Peter Ireland Hilary Term 2017

2nd Law of Thermodynamics and its Applications

Part II - Applications Year 2

1 INTRODUCTION
1.1 Overview
This course is concerned with the efficient use of energy, be it the efficient
conversion of an energy source to mechanical power, or the use of mechanical
power to provide services such as refrigeration or compressed air. The effective
use of energy is important for economic and environmental reasons, so this
course examines the ways of maximising the conversion of heat to work, and of
minimising the work input to power absorbing processes.

1.2 General Reading

These notes are intended to provide you with a self-contained course on Energy
Sources and Cycles. However, it is often useful to have an alternative
presentation of a topic, so at the start of each section appropriate references are
given to the following texts:

R&M GFC Rogers & YR Mayhew Engineering Thermodynamics,

Longman, 4th Ed. 1993

E&McC TD Eastop & A McConkey Applied Thermodynamics, Longman,

5th Ed. 1993.

Numerous other engineering thermodynamics books are also useful, but a book
that provides a particularly thorough treatment of thermodynamic cycles is:

1
Hay R W Haywood Analysis of Engineering Cycles, Pergamon Press,
2nd Ed. 1975

These texts uses a different sign convention for work:

R&M (4th Ed only) E&McC (5th Ed only) treat work like heat, so it is
positive when flowing into a system.

Most engineering thermodynamics texts, and your year 1 course, consider

work to be positive when flowing from a system.

Haywood treats all work terms as positive, and relies on knowing or

anticipating whether the device being analysed is work producing or work
absorbing; this then determines the way in which the 1st Law equation is
written.
1.3 Acknowledgement
These notes are closely based on those written by Professor Richard Stone.

2 REFRIGERATION SYSTEMS
2.1 Introduction
Refrigeration systems also include heat pumps; the only difference being the
way in which the systems are used:

Refrigeration removal of heat from a low temperature to a higher

temperature.
Heat Pumps delivery of heat to a higher temperature from a low
temperature.

2
In both applications heat is being taken from a low temperature source and being
delivered to a high temperature sink. Gas liquefaction is a special class of
refrigeration in which the temperatures are very low, and often the gas being
liquefied forms part of the working fluid.

A refrigeration effect can be obtained by many different techniques, including:

Peltier effect (R&M p489),

Joule-Thomson effect (R&M p107, Hay p241),
Absorption refrigerator (R&M p281, E&McM p511, Hay p225),
Ranque Hilsch tubes and
Reversed power systems - vapour cycles (R&M Ch13, E&McM Ch14,
Hay Ch5).
- gas cycles, either Joule or Stirling cycles

Only reversed vapour cycles will be considered here, since they are very widely
used. However, you will be able to analyse reversed Joule cycles using the
material presented later in Section 5 on Gas Turbines.
For more detailed information you can consult:
W B Gosney Principles of Refrigeration, CUP 1982
R D Heap Heat Pumps, Spon, 2nd Ed 1983.

2.2 Ideal Reversed Vapour Power Cycle -

The Reversed Carnot Refrigeration Cycle is by definition the system for
which the work input will be a minimum, and in a vapour cycle the heat exchange
at constant temperature is readily achieved by isobaric operation within the
saturation region.

A Reversed Carnot refrigeration cycle is shown in Figure 2.1, on both the T-s

3
planes (Figure 2.1a) and p-h planes (Figure 2.1b), and the layout of a physical
system is shown in Figure 2.1c. In all of the figures:
a) the isobars are shown by chain-dashed lines,
b) the isotherms are shown as dashed lines, and
c) the isentropes are shown by dotted lines.

Fig. 2.1a Temperature-Entropy diagrams for reversed power cycles acting

as a refrigerator and a heat pump

Fig. 2.1b&c Pressure-Enthalpy diagram for reversed power cycle, and a

plant diagram

4
The cycle comprises of 4 processes:

12 reversible adiabatic (and thus isentropic) compression

23 isothermal heat rejection
34 reversible adiabatic (and thus isentropic) expansion
41 isothermal heat reception.

In Figure 2.1, it can be seen that the cycle has been arranged so as to occur
wholly within the saturation region, this means that the isothermal heat
transfer processes (23 and 41) are also isobaric. Furthermore, the heat
rejection process has been contrived such that state 3 is saturated liquid,
and the heat reception process ends (state 1), so that after the isentropic
compression (12) the state 2 is saturated vapour.

The two isobars are shown on Figure 2.1a, and it should be noted that:

a) in the subcooled liquid region, the divergence of the isobars from the
saturated liquid line has been exaggerated for the sake of clarity.

b) in the superheated vapour region the isobars have a positive slope that
increases with temperature (the slope would be proportional to the
absolute temperature for a perfect gas), and the isobars thus diverge.

Figure 2.1b has been included, since in the lab you will be making use of a
pressure-enthalpy chart. The interception of the isotherms and isobars
enables the thermodynamic state to be easily identified, and the enthalpy can
then be read directly from the horizontal axis; this facilitates rapid calculation of

5
the thermodynamic processes. The isotherm added to Figure 2.1b is a vertical
line in the subcooled liquid region, since for the pressures being considered
the enthalpy is independent of pressure. For a vapour, the enthalpy increases
when the pressure is reduced and the vapour temperature is constant; the
isotherm takes the curved form shown in Figure 2.1b.

The isotherms become vertical in the superheated region at low pressures,

since the behaviour of the vapour approaches that of a Perfect Gas, for which
the enthalpy is only a function of temperature. The pressure-enthalpy chart uses
a logarithmic pressure scale; this leads to a more uniform spacing of the
isotherms in the saturation region.

The reversed power vapour cycle can be analysed by using a few simplifying
assumptions:

a) The only pressure changes occur in the expander and compressor,

b) there is no extraneous heat transfer, and
c) changes in the Potential Energy and Kinetic Energy can be neglected.

The Steady Flow Energy Equation (SFEE) can then be written in simplified
(and specific i.e. per unit mass) form as:

h1 + q = h2 + w (2.1
The SFEE can be applied in turn to the four components in the cycle:

Compressor: win = h2 - h1

Condenser: qout = h3 - h2
(2.2

6
 Expander: wout = h3 - h4s

Evaporator: qin = h1 - h4s

It should be noted that the changes in enthalpy will correspond to the horizontal
distances on the p-h chart.

The definition of entropy (dq = Tds) enables the heat flows to also be defined as:

Condenser: qout = h3 - h2 = T2,3Δs2,3

(2.3
Evaporator: qin = h1 - h4s = T4,1Δs4,1

Since we have isentropic expansion and compression processes, then

Δs2,3 = Δs4,1

The 1st Law is always valid, so the difference between the heat flow out and heat

flow in will correspond to the net work input (Wnet):

wnet = qout - qin = (T2,3 - T4,1)Δs2,3 (2.4

The net work input of course equates to the area within the cycle.

The Performance Parameter for Reversed Power cycles depends on

whether the cycle is being used as:

7
a) a Refrigerator, for which Qin is of importance, or

b) a Heat Pump, for which Qout is of importance.

In both cases the Performance Parameter is known as a Coefficient of

Performance (CoP):

Refrigerator: CoPRef = Qin / Wnet = Qin / (Qout - Qin)

(2.5
Heat Pump: CoPHP = Qout / Wnet = Qout / (Qout - Qin)

The Coefficient of Performance of the Heat Pump is always greater than unity,
since:
CoPHP = CoPRef + 1 (2.6

Figure 2.1a illustrates the difference between a refrigerator and a heat pump. In
the refrigerator, heat extracted from the medium being cooled (at a temperature
slightly above that of the evaporator) is 'pumped' to a higher temperature, which
is slightly above the environment temperature (To), to which the heat is being
transferred. Heat transfer can only occur with a temperature difference (as
shown in both parts of Fig 2.1a), but as this is an irreversibility that can be
considered as external to the reversed power cycle, it need not concern us here.

