hit. J. ReJHg. Vol. 18, No. 6, pp.

421-433, 1995
[•IUTTERWORTH Copyright :C: 1995 Elsevier Science Limited and IIR
"~E I N E M AN N
Printed in Great Britain. All rights reserved
0140-7007/95/$10.00 + .00

An exergy method for compressor performance analysis

J. A. McGovern and S. Harte*
Department of Mechanical and Manufacturing Engineering, University of Dublin,
T r i n i t y College, D u b l i n 2, I r e l a n d
Received 19 October 1994; revised 6 April 1995

An exergy method for compressor performance analysis is presented. The purpose of this is to identify and
quantify defects in the use of a compressor's shaft power. This information can be used as the basis for
compressor design improvements. The defects are attributed to friction, irreversible heat transfer, fluid
throttling, and irreversible fluid mixing. They are described, on a common basis, as exergy destruction rates
and their locations are identified. The method can be used with any type of positive displacement
compressor. It is most readily applied where a detailed computer simulation program is available for the
compressor. An analysis of an open reciprocating refrigeration compressor that used R12 refrigerant is
given as an example. The results that are presented consist of graphs of the instantaneous rates of exergy
destruction according to the mechanisms involved, a pie chart of the breakdown of the average shaft power
wastage by mechanism, and a pie chart with a breakdown by location.
(Keywords: performance analysis; compressor; exergy; simulation)

Mathode exerg tique d'analyse des performances

des compresseurs
On prgsente une m~thode exerg~tique d'analyse des perJormances des compresseurs, afin d'ident(fier et de
quant(fier h,s &![huts ~;ventuels pouvant qffecter h,ur puissance gt I'arbre. Les donnOes ainsi receullies peuvent
servir h des ameliorations dans la conception du compresseur. Les d~!fituts coneern~s sont d#s ~'t la jHction, au
tran.~'[i,rt de chaleur irreversible, h l'dtranglenwnt des fluides et aux mOlanges irr~versibles de /tuides. On les
caracterise, sur un plan unique, enJonction de leurs taux de destruction exerg{tique et leurs emplacements sont
indiqu{s. La m{thode pr6sent~e convient a tousles types de eompresseur volumdtrique, mais particulE, rement Ih
oit I'on dispose d'un programme inJormatique de simulation d#tailh~e du fimctionnement du compresseur. On
donne comme exen?ple l'analyse d'un eompresseur volumgtrique utilisant le R12. Les r~;sultats sont prOsentOs
sous Jorme de graphiques montrant les taux de destruction instantan~e d'exergie, en Jonction des m~;canismes
irnpliqu~s, ainsi que de deux graphiques 'camembert' montrant la r@artition de la perte mo venne de capacity;
selon le nl~;eanisme et selon la r~partition par endroit.
(Mots cl~s: compresseur; performance: exergie; mesure: simulation: calcul)

This paper describes an exergy analysis technique to valve, or with that due to heat transfer between the
quantify defects in the shaft power utilization of a compressor and the surroundings. Even leakage can be
compressor. The technique provides a breakdown of the characterized in terms of exergy destruction, as it involves
causes of shaft power wastage that is valuable to a design throttling and irreversible fluid mixing. The shaft power
engineer. When the main mechanisms of power wastage wastage due to leakage can thus be quantified.
have been located and quantified, appropriate modifica- The case study included in this paper is that of a
tions can be made to improve the performance of the reciprocating refrigeration compressor. A detailed simu-
compressor. It is shown that the new exergy method has lation program was used and experimental data were
several advantages when compared with conventional available to validate some of the main simulation results.
techniques that are used for the same purpose.
Shaft power wastage in compressors can be attributed
Literature review
to fluid throttling, irreversible fluid mixing, friction, and
irreversible heat transfer. All such defects can be Dossat I adopted a conventional approach and broke
described as exergy destruction rates. The locations down the overall efficiency of a refrigeration compressor
and mechanisms of the defects can be identified. For by defining the mechanical efficiency and the compres-
instance, the exergy destruction at the suction valve can sion efficiency. The mechanical efficiency was the ratio of
be compared with the exergy destruction at the discharge the indicated power to the shaft power input. It was
less than unity due to the effects of mechanical friction
*Present address: Ford Motor Company, Climate Control Division, within the compressor. The compression efficiency was
Dearborn, M1 48120, USA calculated by comparing the indicated power with the

