion and less flexible plant control. 122 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 Fig. 9.22 - Configuration with Frame H for combined cycles, and ATS and CL-SC cooling diagram for the gas turbine blades 9.6 EFFECTS OF REAL GAS IN THE GAS CYCLES For this purpose, a recuperated closed gas cycle plant that can operate with a fluid affected by real gas effects since it is near the saturation curve is considered. One example might be a carbon dioxide closed cycle that has a critical temperature of about 37°C, thus compatible with cooling by ambient air or water. In some transformations the gas can be considered ideal because it is at high temperature (expansion 3-4) or at low pressure (expanded side 4-5-1). In the compression phase (1-2) and in the compressed side (2-6), as the fluid approaches the saturation curve it certainly shows effects of real gas, that on the one hand lead to decreasing its specific volume and on the other to increasing the specific heat. This is shown in Fig. 9.23. It can be considered that as the real gas effects grow, considering cycles with π½ and ππΌπ being equal: • the compression phase requires less power and the final temperature is lower since π decreases as the specific heat increases. The production of entropy and the associated loss of efficiency decreases since it is linked to the medium compressibility factor of the fluid, which is certainly lower than one unit. • If the cycle is not recuperated, the lower end of compression temperature results in a greater incoming thermal power. As it is at the denominator of the terms Δπ, it allows the irreversibility losses linked to the expansion and transfer of heat to the environment to be reduced. An exception to the rule is the loss of introduction of heat since a lower end of compression temperature involves increased power entering the cycle on the one hand, but also decreased mean log temperature on the other. The two effects are opposite, but the result is that on the whole improvement in the efficiency is recorded as the effects of real gas increases. • Instead, if the cycle is recuperated, the greater the mean pressure of the cycle and the real gas effect, the greater is the heat capacity of the cold side with respect to that of the hot side. If βπππ being equal is considered, it will be at the cold end of the heat exchanger, and therefore temperature π5 will be lower in the case of real gas effects. The heat transfer βππππ increases because the βπ increases on the hot end of the exchanger, but the recuperated heat decreases. 123 of 124 Second-law analysis of power cycles - Energy Conversion A – V7.0 The increased thermal power introduced has a beneficial effect on both the losses of compression (already positively influenced by a lower compressibility factor), of expansion and of heat transfer to the environment (further limited by a lower π5 ) Also in this case, an improvement in efficiency is noted as the real gas effects increase. 3 3 4 4 6 6 2 2 5 1 5 1 CO2 Fig. 9.23 - Variation of the carbon dioxide closed gas cycle as the saturation curve is approached due to real gas effects. 9.7 OTHER NUMERICAL EXAMPLES The student has to supplement all the qualitative considerations put forth in this chapter with the numerical results obtained during Exercises 3 and 5, dedicated to the second law analysis for combined cycles at multiple evaporation levels and gas turbines in different configurations. 124 of 124