In the heat pump, the evaporator temperature is slightly below the environment
temperature (the source of the heat), and the condenser temperature is slightly
higher than the temperature of the medium that is being heated. It should be
noted that the heat pump cycle is operating closer to the critical point, and this
has several implications:

8
a) the pressures are significantly higher, and
b) the work and heat flows per unit mass are reduced.

It will be seen later that when compression occurs in the vapour region (as
opposed to within the saturation region), then there is an adverse effect on the
CoP. This means that the refrigerant needs to be selected in accordance with
the temperature ranges to be encountered, and in general, heat pumps use a
different refrigerant to refrigerators.

2.3 Simple Ideal Practical Refrigeration Cycle

In practice, it would be very difficult to ensure that:

a) the liquid entering the expander was saturated, and

b) the state of the mixture entering the compressor was such that it would be
saturated vapour at the end of compression.

With imperfect expanders and compressors, it would clearly be undesirable to

expand vapour from a high pressure to a low pressure, since the irreversibilities
would mean that more work would be input to the compressor than produced by
the expander. The usual practice is to make the refrigerant slightly subcooled
at exit from the condenser (as shown in Figure 2.2. The low specific volume of
the liquid and very wet vapour (i.e. they are dense) means that little work is
produced by the expander. Furthermore, it is difficult to design an expander for
liquids/wet vapours; instead, a simple expansion throttle is used. The
irreversibilities associated with the throttle are in fact reduced by

9
subcooling the refrigerant before expansion, and are really only significant with
low temperature systems, such as used for gas liquefaction.

Fig. 2.2 Temperature-Entropy and Pressure-Enthalpy diagrams for a

practical reversed power cycle

It would be undesirable for liquid droplets to enter most compressors, so the

vapour entering the compressor is superheated slightly. In most refrigerating
systems the small flow rates mean that reciprocating compressors are used,
and liquid droplets would tend to transfer lubricant from the compressor to the
condenser.

10
Fig. 2.3 Plant layout for a practical reversed power cycle, and a
thermostatically controlled expansion valve

Figure 2.3 shows the plant layout, and a detail of the thermostatically
controlled expansion valve. The thermostatically controlled expansion
valve is designed to give a constant level of vapour superheat at entry to the
compressor. The temperature sensing bulb contains a fluid, which expands as
the temperature rises, thereby tending to open the valve, so as to: increase the
refrigerant mass flow rate and increase the evaporator pressure. In small
refrigeration units (say below a heat input of 10kW), then a capillary tube is often
used for controlling the expansion process, since for a given pressure difference
the mass flow rate of liquid will be much higher than that of vapour.

The refrigeration system can be analysed by making the same assumptions as

before. As a slight difference, we will introduce a mass flow rate of refrigerant
(mr).

11
 Heat Flow Out: Qout = mr (h2 - h3)

Heat Flow In: Qin = mr (h1 - h4) (2.7

Work In: W = mr (h2 - h1)

Coefficient of Performance: (CoP)Ref = Qin / W = (h1 - h4)/(h2 - h1)

State 4 is within the saturation region, so pressure and temperature are not
independent. However, since we are:

 neglecting changes in Kinetic Energy & Potential Energy,

 assuming that there is no extraneous heat transfer, and
 clearly there is no work output from the throttle, then:

h3 = h4 (2.8

The throttle is thus isenthalpic, and there will be an increase in entropy due to
the irreversibility, ΔS. The reduction in refrigeration effect is the shaded area in
Figure 2.2, corresponding to:

Tsat,4 ΔS (2.9

2.4 Practical Refrigeration Cycles

The real compressor will not be reversible, so if it is adiabatic, then there will be
an increase in entropy for process 12; for purposes of comparison the
isentropic compression process will be shown as 12s. The T-s plot in Figure

12
2.4 shows the way in which the isobars diverge. This means that there will be an
increase in the work input, since to a first order the enthalpy of the vapour is
proportional to its temperature. As there is no increase in the refrigeration effect,
then the Refrigerator Coefficient of Performance deteriorates.

Fig. 2.4 Temperature-Entropy and Pressure-Enthalpy diagrams for a practical

reversed power cycle

Although the heat output is increased (by an amount equal to the extra work
input), this will lead to a reduction in the Heat Pump Coefficient of Performance,
since both the numerator and denominator have increased by the same amount.
This can also be seen by inspection of equation 2.6:

CoPHP = CoPRef + 1 (2.6

The increase in work input is also shown on the P-h plot in Figure 2.4, since the
isentropes diverge in the vapour region. The increase in work input, leads to the
following definition of compressor isentropic efficiency (ηc):

13
 ηc = isentropic work input = (h2s - h1)/(h2 - h1) (2.10
actual work input

Note the use of a broken line in Figure 2.4 for the irreversible process, and a
chain-dashed line for the hypothetical process.

WARNING In reality, refrigeration compressors are not usually

adiabatic. There will be heat transfer to the
environment, so the enthalpy change in the refrigerant
will be less than the specific work input to the
compressor. The heat transfer will lead to a reduction in
entropy of the refrigerant, and this can be greater than
the entropy creation due to irreversibility within the
compressor. The overall reduction in entropy will lead to
a compressor isentropic efficiency of above 100%! Text
books, tutorials and exam questions invariably assume
adiabatic compressors (without necessarily stating the
assumption).

It must also be appreciated that if the superheat at entry to the compressor is

allowed to increase, then this too will lead to an increase in the work input and a
decrease in the Coefficients of Performance. This is illustrated in Figure 2.5,
for both the T-s and P-h plots. This is due to the divergence of the isobars on
the T-s plot, implying that the increase in heat input is smaller than the increase
in heat rejection, leading to an increase in the work input, and a fall in the CoPs.
The divergence of the isentropes on the P-h plot shows the same result.

14
Fig. 2.5 Temperature-Entropy and Pressure-Enthalpy diagrams that
show the increase in work input as the superheat at entry to the
compressor rises

2.5 Improvements to the Simple Refrigeration Cycle

2.5.1 Separator
It is clearly desirable to minimise the superheat at entry to the compressor, and
this is one of the reasons for using a thermostatically controlled expansion valve
(Figure 2.3). Another option is to use a separator or 'flash chamber', as shown
in Figure 2.6. Since liquid and vapour co-exist in the separator, then both the
liquid and vapour will be saturated, and the vapour entering the compressor
will be saturated. If the refrigerant leaving the evaporator is superheated, then
the direct contact between the vapour bubbles and the liquid in the separator will
cause more liquid to vaporise, until the vapour is saturated.

15
Fig. 2.6 Plant layout and Temperature-Entropy diagram for a simple
refrigeration cycle with a separator

By careful choice of a control volume, we do not need to concern ourselves with

the details of the flows and state of the evaporator, and the heat input to the
cycle is still given by equation 2.7.

Heat Flow In: Qin = mr (h1 - h4) (2.11

The T-s diagram in Figure 2.6 can also be used to illustrate why the vapour
superheating that occurs within the compressor leads to a reduction in the CoP.
The triangle a2b is known as the superheat horn, and its area represents the
extra work input due to the vapour becoming superheated. If the saturation
curve was such that the compression process had remained within the saturation
region, the compression process would have been 1a, and the condensation
process would have been a(b)3. The heat input would remain the same, but
there would be a reduced work input, and thus a higher CoP.

16
2.6 Refrigerant Properties (E&McC)
Desirable refrigerant properties might include:
a) Cheap
b) Non-Toxic
c) Non-Flammable
d) Saturation pressure above atmospheric at the lowest cycle temperatures -
so that air cannot leak in
e) Moderate saturation pressure at the highest cycle temperatures - so that
the mechanical design is simple
f) High enthalpy of vaporisation - so that mass flow rates are small
g) High vapour density at entry to the compressor - so that the compressor is
compact
h) High vapour specific heat capacity (Figure 2.9a) so that on the T-s diagram
the isobars are close to horizontal, and the superheat horn is a small area.
i) Low liquid specific heat capacity, so as to facilitate subcooling in
regenerative cycles
j) Steep saturation lines (Figure 2.9a) in the region of operation, so that the
practical cycle will make a close approximation to the Reversed Carnot
Cycle.
k) The freezing point of the refrigerant needs to be below the operating
temperature of the evaporator.
l) Non-Corrosive) Compatible with any lubricant required in the compressor

Fig. 2.9 Temperature-Entropy diagrams for a good and poor refrigerant

17
Figure 2.9a shows a good refrigerant, with high vapour heat capacity and steep
saturation lines. The operation away from the Critical Point suggests that
excessive pressures are avoided. Figure 2.9b shows the converse to Figure
2.9a, with a significant area in the 'superheat horn'.