421
422 J. A. McGovern and S. Harte

Nomenclature
b Specific flow exergy function (b = h - Tos ) 1) Specific volume (m 3 kg-l)
(Jkg -1) W Rate at which work is done on a system (W)
Cd Coefficient of discharge Compressor shaft power input (W)
cv Specific heat at constant volume (J kg- 1 K - 1) X Distance in x direction (m)
Cp Specific heat at constant pressure Y Distance in y direction (m)
(Jkg -I K -l) Z Distance in z direction (m)
h Specific enthalpy (Jkg 1 K l)
h' Specific enthalpy just downstream of an Greek letters
adiabatic throttling process (J kg -1 K -t) A Change in or increment of a quantity
] Irreversibility rate (exergy destruction rate) (5 Very small change in, or increment of, a
(w) quantity
m Mass (kg) ~,1 Rational efficiency defect
rh Mass flowrate (kg s- l) r/ Rational efficiency
or 0 Absolute temperature (K)
Rate of change of the mass of a system p Density (kgm -3)
(kgs 1) ~. Rate of density change (kg m 3 s l)
P. Absolute pressure (Pa) E Rate of exergy transfer or rate of exergy
Q Heat transfer rate to a system (W) ,.:.,,
change (W)
R Specific gas constant (Jkg -1 K 1) :c Rate of exergy transfer to the stream of fluid
Rate of change of entropy (J K - 1 s J) that passes through the compressor (W)
Scr Rate of entropy creation (JK -1 s 1)
S Specific entropy (J kg i K - I ) Subscripts
Rate of change of specific entropy 1~2, . . . Flow position
(Jkg I K - l s 1) bulk Refers to a characteristic or property of the
s' Specific entropy just downstream of an bulk fluid
adiabatic throttling process (J kg-1 K - 1) C Compressor
T Absolute temperature (K) cr Creation
T' Absolute temperature just downstream of an d Discharge
adiabatic throttling process (K) Identifier for exit flow stream
i/" Rate of change of absolute temperature f Associated with friction
(Ks -I ) i Identifier for inlet flow stream
To Absolute temperature of the environment J Identifier for a thermal mass
(K) k Identifier for a thermal mass
t Time (s) loc Local
U Internal energy (J) mx Associated with mixing
/.) Rate of change of the internal energy of a new New value calculated after a small time
system (J s- 1) interval in a finite difference calculation
u Specific internal energy (J kg- 1) rev Reversible
~i Rate of change of the specific internal energy S Suction
of a system (J kg 1 s-I) th Associated with throttling
V Volume (m 3) O Associated with heat transfer
I? Volume change rate (m 3 s i)

isentropic power required to produce the same refrigera- Exergy techniques have previously been suggested for
tion effect. The compression efficiency took account of application to compressors. Stecco 5 and McGovern 6
effects such as wire drawing (throttling), heat exchange both argued the case for using exergy analysis to measure
between the vapour and cylinder walls, and fluid friction. the efficiency of compressors. Both concluded that the
Dossat pointed out that the relationships between the exergy approach had the advantage that it could be
various factors that influenced the compression efficiency applied to all types of compressors, and that it could be
were difficult to evaluate mathematically. 4 integrated into an analysis of a whole plant or cycle.
FrancoDaCosta 2, Pandaya 3 and Sakaino described Townsend 7 applied exergy techniques to analyse a
the interrelationships between these factors. FrancoDa- multistage refrigeration system. He showed how exergy
Costa outlined a model suitable for use with a simulation techniques were quickly applied and how plant design
program, while Sakaino defined a model suitable for improvements were made due to the added insight the
use with experimental data. Estimations and simplifying exergy techniques gave. Xu 8 used exergy techniques to
assumptions that were somewhat arbitrary or subjective experimentally determine the thermodynamic losses in
had to be made. For example, Sakaino indirectly the components of refrigerators and freezers: in all cases
estimated the power wastage due to heat transfer by the largest exergy destruction rates occurred at the
subtracting the ideal power and the wastage by the other compressor.
mechanisms from the indicated power. However, it is believed that no other authors have
 An exergy method for compressor performance analysis 423

1. Suction plenum
fluid system

A '°

Figure 1 Schematic representation of the compressor. The four fluid

systems within the compressor are named. The four solid masses are
cross-hatched and are identified by the letters A to D. The solid masses
are used in the modelling of the heat transfer processes and do not
represent the assembly details, nor do they correspond in n u m b e r to the
actual components
Figure 1 Repr&entation sch~matique du compresseur. On indique les Figure 3 Heat transfer paths between the fluid systems and solid
quatre circuits de fluides h I'int&ieur du compresseur. Les quatre masses masses of the compressor and between the compressor and the
solides sont marqu&s par des hachures crois&s et sont identifies par les environment. The arrows show the normal directions for heat transfer.
lettres A h D. Ces masses sont utilis&s pour le modellage des processus de In some cases heat transfer also occurs in the direction opposite to that
tran.ffbrt de chaleur et ne repr&en tent pas les d~tails du montage, pas plus of the arrow
qu'elles ne correspondent en nomhre aux composants rdels Figure 3 Chemins de transfi, rt de chaleur entre les circuits et les masses
solides du eompresseur, ainsi qu'entre le compresseur et l'environement.
Les flbches indiquent lessens normaux de tran.ff~,rt de chaleur, mais dans
certains cas, eehd-ci peut s'eff?etuer en sens inverse

Suction Discharge
plenum plenum
in a PhD thesis by M c G o v e r n 12 for the speed range
)
3 0 0 - 9 0 0 r e v m i n -I. All the experimental work was
Suction Discharge carried out before the completion of that PhD, and the
pipe pipe refrigerant used was R12.
The simulation model that was used to represent the
Compression compressor is outlined in this paper, as the particular
chamber exergy analysis results that follow depend on the
assumptions that were made in the simulation. A less
detailed model would have resulted in a less refined
Crankcase exergy analysis, while a more detailed and accurate
Figure 2 Flow paths between the fluid systems of the compressor. The model could have provided a more precise breakdown of
arrows indicate the normal flow directions, but in some cases reverse the causes of shaft power wastage.
flows also occur A major limitation of the ideal gas model with
Figure 2 Chemins d'&'oulement entre les circuits du compresseur. Les constant specific heats that was used was that it was
flbches indiquent lessens normaux d'&'oulement, mais dans certains cas il not possible to obtain simultaneous agreement between
peut se produirc un &'oulement inverse
simulated and experimental results for (1) the mass
flowrate and (2) the refrigerant temperatures. In an
industrially funded project, which cannot be reported on
described the use of exergy techniques to determine the at this time, the authors applied the same techniques to a
instantaneous and average rates of shaft power wastage, more complex rotary positive displacement compressor.
apportioned to either specific mechanisms or specific In this case the refrigerant was R134a. Real gas
systems, within a positive displacement compressor. This equations 13 were used in the computer model and
paper supersedes a paper presented by the authors at the much better agreement between experimental results
1992 International Compressor Engineering Conference and simulation results was achieved.
at Purdue 9. Figure 1 is a schematic representation of the compres-
sor. There are four fluid systems within it, and four solid
masses. The flow paths between the fluid systems are
The compressor model
shown in Figure 2, while the heat transfer paths between
The compressor that was simulated and analysed has the fluid systems and solid masses are shown in Figure 3.
been described in two previous papers 1°'11. It was an A conventional first-law approach was used in
open-type air-cooled belt-driven compressor of relatively modelling the fluid systems within the compressor (see
straightforward design. It had two cylinders, but the Appendix 1). The simulation program was largely based
computer model was for one cylinder only. It had a on those published as Purdue short courses 14,15. Ideal gas
crank radius of 0.025 m, a piston diameter of 0.065 m, behaviour was assumed. In addition, the fluid systems
and a clearance volume of 4.51 ml per cylinder. Detailed were modelled in a special way to facilitate the exergy
experimental results for the compressor were included analysis. The method was fully consistent with the
424 J. A. McGovern and S. Harte