Many substances have been used as refrigerants, including:

water, air, carbon dioxide, ammonia, hydrocarbons (eg propane),

methyl chloride, and chlorofluorocarbons (CFCs).

Water would be very good if it did not freeze, and had a low vapour pressure. Air
has the disadvantage of no phase change. Carbon dioxide results in high
operating pressures, and leaks can lead to suffocation. Ammonia too leads to
high pressures, is flammable and an irritant.

Refrigerant tables are frequently in a more compact form (eg Table 2.1) than
steam tables, and this means that more interpolation is required. The data are a
combination of Saturation and Superheat values on a single line. Each line
corresponds to a particular pressure, the increments of which are chosen to
given uniform steps in the saturation temperature.

The superheat data do not refer to absolute temperatures, but are for fixed
superheat temperature increments relative to the saturation temperature. In
Table 2.1 the superheat increment is 15K, and as an example, the values of
vapour enthalpy that have been underlined all refer to 25C. The three values of
enthalpy are not the same, since the enthalpy of a real gas depends on its
pressure, as well as its temperature.

18
Table 2.1 Thermodynamic Properties of Refrigerant 134a (HLT)

19
In refrigeration cycle analysis, the calculation that usually causes most difficulty
is the compression process, so this will be illustrated by an example.

EXAMPLE
Vapour is leaving a separator at a pressure of 2.01 bar, and being
compressed to a pressure of 8.82 bar in an adiabatic compressor, with an
isentropic efficiency of 85%. Calculate the specific work input to the
compressor.

The vapour entering the compressor will be saturated vapour, at a temperature

of -10C, since this is the purpose of the separator. The compression process is
shown on the T-s diagram in Figure 2.10a, with the broken line representing the
actual compression process.

Fig. 2.10 (a) Temperature-Entropy diagram for a real compression process

(b) interpolation between enthalpy and entropy in the
superheated vapour region

By applying the SFEE (neglecting changes in Kinetic and Potential Energy), the
specific work input (w12) is given by:

w12 = h2 - h1

20
which has to be evaluated from the work input to an isentropic compression
process. From equation 2.10

ηc = isentropic work input = (h2s - h1) / (h2 - h1) (2.10

actual work input

and, w12 = (h2s - h1) / ηc

The enthalpy h1 is the enthalpy of saturated vapour (hg) at -10C, and is found
directly from the tables to be 292.7 kJ/kg. The problem now reduces to finding
h2s. By definition s2s = s1, and s1 is found directly from the tables (as the

entropy of saturated vapour (sg) at -10C) to be 1.733 kJ/kgK.

We now assume a linear relation between the enthalpy and entropy of the
vapour along the 8.87 bar isobar. Inspection of the tabulated data at 8.87 bar
indicates that state 2s lies between the saturation values and those with 15K
superheat. The linear interpolation is illustrated by Figure 2.10b, from which:
h2s - hg = hSat+15 - hg
s2s - sg sSat+15 - sg
Rearrangement gives: h2s = hg + (hSat+15 - hg)  (s2s - sg) / (sSat+15 - sg)
Substitution of numerical values from the 8.87 bar isobar gives:
h2s = 317.2 + (333.2 – 317.2)  (1.733 - 1.7128) / (1.7634 - 1.7128)
= 323.6kJ/kg

and w12 = (h2s - h1) / ηc = (323.6 – 292.7) / 0.85 = 36.35 kJ/kg

The same result is obtained by using the superheat as a temporary variable, and
dividing the linear interpolation into two parts:

21
a) Interpolation between temperature and entropy to find the superheat, and
b) Interpolation between temperature and enthalpy, using the superheat to
find the enthalpy.

Such a calculation in the current example leads to a superheat of 6.1K. When

the enthalpy and superheat are evaluated by NIST software, then for the
superheat of 6.1 K, the enthalpy is 323.80 kJ/kg.

Example 2.1 Refrigeration Cycle with a Separator.

A requirement for a 0.2 kg/s flow of air at -30C is to be met by a vapour

compression refrigeration cycle using R-134a, in which the only pressure
changes occur in the compressor and throttle. The ambient temperature is
20C, and the saturation temperature in the condenser is 30C whilst the
saturation temperature in the evaporator is -35C. The refrigerant is
subcooled by 5K at entry to the throttle valve, and after the throttle it enters
a separator. The vapour leaves the compressor at a temperature of 45C;
the heat transferred from the compressor to its surroundings is equivalent
to 10kJ/kg refrigerant.
(a) Sketch the processes on T-s and P-h plots.
(b) Calculate the mass flow rate of the refrigerant, the power
requirement of the compressor, and the Coefficient of Performance
of the cycle.
(c) Identify the sources of irreversibility.
a)

22
 2
2s

b)
P
Qin = (mcp)air ΔTair = mR(h1-h4)
h1 = (hg) -35C = 277.2 kJ/kg.K
h4 = h3 ~ (hf) 25C = 134.6 kJ/kg.K
mR = (mcp)air ΔTair / (h1-h4)
= 0.2 x 1.01 x 50/(277.2-134.6)
mR = 0.071 g/s

SFEE: In = Out
w12 + h1 = q12 + h2 h
h2 = (h7.70bar)Tsat + 15 K = 330.5
kJ/kg.K
w12 = q12 + h2 - h1 = 10 + 330.5 – 277.2 = 64.1 kJ/kg.K
W12 = mR x w12 = 0.071 x 64.1 = 4.55 kW (work input)
Qin = (mcp)air ΔTair = 0.2 x 1.01 x 50 = 10.1 kW
CoP = Qin / W12 = 10.1 / 4.55 = 2.20

c) Expansion process, Compression Process, Heat Exchange with a finite

temperature difference

23
To be able to attempt past exam questions you will need data for R12. The data
below are from the previous edition of HLT p67.

24
 3 STEAM CYCLES (Hay Ch1&7)
3.1.1 Introduction (Hay Ch1)

The simple steam plant shown in Figure 3.1a consists of: a turbine, condenser,
feed pump and boiler. The following assumptions are made to facilitate the
analysis:

a) the processes in the condenser and boiler will be assumed to occur

reversibly at constant pressure, and

b) the expansion and compression processes will be assumed to be both

adiabatic and reversible (and thus isentropic).

Fig 3.1a Vapour power cycle,

25
Fig 3.1b&c Temperature-Entropy and Enthalpy-Entropy plots showing
the Rankine Cycle

A cycle (known as the Rankine Cycle) is to compress saturated liquid in the

feed pump, and to expand superheated steam in the turbine. This is illustrated
by the T-s diagram in Figure 3.1b, and the h-s diagram in Figure 3.1c.

The Rankine Cycle comprises the following processes:

12 isentropic expansion in the turbine.

23 isobaric condensation in a condenser, with the condensate leaving as
saturated liquid.
34 isentropic compression in the feed pump.
41 isobaric heat input in the boiler, with -

26
 4a heating of liquid water to its saturation temperature,
ab evaporation of the saturated water to saturated vapour, and
bc heating of saturated vapour to become superheated vapour.

The h-s diagram emphasises that a significant energy input occurs at constant
temperature. The advantage of the h-s plot is that accurate charts are published
that encompass the region of turbine expansion, and these permit the rapid
graphical calculation of turbine performance.