Pressures Temperatures of the solid masses

1.4

1.2

1
! E sTrg n°m ~ 340
350
Discharge plenum

1 1 Head partition
i
!
I

330 i
a. Suction plenum I
~ 320 Cylinder, crankcase, and piston
0.8

0.6
310 I
] Crankcase : 3O£ , 1
n
0.4 . . . . . .
45 90 135 180 225 270 315 360
Crank angle/[deg]
f
0.2

I Suction plenum j [
0
Temperatures of the fluid systems
0 45 90 135 180 225 270 315 360

Crank angle l ( d e g ]
Discharge plenum Crankcase
~' 36o
Valve lift "~ 340 I
~ , ! ~ Cylinder
& 320 ! ! ~\/.v---- . . . .
I I E
4 uconvave i
300 "4.. I ~t
~ - Suction plenum ] !
280 , h , ,
0 45 90 135 180 225 270 315 360
~, ~/ Discharge valve ~ I~
Crank angle/[deg]

Figure 5 Diagrams, as predicted by the simulation program, of the

instantaneous temperatures of the eompressor's fluid systems and solid
masses
45 90 135 180 225 270 315 360 Figure 5 Diagrammes ( comme pr~;vues par le programme de simula-
Crank angle/[deg] tion) des temp#ratures instantan#es des ('ircuits et des masses solides du
'ompresseur
Figure 4 Diagrams, as predicted by the simulation program, of
pressures within the fluid systems of the compressor and valve lifts
Figure 4 Diagrammes, pr#vues par le programme de simulation, des
pressions ~t l'intOrieure des circuits du eompresseur et des lev6es des the suction plenum and entering the discharge plenum. A
soupapes constant convection coefficient was used for the fluid
system in the crankcase also: this was given an estimated
value. In the compression chamber the instantaneous
conventional first-law approach, but incorporated heat transfer area was used and the heat transfer
specific models for the types of thermodynamic irrever- coefficient was taken to be proportional to the cylinder
sibility that were assumed to occur. For each of the non- pressure. The constant of proportionality was adjusted in
equilibrium fluid systems within the compressor, the an attempt to match the experimental data for indicated
exergy analysis model included a similar fluid system in power, the temperature of the compressor body, and the
equilibrium. rate of heat loss from the compressor. Like the model of
The solid parts of the compressor were modelled, Rectenwald 17, which was very detailed, this crude heat
somewhat crudely, using a lumped heat capacity transfer model involved large variations in the heat
method 16 in which the first law of thermodynamics was transfer rate over the cycle. An estimated value was used
applied to calculate the new temperatures after each time for the external surface emissivity of the compressor. A
step. The calculation step used for all aspects of the constant in an empirical expression for the external
simulation was 0.5 ° of crank rotation at a speed of convection coefficient was chosen to give reasonable
600revmin -t. Estimated values were used for the agreement with the experimental results for the rate of
thermal resistances between the lumped masses. heat loss from the compressor.
The suction and discharge reed valves were modelled The friction power was determined experimentally by
using a mode shape dynamics model and experiment- subtracting the indicated power from the shaft power. In
ally derived effective flow and force area functions 14J5. the simulation model the instantaneous friction power
Other flow restrictions were modelled as fixed-area was assumed to be proportional to the product of the
orifices with appropriate Cd values. In all these cases piston area, the piston velocity, and the pressure
the flow equation for an isentropic nozzle was used to difference across the piston. All the friction energy was
calculate the flowrate. The isentropic condition applied assumed to be dissipated as heat transfer into the solid
only as far as the vena contracta: the overall enthalpy mass around the crankcase and compression chamber
difference across each flow restriction was taken to (Figure 3).
be zero. Figure 4 shows simulation results for the pressures
Constant convection coefficients were assumed for within the fluid systems of the compressor and the
heat transfer between the fluid systems in the suction and corresponding suction and discharge valve lifts. Figure 5
discharge plenums and the solid masses. Their values shows the corresponding simulation results for the
were selected to give reasonable agreement with experi- temperatures of the solid masses and of the fluid systems
mental measurements of the vapour temperature leaving within the compressor.
 An exergy method for compressor performance analysis 425

which the exergy method for compressor performance

Analysisboundary7 W= -p.~9 analysis is based.
For a compressor that operates in a steady-state
_ ..... /___" : ........ manner within a particular environment, the rate of
,-'1 I','ll exergy transfer to the stream of fluid being compressed
~1
i )- > can be evaluated from measurements of the mass
il I
flowrate, the thermodynamic properties at suction and
discharge, and the temperature of the environment. It is
rh, I-~-] ~ U, V, The,k,Phi,k, m ,,| ,
rh given by
7.'1
,, 0, ,:, ~-c : rn(bd - bs) (1)

i . . . . . . . . . . . . I The total exergy destruction rate is the difference

between the exergy input rate and the rate of exergy
transfer to the fluid stream. The exergy input rate is equal
Figure 6 A general fluid system, which could represent one of the fluid to the shaft power, which can be measured. Hence
systems within a compressor, lnlet flows enter through simple orifices.
Any exit flow restrictions are assumed to be outside the system
boundary: they will be inlet flow restrictions of adjacent downstream
I' = W~ - Ec (2)
systems
Figure 6 Sch&na d'un circuit g~n~ral qui pourrait repr&enter l'un des The rational efficiency is the rate of exergy transfer to the
ciruits de l'int&ieur d'un compresseur. Les &oulements h l'entr&s fluid stream divided by the rate of exergy input as shaft
s'effectuent par des oroqces simples. On prend comme hypothbse que toute power:
restriction h ['Ocoulement iz la sortie est en dehors des limites du systbme
~¢

r/- I~ (3)
Overall exergy analysis
Exergy analysis is based on the assumption that there is The difference between unity and the rational efficiency is
an infinite equilibrium environment that ultimately the total rational efficieno, defect. It is a decimal fraction
surrounds all systems that are to be analysed. The representing the rational inefficien O' of the compressor:
exergy (or available energy) of a system is the maximum
work that could be derived if the system were allowed to ~,/= 1 - r/ (4)
come to equilibrium with the environment. It is a
consequence of the second law of thermodynamics that
the combined exergy of all systems can only decrease or
Exergy analysis of a general fluid system
remain unchanged. Exergy can be destroyed, but not
created. Exergy can be exchanged between systems, but if For a general system as shown in Figure 6, which could
there are thermodynamic irreversibilities (such as friction correspond to one of the fluid systems within a
or heat transfer with finite temperature differences) some compressor, there are several types of change occurring
of the potential for the production of work is destroyed. simultaneously. There are unsteady flows of fluid in
In all real processes, therefore, the total exergy of all and out and there are unsteady heat transfer rates
systems decreases. For a specified system boundary a (which can be positive or negative). The mass, the
clear distinction can be made between exergy destruction volume, and all the thermodynamic properties of the
and exergy loss. Exergy loss is exergy that is passed on to system are changing.
some other system (often the environment) and which The rate of exergy destruction for the system can be
cannot be considered useful in the context of the purpose described as follows (Kotas TM, appendix B, Equation
of the plant. The term exergy destruction is used when the (B.14); Bejan 19, p 114):
potential for the production of work is destroyed within
the system boundary.
The exergy of a system is a co-property of the system ' ' Z<
i 7"/.] (5)
and the environment. In exergy analysis of compressors
the environment consists of the local surroundings of the The term S is the rate of change of entropy of this system.
compressor. These local surroundings are modelled as Equation (5) is related to Equation (35) of Appendix 2 by
being in equilibrium and infinite. Given sufficient infor- the G u o y Stodola relation:
mation, the exergy of all the systems within a compressor
can be determined at any time. By determining ] = ToScr (6)

1. the rate of change of the exergy of each of the For a system that was in a steady state or in equilibrium,
systems that comprise a compressor, Equation (5) could be evaluated provided there was
2. the rates of exergy exchange between them, enough information to describe the density and the
3. the rates ofexergy exchange between the compressor specific entropy as continuous functions over the volume
and systems outside the compressor, (right to the boundary) and over time:

the rates of exergy destruction associated with the S = I][(~p+ sfl)dxdydz (7)
compressor can be calculated. This is the principle on
426 J. A. McGovern and S. Harte

Irreversible Reversible themselves to define the total irreversibility rate for a

mixing mixing system as shown in Figure 6.
T,p T,p The equilibrium system of the exergy analysis model is
~7 _p . contained within the non-equilibrium system. The word
'system' on its own will refer to the non-equilibrium

I Irreversible heat transfer /

system. In addition to the assumptions made in
Appendix 1 for the first-law analysis, the following
assumptions are made to separate out the different
in a thin boundary layer " irreversibility mechanisms and to define the equilibrium
Ca) (b) system so that it is amenable to analysis:
Reversible
Irreversible heat heat transfer - - - - ' ~ //- Rmeve;ible 1. It is assumed that at any given time the mass of fluid
transfer within a ~ l within the system that has not yet entered the equilibrium
thermal resistance ---____._____.___. ] [ Q"~I [ system is small enough to be neglected. Also, it is
of negligible mass ] J'~ / T,p
anvoume hrnnd I/ assumed that the volume of the fluid that is not part of
the equilibrium system can be neglected.
2. There is no irreversibility associated with either
R:vtet~bn~er with a ~ . ' p~ expansion or contraction of the system boundary.
temperature glide
(c) Analysis boundary vc=-p~
1
Figure7 These diagrams explain the model used in the exergy analysis
of mixing of an incoming flow with the fluid already in a system. The
irreversible mixing process in (a) is replaced by irreversible heat transfer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

between the stream of incoming fluid and the fluid already in the
system. Two different representations of the same model are shown in
(b) and (c). The model applies while a small quantity of fluid enters the --*0=~ Equilibrium
system
fluid system
& --, , th
Figure 7 Ces schemas montrent le modble utilis~ pour l'analyse U, V,T,p,m
exergOtique du mOlange d'un ~coulement entrant dans le systkme avec le
fluide dgjit prOsent dans celui-ci. Le processus irrkversible de mdlange dans
Ca) est remplac~ par le transfert irreversible de chaleur entre l~coulernent
de fluide entrant clans le systkme et le fluide dkja present dans celui-ci.
Deux representations diff&entes d'un mdme modble sent prdsentOes clans
(b) et (c). Le modkle est valable pour les cas olt une petite de fluide entre
dans le systkme