The thermal efficiency of any cycle is given by equation 1.3:

η = wnet / qin

By applying the Steady Flow Energy Equation (SFEE), and making the usual
assumptions about negligible changes in Kinetic Energy and Potential Energy,
then for the Rankine Cycle:

ηRankine = wnet / qin = (wt -wp) / qin = {(h1 - h2) - (h4 - h3)} / (h1 - h4) (3.1

The enthalpy at state 1 can be found from the steam tables or chart, the enthalpy
at state 3 can be found from the steam tables, and the enthalpy at state 2 can be
calculated. However, the enthalpy at state 4 (subcooled water) is not
tabulated, so the feed pump work (wp) has to be evaluated by:

4
wp = vdp and, h4 = h3 + wp (3.2
3

Since water is almost incompressible, the specific volume can be taken outside
the integral, and:

27
 4 4
wp = vdp  vdp = v [p4 -p1] (3.3
3 3

In which v is now a mean specific volume

In low rated steam plant the feed pump work can often be neglected, but in
steam plant that operates with temperatures and pressures that are chosen to
give the highest efficiency (say, 250 bar and 600C at turbine inlet), then the feed
pump work is in the region of 5% of the turbine output.

Some terms that will be encountered with steam cycles are:

Work Ratio - Net Work Out / Turbine Work.

Quality - The dryness (or wetness) of steam, low quality meaning

comparatively wet, and high quality meaning comparatively dry.

Heat Rate - A reciprocal of efficiency, and analogous to specific fuel

consumption in Internal Combustion Engines, the units might be
MJ(fuel)/kWh(work) or kg(fuel)/kWh(work) - 1kWh  3.6MJ

Steam Data HLT

Saturated Data (pp 54-55

and 56-65)
Superheated Data (pp 66-70)
Supercritical Data (pp 71)

28
Mollier Diagram

29
Fig 3.2 Temperature-Entropy and Enthalpy-Entropy plots showing
irreversible turbine expansion
The real turbine will not be reversible, but it can be treated as adiabatic since it
will be insulated, and there will be an increase in entropy for process 12
(Δs12); for purposes of comparison the isentropic compression process will be
shown as 34s - this is shown on the T-s plot in Figure 3.2.

The decrease in work output, leads to the following definition of turbine

isentropic efficiency (ηt):

ηt = actual work output = (h1 - h2)/(h1 - h2s) (3.7

isentropic work output

Note the use of a broken line in Figure 3.2 for the irreversible process, and a
chain-dashed line for the hypothetical process.

30
Example 3.1

a) Calculate the thermal efficiency, work ratio and steam quality at turbine exit
for a steam cycle with water as the working fluid, a boiler pressure of 100 bar
and turbine inlet temperature of 5500C, and a condenser pressure 0.2 bar. The
turbine is adiabatic with an isentropic efficiency of 82%. Include the feed pump
term.

b) If the gas temperature into the boiler is 600°C , the outlet temperature from
the boiler is 170°C, and the combustion gases have a specific heat capacity of
1.1 kJ/kg_K, then calculate the mass flow rate of the combustion gases (per kg
of H2O) and the temperature difference at the
pinch point.

31
 h1 = ½ (h600 + h500)100 bar = ½ (3622.7 + 3374.6) = 3498.7 kJ/kg
s1 = ½ (s600 + s500)100 bar = ½ (6.9013 + 6.5994) = 6.5704 kJ/kg.K
s1 = s2s = (sf + x2s sfg ) 0.2 bar = (0.8321 + x2s 7.0773)

dryness, x2s = (s1- sf) / sfg = (6.5704 – 0.8321) / 7.0773 = 0.811

h2s = (hf + x2s hfg ) 0.2 bar = (251.5 + 0.811 x 2358.4) = 2163.7 kJ/kg

w1-2s = (h1 - h2s ) = 3498.7 – 2163.7 = 1335.0 kJ/kg

w1-2 = w1-2s x ηs = 1335.0 x 0.82 = 1095 kJ/kg

w1-2 = (h1 - h2) ; h2 = h1 - w1-2 = 3498.7 – 1095 = 2404 kJ/kg

h2 = (hf + x2 hfg ) 0.2 bar = (251.5 + x2 x 2358.4) = 2404 kJ/kg

x2= (2404 - 251.5) / 2358.4 = 0.912, dryness at turbine exit

w34 = ∫vdp ≈ v ∫dp = vf, 0.2 bar Δp = 1.0172X10-3 x (100 – 0.2) x 105 = 10.2 kJ/kg

h4 = h3 + w34 = hf, 0.2bar + 10.2 = 251.5 + 10.2 = 261.7 kJ/kg

η = wnet / qin = (w12 - w34 ) / (h1 - h4) = (1095 – 10.2) / (3498.7 – 261.7) = 0.335
Work Ratio = Wnet/W12 = (1095 – 10.2) / 1095 = 0.991

32
 b)

Temperature/Enthalpy Plot

700
600
500
T (deg C)

400
300
200
100
0
0 500 1000 1500 2000 2500 3000 3500 4000
h (kJ/kg)

T1 = 550°C h1 = 3498.7 kJ/kg Gas Inlet Temperature, T5 = 600°C

Tsat = 311°C hg = 2727.7 kJ/kg hf = 1408.1 kJ/kg

h4 = 261.7 kJ/kg Gas Outlet Temperature, T6 = 170°C

For interest
T4 = 62.16+(64.08-62.16) * ( 261.7- 260.1) /( 268.2 - 260.1) = 62.4°C

33
3.2 Means of Improving Steam Cycle Efficiency
3.2.1 Steam Condition (Hay Ch1)

The steam condition refers to the steam pressure and temperature at entry to
the turbine; the effect of each of these variables is shown in Figure 3.3, in which
the change to the cycle is shown by a broken line. Raising the turbine inlet
temperature leads to:

a) a raising of the Mean Temperature of Heat Reception Ti , and thus

an increase in the cycle efficiency, and

b) a reduction in the wetness of the steam at exit from the turbine, and this
leads to two benefits - firstly a reduction in blade erosion (the steam
wetness should be no more than 15%), and secondly an increase in the
isentropic efficiency of the turbine.

Fig 3.3 Temperature-Entropy plots showing the effect of: a) raising the
temperature, and b) raising the pressure on the Rankine Cycle

34
Although the turbine expansion is shown in Figure 3.3 as an isentropic process,
it will in reality be irreversible. A useful design guide is that:

Every 1 percentage point increase in the wetness of the steam

in the turbine, leads to a 1 percentage point reduction in the
turbine isentropic efficiency.

An alternative way of discussing the changes in the cycle, is to consider the

change as an additional cycle, 211*2*. This additional cycle clearly has
an increased Mean Temperature of Heat Reception _i, and thus the
combined cycle will be more efficient.

The other advantage of raising the turbine inlet temperature is an increased

specific work output. The higher temperature leaving the boiler implies a
greater specific heat input, and when this is combined with the higher cycle
efficiency, then there will be an increase in the work output.

Raising the boiler pressure, and thus the turbine inlet pressure is illustrated in
Figure 3.3b, which shows:

a) An increase in the Mean Temperature of Heat Reception _i, but

b) An increase in the wetness of the turbine exhaust, which will lead to a
lower turbine isentropic efficiency.

The overall effect on efficiency and output of raising the system pressure with a
fixed turbine inlet temperature cannot be generalised. However, a combination
of raising the turbine inlet pressure and temperature will lead to an increase in
output and efficiency.

35
3.2.2 Regenerative Feed Water Heating (Hay Ch7)

Regenerative feed water heating is used to improve steam cycle efficiency, by

means of discrete feed water heaters. Up to 6 or so stages of feed water
heating can be used, but for simplicity only one stage will be considered here.
Figure 3.5 shows a Direct Contact Feed Heater (DCFH), in which the
steam bled from the turbine (2) heats the subcooled water from the first feed
water pump (45) to its saturation temperature. Since the water and steam
can be assumed to be at equilibrium within the DCFH, then the water leaving the
feed heater (6) must be saturated.