The basis of the exergy analysis model for a fluid system Figure 8 Schematic representation of the set of models for the
irreversibility mechanisms within a fluid system. The equilibrium fluid
When a system like that shown in Figure 6 is not in a system is included within the system
steady state and not in equilibrium, there will not be Figure 8 Reprksentation schdmatique des groupe de modbles pour les
enough information to evaluate the S term of Equation m~canismes d'irr~versibilit~ it l'int~rieur d'un circuit. Le circuit it
l'~quilibre est inclus clans le systkme
(7), or Equation (5). However, an exergy analysis model
is proposed whereby the o6 term can be estimated. This
involves specifying quasi-steady models for each of the Analysis boundary -7
mechanisms of irreversibility that are assumed to occur. ! ~,=-pv
Each of these irreversibility models involves entropy
creation, but no entropy change rate. The sum of the _
i i
; ....... -,
entropy creation rates can be represented by the t e r m Scr" __5_ q---
According to the assumptions that are made, when the ',, 0o,, ,, i

models for heat transfer, throttling and mixing are

applied to a fluid system in a compressor there is no rh I"
U,V,T,p,m
,, m~
further irreversibility in the bulk fluid and it can then be
regarded as being in equilibrium. The reversible entropy i

change rate, Srev, for the equilibrium system that is " (Z :'
complementary to the irreversibility models, and con-
sistent with the first law analysis of the original system,
can then be evaluated.
The exergy analysis model thus consists of separate
components that can be used to calculate the entropy Figure9 A fluid system in equilibrium. The intensive thermodynamic
creation rates for each mechanism of irreversibility and state of all inlet and outlet flows is the same as that of the system itself.
the entropy change rate of a complementary equilibrium All heat transfer rates to or from the system occur at the temperature of
the system
system; the sum of these entropy creation and entropy
Figure 9 Un circuit it l'~quilibre. L'~tat thermodynamique de tousles
change r a t e s (o6cr -4:"o6rev) equals the entropy change rate Ocoulernents aux entrdes et aux sorties est similaire it celui du systbme lui-
of the non-equilibrium system. Alternatively, it will be m~rne. Tous[es taux de transfert de chaleur vers ou en provenance du
shown that the irreversibility models are sufficient in systkme se produisent it la temp&ature du systbme
 An exergy method for compressor performance analysis 427

3. Any heat transfer rate between a solid surface at the Table l Comparison of simulation results and experimental results
boundary and the equilibrium fluid system occurs, with a Tableau l Comparaison des r~sultats de la simulation avec ceux obtenus
finite temperature difference, across a boundary layer of lors des experiences
fluid that has negligible thickness and mass within the Experimental Simulation
system. The associated exergy destruction rate occurs in results results
the system but outside the equilibrium fluid system
Shaft power (W) 725 742
irrespective of the direction of heat flow. Indicated power (W) 674 694
4. The flow through any restriction is isentropic Mass flow rate (kg s -I) 0.013 0.0125
upstream of the throat. This is a standard assumption, Suction plenum exit temperature (K) 298 289
which underlies the orifice flow equation, in compressor Dsch. plenum exit temperature (K) 358 344
Crankcase temperature (K) 320 321
simulation. The irreversibility associated with throttling Energy to environment (W) 200 213
at a flow restriction therefore occurs downstream of the
restriction.
5. Any fluid that enters a system through a flow derived in Appendix 3:
restriction undergoes adiabatic throttling, followed by
irreversible heat transfer (as illustrated in Figure 7) with
the equilibrium fluid system. This irreversibility occurs (J T cpdO ( h T h i ) ) (10)
Jmx' = thiT° r; 0
within the system, but outside the equilibrium system.

This can also be written as

Expressions for exergy destruction mechanisms within a / . fT cpdO Omx"~

fluid system To(m,]T;T + T ) (ll
In this section a specific model is proposed for each type
of irreversibility that is assumed to occur within a fluid where
system of a compressor. Of necessity, these models will
be approximations, but they are of considerable practical Qmx = rib(hi - h) (12)
value. They must be fully consistent with the first- and
second-law analyses of the system.
The irreversibility rate due to heat transfer from a solid The equilibrium fluid system
mass at temperature T~ to a fluid system at temperature The relationship of the equilibrium fluid system to the
T is given by Equation (8); see, for instance, Szargut system and to the models for the irreversibility mechan-
et al 2° [Equation (1.33), p. 27)]. isms is shown in Figure 8. The equilibrium system is
shown separately in Figure 9. The first law analysis of the
(8) equilibrium fluid system (which transfers heat reversibly
at the temperature of the equilibrium fluid and receives
mass flow inputs at the intensive state of the equilibrium
As heat transfer only occurs from a higher to a lower fluid) is the same as that of the system with the analysis
temperature, T i will be smaller than T if the heat transfer boundary as shown in Figure 6.
rate is negative (i.e. from the fluid system). Therefore The rate of change of the equilibrium fluid system's
Equation (8) applies and has a positive value whether the entropy can be evaluated in terms of its changing
heat transfer rate is positive (to the fluid system) or thermodynamic properties by applying an entropy
negative (from the fluid system). balance, as the rate of entropy creation is zero:
For fluid flowing into the system, the exergy destruc-
tion rate due to adiabatic throttling is given as follows;
see, for instance, Kotas 18 [Equation (4.38), p. 112]:
i

]th, = Torih(s~," - si) (9) (13)

For exergy analysis purposes, therefore, the fluid
where s~, the specific entropy after throttling, is evaluated system shown in Figure 6 can be represented as shown in
at the pressure of the bulk fluid in the system and at Figure 8. Each of the devices shown that models an
ffi = hi. irreversibility mechanism has negligible mass, volume,
Just after it has been throttled, this fluid will be at a and entropy. Therefore the entropy change rate of the
temperature T,~ that may still differ from the temperature system within the overall analysis boundary equals that
of the fluid within the equilibrium system. The mixing of of the equilibrium fluid system. Subject to the assump-
the incoming fluid with the fluid already in the system tions underlying the irreversibility models, Equation (13)
when they have different temperatures is another cause also gives the entropy change rate for the fluid system
of irreversibility. According to the model represented in shown in Figure 6. There are thus two distinct, but fully
Figure 7 the entering fluid rejects or receives heat over a compatible, methods of evaluating the irreversibility or
temperature range, while the fluid in the equilibrium exergy destruction rates: the application of Equations (5)
system receives or rejects heat at a constant temperature. and (13) for the entire fluid system, or the summation of
Unless the entering fluid has the same temperature as the the exergy destruction rates for the different mechanisms.
fluid in the equilibrium system, there is exergy destruc- Both methods depend on the models for the irreversi-
tion. This is given by the following expression, which is bility mechanisms.
 Exergy destruction rate due to throttling Exergy destrn, rate d u e to int. convn.

1400

/ /
45C
1200 40C
356
1000
300

/
8OO 25C
6O0 200
150
400
>, 100
P 200
uJ 0
_J 50
0
0 45 90 135 180 225 270 315 360 45 90 135 180 225 270 315 360
Crank angle/[deg] Crankang~/[deg]

Exergy destruction rate due to friction Exergy destrn, rate due to condn.

/
250

200

150

100

50
/ ,v
0
0 45 90 135 180 225 270 315 360 45 90 135 180 225 270 315 360
Crank angle I[deg] Crank angle I[deg]

Exergy destruction rate due to mixing Exergy destrn, rate due to ext. conv.

5O 20
45
IRl'l _ I
18
40
^IVW~_I 16
35
30
Iv. " N 14
12
25 10

t
8
15 ~A 6
10 /I II ~1 4
5
~lV~l 2
0 - l 0
0 45 90 135 180 225 270 315 360 0 45 90 135 180 225 270 315 360
Crank angle I[deg] Crank angle/[deg]

Figure 10 Diagrams of the instantaneous rates of exergy destruction according to the mechanisms involved
Figure 10 Diagrammes des taux de destruction instantan~e d'exergie, selon les rnkcanismes coneern~s

Other exergy destruction rates friction work that is dissipated as heat transfer to a mass
Other causes of irreversibility that occur in the com- at temperature Ti is given as follows [e.g. see Kotas 18,
pressor are heat transfer between the lumped solid Equation (10)]:
masses, heat transfer to the environment, and friction.
Equations (14) and (15) respectively, which have /:f= ljVf(1 T/ -To
. .) (16)
the same form as Equation (8), are applicable for the
first two of these. The subscripts j and k refer to
solid masses.
The overall exergy destruction rate
Equations (5) and (8)-(16) can be used to evaluate the
•

ze,j~ = Qj~TO ~ - (14)

instantaneous exergy destruction rates for the entire
compressor or for particular subsystems. They can also
•
IQjo "
= QjoTO( 1~ - T~ ) (15) be used to obtain a detailed breakdown of the shaft
power wastage according to particular mechanisms or at
particular locations within the compressor. Breakdowns
The exergy destruction rate associated with a rate of of average exergy destruction rates and cumulative
 An exergy method for compressor performance analysis 429