Fig 3.4 System diagram and Temperature-Entropy plot for a cycle with a
single stage of regenerative feed water heating
The reason that regenerative feed water heating leads to an improvement in the
cycle efficiency, is that there is a raising of the Mean Temperature of Heat

Reception, Ti , with no change to the Mean Temperature of Heat

Rejection To . The heat input is now occurring from T7 to T1, whilst without

36
feed water heating the heat input would have been from T5* to T1.

Feed water heating can give up to about a 10% improvement in the cycle
efficiency, with an infinite number of feed heaters. Increasing the number of feed
heaters is subject to a law of diminishing returns, since the improvement in cycle
efficiency is approximately proportional to:

number of feed heaters / (number of feed heaters + 1) (3.8

Thus one feed heater gives about half of the maximum gain in cycle efficiency,
two feed heaters gives two-thirds, and so on.

Other benefits of feed water heating are:

a) the reduction in the mass flow rate of steam at the low pressures (where it
has a particularly low density) and the blades tend to be excessively long.

b) the bled steam can be used to withdraw any condensate; this is particularly
useful in the low pressure stages where the steam is quite wet.

Whilst feed water heating will lead to an improvement in the steam cycle
efficiency, consideration must be given to the boiler, if there is not to be a
reduction in the overall system efficiency. The feed water is heated by the
combustion products (in the economiser), just before they leave the boiler
(see Figure 3.6). If the feed water enters the boiler at a higher temperature,
then the combustion products will leave the boiler at a higher temperature,
and the improvement in the steam cycle efficiency will be offset by a reduction in
the boiler efficiency. The fall in boiler efficiency can be recovered by using the
flue gases to pre-heat the combustion air.

37
The optimum gain in cycle efficiency occurs when the feed water heating is at
carefully chosen pressures; these approximate to giving equal enthalpy rises in
the economiser and each of the feed water heating stages.

Analysis of regenerative feed water heating is comparatively straightforward,

since an energy balance can be applied to the DCFH. However, it is essential
to remember that the mass flow rates vary with position in the cycle.
Assuming the DCFH is adiabatic, and with reference to Figure 3.4, apply the
SFEE:

y h2 + (x-y) h5 = x h6 (3.9

where: x is the mass flow rate through the boiler, and

y is the mass flow rate of steam bled from the turbine.

Care is needed in evaluating the turbine output (Wt), since the mass flow rate
through the turbine is reduced by the steam extraction to the feed water heater.

Wt = x(h1 - h2) + (x - y)(h2 - h3) = xh1 -yh2 + (x-y)h3 (3.10

Similarly, when the feed pump work is evaluated, there is a different flow rate in
each pump:

Wp,45 = (x - y)(h5 - h4), Wp,67 = x (h7 - h6) (3.11

3.2.3 Reheating
Reheating would almost inevitably be used in conjunction with feed water
heating, but for analytical simplicity, reheating will be considered here alone.

38
Figure 3.5 shows a plant layout and an associated T-s diagram, which has been
distorted to clarify the reheat part of the cycle. The extra heat input will lead to
an increase in the work output, since the enthalpy change across the second
turbine (34) will be greater than if the steam had been expanded with no
reheat (2a), but will there be an increase in the cycle efficiency?

Fig 3.5 System diagram and Temperature-Entropy plot for a cycle with
re-heating

If the Mean Temperature of Heat Reception Ti in the reheater (23) is

higher than in the original part of the cycle (61), then there will be an
improvement to the cycle efficiency. Figure 3.5 indicates, that by a suitable
choice of pressure for the reheater, then there will indeed be an improvement in
the cycle efficiency. The higher the reheat pressure, then:

a) the higher the Mean Temperature of Heat Reception _i in the

reheater (23), but

39
b) the lower the work output from the second turbine (34), and the
smaller the contribution to the overall cycle efficiency.

These two competing effects can be quantified by treating the reheat cycle as
two linked cycles (A and B), with identical mass flow rates.

Cycle A 1a56, Corresponding to a simple Rankine cycle

Cycle B a236, Corresponding to the reheat part of the cycle

The work outputs from cycles A and B are respectively:

WA = ηAQ61, and WB = ηBQ23 (3.12

The combined cycle efficiency, ηcy, is given by

W A+WB  Q + Q
 cy = = A 61 B 23 (3.13
Q61 + Q23 Q61 + Q23

and the increase in cycle efficiency caused by adding the reheat cycle
(ηcy - ηA) is

Since the steam is always wet when leaving the turbine, the Mean

Temperature of Heat Rejection T o will be the same for both cycles A and
B, and the combined cycle, and the efficiencies of cycles A and B will only

 A Q61 +  B Q23  A Q61 + Q23 

 cy -  A = -
Q61 + Q23 Q61 + Q23
(3.14
Q23  B -  A 
=
Q61 + Q23

40
depend on the Mean Temperatures of Heat Reception. Equation 3.14
shows that the effect of raising the reheat pressure on the cycle efficiency will
be a balance between: the increase in ηB, and the decrease in Q23.

Fig 3.6 The arrangement of a boiler that incorporates an air pre-heater

and re-heating

For isentropic turbines without regenerative feed water heating, the optimum
reheat pressure is about one quarter of the boiler pressure. However, real
turbines are irreversible, and an important benefit of reheating is the reduction
in the wetness of the steam in the low pressure turbine stages (with the ensuing
improvement in turbine isentropic efficiency); this results in a much lower
optimum reheat pressure.

For a reheat cycle with regenerative feed water heating and non-isentropic

41
turbine expansion, then the reheat pressure should be in the region of 20-25% of
the boiler pressure, with an ensuing 4-5% improvement in the cycle efficiency.

Finally, Figure 3.6 shows the arrangement of a boiler that incorporates re-
heaters. It should be noted the way in which the combustion products are
cooled by the economiser and finally the air pre-heater. The steam
drum ensures that saturated steam (and no water) enters the superheater - this
is analogous to the separator in the reversed power vapour cycles. The flow in
the water tubes is by natural convection, but when boiling occurs there is a two-
phase flow, and the vapour enhances the density difference.

42
4 GAS TURBINES
4.1 Introduction

As with Reciprocating Internal Combustion Engines, Gas Turbines invariably

operate as Open Circuit devices. When the heat input is from the combustion
of a fuel, then the cost, complexity and bulk of heat exchangers can be avoided,
if air is used as a working fluid, and the fuel is burnt in the air. None the less,
some closed cycles have been constructed, and these are most attractive when
the energy source is not from combustion.

Useful texts include:

Richard Harman Gas Turbine Engineering, Macmillan 1981.
with more analytical information available in:
H Cohen, GFC Rogers & HIH Saravanamuttoo, Gas Turbine Theory,
2nd Ed, Longman 1972.

4.2 Gas Turbine Cycle Analysis - REVISION

The ideal cycle for the gas turbine is known as the Joule Cycle (UK) or the
Brayton Cycle (USA). It consists of the following four processes, that are also
illustrated in Figure 4.1:

12 Adiabatic and reversible (thus isentropic) compression

23 Constant pressure (isobaric) heat addition
34 Adiabatic and reversible (thus isentropic) expansion
41 Constant pressure (isobaric) heat rejection.

43
Figure 4.1a shows the Joule Cycle on the P-V state diagram.
The heat input at constant pressure (23) causes a reduction in density (ie an
increase in volume), so that the volume flow rate through the turbine is greater
than the volume flow rate through the compressor. Since the work is vdp, then
the power output from the turbine will be greater than the power input to the
turbine.
However, Figure 4.1a also shows that the power input to the compressor (Area
1ab2) is large compared to the net power output (Area 1234), so that slight
imperfections in the compressor and turbine will lead to a rapid reduction in the
net power output, and thus also the cycle efficiency.

Fig 4.1 The simple ideal gas turbine cycle

The Joule Cycle processes are identical to the Rankine Cycle (Section 3.1), but
since there will be no phase change, the heat transfer processes are always
accompanied by temperature changes, as can be seen in Figure 4.1b. Since we
are assuming here Perfect Gas behaviour, the specific heat capacities at

constant pressure and constant volume are constant, and dh = cpdT can be
integrated directly. In other words, the T-s and h-s plots will be identical, except

44
for a vertical scaling factor, which is the specific heat capacity at constant
pressure ,cp.
In order to understand gas turbine performance, it is important to appreciate the
form of the isobars on the T-s or h-s plots. For any incremental process:

dq = Tds (4.1

the General Thermodynamic Relation

dh = Tds - vdp (4.2

From the SFEE dh = dq - dw, thus in steady flow dw = vdp,

Partial differentiation of Equation 4.2 gives:
a) (h/s)p = T That is, the isobars have a slope that is proportional to
the absolute temperature. At a given temperature all
isobars have the same slope, and if an isobar is followed
with increasing temperature, then its slope increases.
(4.3
b) (h/p)s = v That is, the vertical separation between the isobars
increases with temperature, and for an isentropic
process the high pressure isobar has a higher enthalpy
or temperature.

The compression (12) and expansion (34) processes are vertical lines on
the T-s and h-s plots, since we have specified them to be isentropic. If we apply
the SFEE with the usual assumptions of negligible changes in Kinetic Energy
and Potential Energy, then since the processes are adiabatic:

wc = (h2 -h1) = cp(T2 -T1) (4.4

45
and
wt = (h3 -h4) = cp(T3 -T4) (4.5
Thus the vertical distance of the compression and expansion processes on the
T-s and h-s plots will correspond to the work flow.

For a Perfect Gas (in which the ratio of heat capacities, γ, is constant) we can
combine the equation for an isentropic process (pvγ = const.) with the
Equation of State (pv = RT), to give:

T/p(γ-1)/γ = konst., or T2/T1 = (p2/p1)(γ-1)/γ and T3/T4 = (p3/p4)(γ-1)/γ (4.6

Since p3 = p2 and p4 = p1, the pressure ratios (rp) across the compressor

(p2/p1) and the turbine (p4/p3) are identical, and they will have the same

isentropic temperature ratio (ρ). In other words:

T2 / T1 = (p2 / p1)(γ-1) / γ = T3 / T4 = (p3 / p4)(γ-1) / γ = rp(γ-1) / γ = ρ (4.7

The thermal efficiency of any cycle is given by equation 1.3:

η = wnet / qi
By applying the Steady Flow Energy Equation (SFEE), and making the usual
assumptions about negligible changes in Kinetic Energy and Potential Energy,
then for the Joule Cycle:

ηJoule = wnet / qi = (qi -qo) / qi = 1 - qo / qi = 1 - (h4 - h1)/(h3 - h2) (4.8

For a Perfect Gas: h = cpT, and substituting for T2 and T3 in terms of T1 and T4
(using equation 4.7) gives:

ηJoule = 1 - cp(T4 - T1)/cp(T3 - T2) = 1 - (T4 - T1)/(ρT4 - ρT1) = 1 - 1/ρ

46
 (4.9
or ηJoule = 1 - 1/rp(γ-1) / γ
This important result tells us, that for the ideal gas turbine cycle, the efficiency
is only a function of the pressure ratio (rp) or ρ. However, it will be seen next,
that the specific work output and the efficiency of the irreversible cycle are both
dependent on the ratio of the maximum and minimum cycle
temperatures (θ).

4.3 Specific Work Output and Irreversible Cycle Analysis

The specific work output (w, the work output per kg of fluid circulated), is
influenced by the pressure ratio and the ratio of the maximum and minimum
cycle temperatures (θ). The minimum cycle temperature will be determined by
the environment and the economics of the heat exchanger - the closer T1 is to

To, then the larger and more expensive the heat exchanger. The maximum

cycle temperature (T3) is limited by the materials available for turbine blades -
about 1750 K for cooled blades, and 1200 K for un-cooled blades.

Fig 4.2 The influence of the pressure ratio on the simple ideal gas
turbine cycle specific work output

47
Figure 4.2 illustrates the way in which the pressure ratio will affect the work
output for operation between specified temperature limits. Three cases are
considered:

a) A very small pressure ratio, which shows that when rp is unity, then w = 0

b) A pressure ratio which shows that when the compressor delivery

temperature equals the maximum cycle temperature (ρ = θ), again w = 0.

c) An intermediate pressure ratio, in which the turbine and compressor outlet

temperatures are equal, which it will be shown corresponds to the
maximum work output.

Whilst case (b) will have a higher cycle efficiency than case (c), it is of no
practical significance.

The specific work for the compressor and turbine can be expressed as follows,
in terms of the isentropic temperature ratio (ρ) from equation 4.7:

wc = cp(T2 - T1) = cpT1(ρ - 1) = cpT1ρ(1 - 1/ρ) (4.10

wt = cp(T3 - T4) = cpT3(1 - 1/ρ) = cpT1θ(1 - 1/ρ) (4.11

Where:
the ratio of the maximum and minimum cycle

temperatures is θ = T3 / T1
and the isentropic temperature ratio is ρ

48
The net work output (wnet) is the difference between the compressor work and
the turbine work:

wnet = wt - wc = cpT1θ (1 - 1/ρ) - cpT1ρ(1 - 1/ρ) = cpT1(θ - ρ)(1 - 1/ρ) (4.12

Inspection of equations 4.9 and 4.12 confirms that:

for case (a), when ρ = 1, then wnet = 0 and ηJoule = 0, and

for case (b), when ρ = θ, then wnet = 0 but ηJoule = (1 - 1/ρ) = ηCarnot

To find the isentropic temperature ratio (and by implication the pressure

ratio) that will give the maximum specific work, then equation 4.12 has to be
differentiated with respect to ρ and equated to zero:

dwnet/dρ = 0 = cpT1θ(1/ρ2) - cpT1(1 - 0); or ρ2 = θ (4.13

In other words the maximum specific work output occurs when:

(T2/T1)(T3/T4) = T3/T1 or T2 = T4,

as shown in Figure 4.2 case (c). (4.14

4.3.1 Irreversible Cycle Analysis

As will be realised from the treatment of the vapour cycles, neither the
compression process nor the expansion process will be isentropic.
The real compressor will not be reversible, so if it is adiabatic, then there will be
an increase in entropy for process 12; for purposes of comparison the

49
isentropic compression process will be shown as 12s. The h - s plot in Figure
4.3 illustrates the way in which this irreversibility causes an increase in the work
input, and the following definition of compressor isentropic efficiency (ηc):

ηc = isentropic work input = (h2s - h1)/(h2 - h1) (2.10

actual work input

Fig 4.3 Irreversible compression and expansion processes, compared

with the ideal gas turbine cycle

Since we are assuming Perfect Gas behaviour (for which h = cpT), equation
2.10 can be re-written as:
ηc = (T2s - T1) / (T2 - T1) (4.15

For which T2s is determined from an adaptation of equation 4.7 (in which the
subscript s is used for an isentropic process):

T2s/T1 = (p2 / p1)(γ-1) / γ = rp(γ-1) / γ = ρ (4.16

50
Similarly, the turbine will not be reversible, but it can be treated as adiabatic, and
there will be an increase in entropy for process 34. For comparison, the
isentropic expansion process will be shown as 34s - this is also shown on the
h - s plot in Figure 4.3.

The decrease in work output, leads to the same definition of turbine isentropic
efficiency (ηt) as in the steam cycle:

ηt = actual work output = (h3 - h4)/(h3 - h4s) (3.7

isentropic work output

Since we are again assuming Perfect Gas behaviour, equation 3.7 can be re-
written as:

ηt = (T3 - T4) / (T3 - T4s) (4.17

Equation 4.7 can again be adapted to evaluate T4s

T3/T4s = (p3 / p4)(γ-1) / γ = rp(γ-1) / γ = ρ, or T4s = T3 / ρ (4.18

Note the use of broken lines in Figure 4.3 for the irreversible processes, and
chain-dashed lines for the hypothetical processes.