Figure 11 shows the contributions of the average rates of

exergy destruction, by mechanism, to the overall shaft
power wastage. The most significant rate of exergy
destruction is that due to throttling, while the next most
significant mechanism is internal convection. The per-
Co n = , o n
Mixing 4.2% formance of this compressor could have been improved
5.7% significantly by increasing the areas of the suction and
discharge valve ports. This could have been done as these
/ Internal ports occupied only a small part of the area of the
Friction External convection cylinder head.
12.5% convection & 18.7% The steps necessary to optimize the exergy destruction
radiation
5.2%
due to internal convection are less obvious and would
require detailed consideration. An insulating cylinder
Figure 11 Breakdown of the contributions of the mechanisms of heat
transfer, friction, throttling, and mixing to the overall shaft power liner that had low heat capacity could reduce the exergy
wastage in the compressor destruction by this mechanism, as could any means of
Figure I I Re;partition des apports des m~canismes des transferts de reducing the internal convective heat transfer coefficient.
chaleur, de la friction, de l'~tranglement et du m~lange, h la perte glohale An example would be to reduce the swirl and turbulence
de puissance ~ l'arbre du compresseur within the cylinder. The situation is not as simple as this,
however, as these modifications would also increase the
work of compression and increase the discharge tem-
Compression
chamber 33.6% perature. The exergy transfer to the refrigerant stream
Suction Suction would be increased, but this might not be the desired
pipe \ ~ ~ plenum outcome in the context of a complete refrigeration
0.6% ~'~ ~aRb~ ~ I 8.7% system. The authors are in agreement with one of the
Discharge reviewers of this paper, who suggested that the methods
plenum - ~ External of Bejan 21 would be relevant to the problem. They intend
19.3% ~ ~ boundary to follow this line in their ongoing work.
layer While the simulated energy loss rate to the environ-
Discharge 5.2% ment was 213 W, the associated exergy destruction rate
pipe / \ Friction surfaces was just a small fraction of this: 18.3 W. Only 4.2% of the
1.1% / ~ 12.5% total exergy destruction was by the mechanism of
Crankcase Solidmasses conduction. However, the thermal conductivity and the
14.8% 4.2% heat capacity of the solid masses influence the exergy
destruction rates that occur due to heat transfer in the
Figure 12 Breakdown of the power wastage in the compressor by boundary layers of the fluid systems and at the outside
location surface of the compressor.
Figure 12 Re;partition sehm l~ndroit, de la perte d~nergie darts h' The exergy destruction rate due to friction is slightly
compresseur less than the friction work (as is evident from Equation
(16) where ~ is a little bigger than To). The difference,
while small, is due to the fact that the energy is dissipated
exergy destruction rates can also be obtained. The total within the compressor as heat transfer at a temperature
instantaneous exergy destruction rate can be written in other than that of the environment. This heat transfer
the form corresponds to an exchange, between the solid masses, of
potential to produce work by interacting with the
/: + ZJf (17) environment.
Figure 12 is a representation of the average exergy
Equation (17), without the friction term, can also be destruction rates of the compressor according to the
applied to a fluid system such as that shown in Figure 6 in locations where they occur.
order to evaluate its exergy destruction rate. It is shown The average rate of exergy transfer to the fluid was
in Appendix 4 that this is equivalent to applying 390 W, while the exergy destruction rate was 352 W. The
Equation (5).to the system and using Equation (13) to rational efficiency of the compressor was 52.6%.
evaluate the S term (see also Appendix 2).
Discussion
Analysis results Heat transfer defects are shown to be a significant
Table 1 compares the results of the simulation program portion of the total rational efficiency defect in a
with experimental results. There is quite good agreement compressor. It is therefore important that the mechan-
between these, although the values for the power and the isms by which these defects occur can be fully under-
discharge temperature differ somewhat. It is not within stood. The technique proposed in this paper can quantify
the scope of this paper to discuss the differences. The defects due to heat transfer. In contrast, methods that are
discrepancies are incidental, as the primary purpose of based only on the first law of thermodynamics do not
these simulation results is to demonstrate the new assign defects in compressor efficiency to the mechanism
technique. of heat transfer with finite temperature differences.
Figure 10 shows the instantaneous exergy destruction It would be difficult and costly to obtain by direct
rates according to the mechanisms that cause them. experiment all the information necessary to examine the
430 J. A. McGovern and S. Harte

mechanisms of shaft power wastage in detail. This would ConJerence at Purdue (Ed. J. F. Hamilton and R. Cohen) Purdue
include detailed data for temperatures, heat transfer University, W. Lafayette, IN, USA, (4 7 Aug 1986) 2 415 427
4 Sakaino, K. Some approaches towards a high efficiency rotary
rates, and mass flowrates within the compressor. A good compressor Proe 1984 International Compressor Engineering
practical solution is to use simulation techniques with Con/~'renee at Purdue (Ed. R. Cohen) Purdue University, W.
experimental verification of critical predictions such as Lafayette, IN, USA (11-13 Jul 1984) 315-322
the shaft power input, the mass flowrate and the 5 Stecco, S. S. Exergy analysis of compression and expansion pro-
discharge temperature. The irreversibility models pro- cesses Energy (1986) l l (6) 573 577
6 McGovern, J. A. Compressor rational efficiencies Developments
posed in this paper reduce to a practical minimum the in bMustrial Compressors, Proc lnst Mech Eng Conference 10
amount of data necessary for an exergy analysis that can (1989) 59-67
discriminate between different mechanisms and locations 7 Townsend, D. W. Second law analysis in practice The Chemical
of power wastage. Engineer (1980) Oct 628-633
8 Xu, X. Energy analysis on a vapor compression system using
R12, R134a and R290 as refrigerants Proc 1992 International
R¢f[rigeration Con/~'rence at Purdue (Ed. R. Tree and J. E.
Conclusions Braun) Purdue University (14-17 Jul 1992) 1 231-240
9 McGovern, J. A., Harte, S. Computer simulation of exergy
A method has been proposed to break down and destruction within a reciprocating compressor Proe 1992 Inter-
quantify the shaft power wastage mechanisms in a national Compressor EngineerhTg Conference at Purdue (Ed.
compressor. It is particularly suitable for application in J, F. Hamilton) Purdue University, W. Lafayette, IN, USA
(14 17 Jul 1992) I 309-316
computer simulations of compressors, and provides a 10 McGovern, J, A. Performance characteristics of a reciprocating
sound basis for design optimization. The method also compressor over a range of speeds Proc 1988 International Com-
fills a gap in the literature by allowing for well-founded pressor Engbwering ConJi'rence at Purdue (Ed. W. Soedel) Pur-
quantification of defects in shaft power utilization or due University, W. Lafayette, IN, USA (18-21 Jul 1988) 1
146 153
rational efficiency. The technique was applied to a 11 McGovern, J. A. Performance testing of an open reciprocating
reciprocating refrigeration compressor, and graphs of refrigeration compressor over a range of speeds Proc 1986 Inter-
the instantaneous rates of exergy destruction (power national Compressor Engineering Conjerence at Purdue (Ed. J. F.
wastage) were generated. Added insight was gained into Hamilton and R. Cohen) Purdue University, W. Lafayette, IN,
how heat transfer mechanisms increase the power USA (4 7 Aug 1986) 1 242-253
12 McGovern, J. A. On refrigerant compressors PhD thesis Univer-
consumption of a compressor per unit of exergetic sity of Dublin, Trinity College (1988)
output rate. 13 Kabelac, S. A simple set of equations of state for process cal-
The main limitation of the approach is that very culations and its application to R134a and R152a lnt J Re/'r~g
detailed information is required about the functioning of (1991) 14 217-222
14 Soedel, W. Introduction to Computer Simulation of Positive
a compressor in order to be able to prepare a detailed Displacement Type Compressors Short Course Notes, Purdue
and accurate exergy analysis. However, there is no University, W. Lafayette, IN, USA (1972)
competing technique that can give the same type of 15 Hamilton, J. F. Extensions q[ Mathematical Modelling of
detailed results more easily. Positive Displaeenwnt Type Compressors Short Course Notes,
Exergy analysis techniques have several advantages Purdue University, W. Lafayette, IN, USA (1974)
16 Lin, Z. Modelling temperatures in high speed compressors for
compared with conventional techniques for the analysis the purpose of gas pulsation and valve loss modelling Proc
of shaft power use in compressors. Among the advan- 1992 hlternational Compressor Engineering Conference at Purdue
tages are: (Ed. J. F. Hamilton) Purdue University, W. Lafayette, IN, USA
(14 17Jul 1992) I 1375 1384
17 Reetenwald, G. W. Prediction of heat transfer in compressor
1. Efficiency defects due to heat transfer can be cylinders Proc 1986 International Compressor Engineering Con-
calculated on the same basis as other defects such as ,[erenee at Purdue (Ed. J. F. Hamilton and R. Cohen) Purdue
those due to friction or flow throttling. University, W. Lafayette, IN, USA (4 7 Aug 1986) 1 159 174
2. The efficiency defects are not defined according to 18 Kotas, T. J. The Exergy Method of Thermal Plant Anah'sis
some hypothetical or somewhat arbitrary measure of Butterworths, London (1985)
19 Bejan, A. Advanced Engineering Thermodynamies John Wiley &
what the ideal work of compression should be, such as Sons, New York (1988)
the isentropic, polytropic, or isothermal work. Instead 20 Szargut, J., Morris, D. R., Steward, F. R. Exergy Analysis of
they are defined in terms of the minimum theoretical Thermal, Chemical, and Metallurgical Processes Hemisphere
work necessary to bring the fluid being compressed from Publishing Corporation, New York (1988)
21 Bejan, A. Entropy Generation Through Heat and Fluid Flow John
the actual suction state to the actual discharge state. Wiley & Sons, New York (1982)
3. The exergy destruction rates within a compressor
can be incorporated readily into an analysis of the whole
system, whether this is a refrigeration plant or, for Appendix 1 First-law analysis and mass balance
instance, a pneumatic power installation.
A first-law analysis and a mass balance is carried out on a
system, as shown in Figure 6, to determine how the
energy and the mass change over a small time interval.
References
The analysis is subject to the assumptions listed in the
1 Dossat, R. J. Principles of Refrigeration 2nd edn, SI Version, following section.
John Wiley & Sons, New York (1981)
2 FrancoDaCosta, C. M. Use of a simulation model for theoretical
optimisation analysis of a rolling-piston type compressor Proc
1986 International Compressor Engineering Conference at Purdue Assumptions
(Ed. J. F. Hamilton and R. Cohen) Purdue University, W.
Lafayette, IN, USA (4 7 Aug 1986) 3 824 839
1. It is recognized that the processes occurring within
3 Pandaya, P. N. A simplified procedure for designing hermetic the system can be highly irreversible, involving fluid
compressors Proe 1986 International Compressor Engineering throttling, mixing of incoming streams with the bulk
 An exergy method for compressor performance analysis 431