Equation 4.8 is valid for any cycle, so the thermal efficiency of the irreversible
cycle is:

ηcy = wnet / qi = (qi -qo) / qi = 1 - qo / qi = 1 - (h4 - h1) / (h3 - h2) (4.8

51
Or for a Perfect Gas:

ηcy = 1 - (T4 - T1) / (T3 - T2) (4.19

Equation 4.19 can be evaluated, by expressing temperatures T2, T3 and T4 in

terms of T1.

Combining and rearranging equations 4.15 and 4.16 gives:

T2 = T1 + (T2s - T1) / ηc = T1 {1 + (ρ - 1) / ηc} (4.20

From the definition of the maximum to minimum cycle temperature (θ):

T3 = θT1 (4.21

Combining and rearranging equations 4.17 and 4.18, and substitution for T3 from
equation 4.23 gives:

T4 = T3 - (T3 - T4s) ηt = θT1{1 - (1 - 1/ρ)ηt} (4.22

Substitution from equations 4.20 to 4.22 into equation 4.19 gives:

ηcy = 1 - [θ {1 - (1 - 1/ρ)ηt} - 1] / [θ - 1 - (ρ - 1)/ηc]

Rearranging gives:

52
  c   - (  - 1)  t  -   c
 cy = 1 -
 c  -  c  -  (  - 1) -  c (  - 1)

 c  -  c  -  (  - 1) -  c  + (  - 1)  t  c  +   c
=
  c (  - 1) - (  - 1)  (4.23

(  - 1)  t  c  -  
=
  c (  - 1) - (  - 1) 

This is not a result that is especially amenable to further analysis, and an

alternative approach is to establish the operating conditions, for which the turbine
work output (wt) is greater than the compressor work input (wc), that is:

wt / wc > 1 (4.24
Re-writing equations 4.10 and 4.11, and incorporating the expressions for the
isentropic efficiencies (equations 4.15 and 4.17) gives:

wc = cp(T2s - T1) / ηc = cpT1(ρ - 1) / ηc = cpT1ρ(1 - 1/ρ) / ηc (4.25

wt = cp(T3 - T4s)ηt = cpT3(1 - 1/ρ)ηt = cpT1θ(1 - 1/ρ)ηt (4.26

Substitution from equations 4.25 and 4.26 into equation 4.24 gives:
cpT1θ(1 - 1/ρ)ηt > cpT1ρ(1 - 1/ρ)/ηc

or θηt > ρ/ηc or θ > ρ/(ηcηt) (4.27

Because of materials limitations on the turbine entry temperature, and the low
isentropic efficiencies of early compressors and turbines, it was not until the
1940s that gas turbines were produced that had a net work output.

The net work output (wnet) is the difference between the compressor work and
the turbine work, and equation 4.12 becomes:

53
 wnet = wt - wc = cpT1θ(1 - 1/ρ)ηt - cpT1ρ(1 - 1/ρ)/ηc

= cpT1(θηt - ρ/ηc) (1 - 1/ρ) = cpT1(θηcηt - ρ) (1 - 1/ρ) / ηc (4.28

wnet = cpT1(α - ρ)(1 - 1/ρ)/ηc where: α = θηcηt

As before, we can find the isentropic temperature ratio (and by implication

the pressure ratio) that will give the maximum specific work, by differentiating
equation 4.28 with respect to ρ and equating to zero:

dwnet/dρ = 0 = θ(0 + 1/ρ2)ηt - (1 - 0)/ηc or: ρ2 = θηcηt = α (4.29

The efficiency and specific work output as a function of the isentropic
temperature ratio are plotted in Figure 4.4, for the:

a) Joule cycle (equations 4.9 and 4.12), and

b) the irreversible cycle (equations 4.23 and 4.28)

Figure 4.4 shows:

a) The substantial reduction in cycle efficiency caused by quite small
reductions in the isentropic efficiencies - even at the isentropic temperature
ratio for maximum efficiency.
b) That as the ratio of the maximum to minimum cycle temperature (θ) is
increased, then it is necessary to increase the isentropic temperature
ratio if the cycle efficiency is to be optimised.
c) That the isentropic temperature ratio for the maximum specific work
output is slightly below the isentropic temperature ratio for maximum
efficiency - result that can be confirmed by differentiating equation 4.23.
d) That a pressure ratio in the region of 25:1 is needed for operating
contemporary gas turbines (θ = 5, ηc = 0.85, and ηt = 0.88) in the region of
maximum efficiency.

54
Fig 4.4 Cycle efficiency and specific work output of simple gas turbine
cycles

4.4 Practical Gas Turbine Plant

When air is used as the working fluid, and the heat input is a result of
combustion, then there is no need for heat exchangers, and the resulting open
circuit plant is shown in Figure 4.5.

55
Fig 4.5 An Open Circuit gas turbine with h-s plot

This is clearly not a thermodynamic cycle, but the T - s and h - s plots look much
the same. However, it must be appreciated that:

a) The mass flow rate through the turbine (mex) is larger than the flow rate

through the compressor (ma), because of the addition of the fuel (mf):

mex = ma + mf = ma(1 + 1/AFR) where: AFR = ma/mf (4.30

b) The exhaust gases have different thermodynamic properties from air,
and for a contemporary gas turbine at its design point, then:

Property Air Exhaust

cp (kJ/kgK) 1.05 1.15
ratio of heat capacities, γ 1.38 1.33

The lower value of γ for the exhaust will cause a lower isentropic
temperature ratio across the turbine (for a given pressure ratio), and
thus lowering the work output. However, this is more than compensated
for by the higher heat capacity of the exhaust products.

56
c) There will be a pressure drop across the combustion chamber, and inlet
and exhaust ducting losses will further:

i) increase the pressure ratio across the compressor, and

ii) decrease the pressure ratio across the turbine.

The T - s diagram in Figure 4.5 takes into account these effects.

Fig 4.6 An Open Circuit gas turbine with a separate power turbine,
shown with its h - s plot

The open circuit plant shown in Figure 4.5, is only likely to be used in small
applications. A more realistic arrangement for power generation is shown in
Figure 4.6, along with its associated h - s diagram. The coupled turbine and
compressor are sometimes referred to as a gas generator. The advantage of
this arrangement is that the speed of the load (coupled to the power turbine) is

57
independent of the gas generator speed, and it is possible for the gas generator
speed to be controlled so as to obtain the maximum cycle efficiency. In a single
shaft system operating away from its design point, the overall efficiency is
reduced because:

a) the isentropic efficiencies of the turbine and compressor will be reduced,

and
b) the air fuel ratio will be weakened, and the turbine entry temperature will
be reduced.

In the gas generator, the turbine has to generate enough power to drive the
compressor:
ma wc = mex wt ηmech (4.31

where: ηmech is the mechanical efficiency (this is often so close to unity that
it is ignored)

To calculate the performance of such a twin-shaft gas turbine system, it is

usually a matter of having to find the pressure ratio across the gas generator
turbine. The approach is along the following lines:

a) evaluate the power requirement of the compressor (equation 4.25

multiplied by ma).
b) calculate the specific work required from the turbine (equations 4.31 and
4.30).
c) determine the temperature change across the turbine (using an
appropriate value of cp), and thus the isentropic temperature change and
the isentropic temperature ratio. Thence find the corresponding pressure
ratio (using an appropriate value of γ). All these steps are achieved

58
 through use of parts of equation 4.26.
d) the pressure ratio across the first turbine will enable the pressure ratio
across the power turbine to be determined. The power turbine entry
temperature is the exhaust temperature from the first turbine, so it is thus
possible to calculate the power turbine performance.

Fig 4.7 An Open Circuit gas turbine with an expansion nozzle (a turbo-
jet engine), shown with its h - s plot

A very important use of gas turbines is in aircraft propulsion, and this is

illustrated by the turbo-jet engine in Figure 4.7. The arrangement is very
similar to the gas generator/power turbine arrangement in Figure 4.6, except
there is now no power turbine. Instead, the momentum flux of the jet provides
a propulsive force, and clearly it is no longer possible to neglect the Kinetic
Energy of the exhaust.

59
By applying the Steady Flow Energy Equation across the nozzle (45) it is
possible to determine the jet velocity (vj):

h4 = h5 + ½vj2 (4.32a

or for a Perfect Gas: T4 = T5 + vj2 / 2cp (4.32b

The expansion in the nozzle from the turbine outlet pressure (p4) to the ambient

pressure (p5) is like the expansion in a turbine. Ideally, the expansion across the
nozzle is isentropic, but fluid friction causes a slight irreversibility which can be
described by the nozzle (isentropic) efficiency - ηn.

ηn = (h4 - h5)/(h4 - h5s) = (T4 - T5)/(T4 - T5s)

(4.33
Where: T5s / T4 = (p5s / p4)(γ-1) / γ

The analysis of this simple turbojet engine (Figure 4.7) is exactly the same as the
two-shaft gas turbine of Figure 4.6.

When a simple turbojet engine (Figure 4.7) is stationary:

 ex vj
Thrust  m
and
 ex vj2
the Kinetic Energy Flux, or Jet Power  m (4.34

In by-pass engines, turbo-fan engines or turbo-prop engines, some of the power

that would otherwise be available in the jet is used to drive a 'pump' (compressor,
fan or propeller), so that the jet velocity is reduced but with an increased mass
flow rate - it can be shown that this leads to a higher thrust, and thus a more

60
efficient propulsion system. An additional and very significant advantage of
reducing the jet velocity is the reduction in noise.

Digression for Enthusiasts

Neglecting Kinetic Energy terms may see somewhat questionable, when clearly
gas turbines operate with high flow rates and small flow areas. However, the
Kinetic Energy terms can readily be incorporated by use of the stagnation

enthalpy (h0), which incorporates the Kinetic Energy as follows:

ho = h + ½v2
and for a Perfect Gas (4.35
To = T + v2 / 2cp

When (for example) a thermocouple is placed in a high velocity flow, it is in fact

the stagnation temperature (To) that is measured. T and h are now referred

to as the static temperature and enthalpy.

When a flow is accelerated or decelerated reversibly and adiabatically (ie

isentropically) there will be a pressure change, and this leads to the concept of
stagnation pressure (p0)

po = p(To/T)γ / (γ-1) (4.36

By using stagnation properties in all the equations presented here for gas
turbine calculations, then the Kinetic Energy terms would be included correctly.

.o0o.

61
5.5 Means of Improving the Efficiency of Simple Gas
Turbines

4.5.1 Simple Reversible Gas Turbine Cycle with Heat Exchanger

If the outlet temperature from the turbine is higher than the compressor outlet
temperature, then a regenerative heat exchanger can be used to heat the
gas leaving the compressor, prior to any external heat input. Such a cycle is
shown in Figure 4.9, for which it will be assumed that the regenerative heat
exchanger is perfect. When this is combined with the Perfect Gas
assumption, then it means that:
a) there will be no temperature difference between the fluid being heated,
and the fluid being cooled, and
b) the temperature rise of the fluid being heated will equal the temperature
fall of the fluid being cooled.

Fig 4.9 An ideal regenerative gas turbine cycle, with a T-s plot

62
The result of the perfect heat exchanger assumption is illustrated by the T-s plot
in Figure 4.9, which shows that:

the regenerative heat exchange: qx = h3 - h2 = h5 - h6 (4.39

The heat input is now from 34 (without the regenerative heat exchanger it
would have been from 24), so Ti is raised, and the cycle efficiency will be
raised.

The heat output is now from 61 (without the regenerative heat exchanger it
would have been from 51), so To is lowered, and the cycle efficiency will be
raised.

Example 4.1

(a) Calculate the thermal efficiency and work ratio for the air-standard gas
turbine cycle with a pressure ratio of 5, and maximum and minimum
temperatures of 7500C and 200C. The compressor isentropic
efficiency is 0.80 and the turbine isentropic efficiency is 0.85; both the
compressor and turbine are adiabatic.

(b) What is the efficiency of the Joule cycle operating with the same pressure
ratio for: air, ii) helium. The ratio of specific heat capacities for helium
is 1.66.

(c) Calculate the thermal efficiency if a perfect regenerative heat exchanger

is added to the cycle in part (a).

63
 T1  293.15K, T3  1023.15K, r p  5, c  0.80, t  0.85
a)
 1 1.4  1

T2s  T1 r p  293.15 x 5 1.4  464.3
1 1 .4  1

T 4s  T 3 rp  1023.15 5 1.4  646.0K

64
 1,2s c p T2S  T1  T2s  T1 

c     for an adiabatic compressor with a
1,2 c p T2  T1  T2  T1 
perfect gas, if KE  PE  0

 3,4 c p T3  T4  T3  T4 
t     Adiabatic turbine, perfect gas, if
 3,4s c p T3  T4s  T3  T4s 

KE  PE  0
T2  T2s  T1  c  T1  464.3  293.15 0.8  292.15  507.1K
T4  T3  T3  T4s t  1023.15  1023.15  646.00.85  702.6K

 c p T3  T4   T2  T1  1023.15  702.6  507.1  293.15

 cycle  net    0.207
q in c p T3  T2  1023.15  507.1

/
b) ηJoule = 1 - 1/rp(γ-1) / γ Air: ηJoule = 1 - 1/5(1.4-1) 1.4 = 0.367

/
Helium: ηJoule = 1 - 1/5(1.66-1) 1.66 = 0.473

c) T1, T2, T3 and T4 are all unchanged by adding the heat exchanger
Perfect Regeneration => T
T2 = T4* = 507 K , T3
2* 3
T4 = T2* = 703 K
Work Ratio is unchanged,
but heat input reduces:
T2*
4* 4 T4
qin = cp(T3 – T2*) T2
T4*

T1 s

 net c p T3  T4   T2  T1  1023.15  702.6  507.1  293.15

 cycle     0.333
q in c p T3  T2*  1023.15  702.6.1

65
5.5.2 Combined Cycle Gas Turbines - Open Circuit Gas
Turbine with a Steam Cycle

The type of gas turbine used for stationary power generation has an efficiency in
the region of 33%. When the turbine exhaust gases are used as the energy
input to a steam cycle, then the overall efficiency rises to over 50%. The
arrangement of the simplest possible combined cycle is shown in Figure 4.10.

Fig 4.10 A Combined Cycle Gas Turbine (CCGT) plant, that combines a
gas turbine with a steam cycle

The analysis of a combined cycle is entirely straightforward, but it must be

appreciated that there will be different mass flow rates in the gas turbine and
the steam cycle, so that the use of specific work alone would lead to errors. The
ratio of mass flow rates is found from an energy balance on the boiler.

Table 4.1 summarises the operating conditions of a combined cycle (Figure

4.15); the data are based on the 900MW Killingholme plant, but the steam
cycle has been simplified here.

66
Table 4.1 Operating conditions of a Combined Cycle (Figure 4.10)
___________________________________________________

Position Pressure (bar) Temperature (C)

___________________________________________________

1 1.0 15
2 10.7 323
3 10.7 1050
4 1.0 553
5 1.0 107
6 80.0 522
7 0.0685 Dryness - 0.87
8 0.0685 39
9 80.0 40
___________________________________________________

To find the ratio of mass flow rates in the two parts of the system, we apply an
energy balance (SFEE) to the boiler:

Let: X = mass flow rate of combustion products

Y = mass flow rate of water

Applying the SFEE equation to the boiler gives:

X  cp,ex(T4 - T5) = y (h6– h9)

Assuming a mean value of cp,ex to be 1.15 kJ/kgK, and taking the tabulated
enthalpy values for the steam and water, gives:

X  1.15 (553 - 107) = Y (3453 - 163) or X/Y = 6.35

The Killingholme plant consists of 3 modules, each of which comprises two gas
turbines, two boilers and a steam turbine: the mass flow rate through each gas
turbine is 500 kg/s.

67