fluid, and heat transfer with significant temperature The new specific internal energy of the bulk fluid in the
differences. The system is not in thermodynamic system after a short time interval is thus given by
equilibrium. For the first-law analysis a temperature Equation (23) and the new mass by Equation (24):
and a pressure of the bulk fluid, Tb~lk and Pbulk
//bulkne,~ =//bulk "4- /)bulk A / (23)
respectively, are defined as the temperature and pressure
of a corresponding equilibrium system having the same mnew = m + t h a t (24)
values of U, V and m.
The new system volume is given by
2. It is assumed that a quasi-equilibrium state exists
over each portion of the system boundary where a fluid Vnew = V + VAt (25)
stream crosses. and the new specific volume of the bulk fluid can be
3. The gravitational potential energy and the kinetic determined as
energy of the fluid streams and the fluid in the system are
assumed negligible. ]/new
Vbulk.... -- (26)
4. It is assumed that there is good mixing of the fluid renew
within the system and that therefore all exit streams can The new bulk fluid temperature, Tbulk,0~, is the equili-
be assumed to have specific internal energy values and brium temperature of fluid that would have the new bulk
other thermodynamic properties that equal the bulk fluid fluid values for the specific internal energy and the
values. specific volume. By using the equation of state of the
5. It is assumed that the bulk fluid pressure is an fluid, an estimate of the new pressure of the bulk fluid
adequate estimate of the pressure on the local boundary can be made:
for evaluating the displacement work. Owing to the inlet
flow restrictions, there can be significant differences Pbulk .... = J'('l~bulk ...... Tbulkn~w) (27)
between the pressures at which the inlet fluid streams
cross the system boundary and the bulk fluid pressure.
6. Owing to the heat transfer resistance of the fluid Comments
boundary layer, there can be significant temperature According to the assumptions underlying the analysis,
differences between the surface of the containment and the new specific internal energy, specific volume,
the bulk fluid. temperature and pressure of the bulk fluid are predicted
7. It is assumed that the bulk fluid temperature is after a short time interval. These data also define the
an adequate estimate of the fluid temperature for thermodynamic state of the fluid in the exit streams.
evaluating the heat transfer rates between the fluid and However, the first-law analysis and mass balance do not
its containment. uniquely define a final equilibrium state of the system.
Because a state of equilibrium does not exist, the
Derivation of equations pressure, temperature, and other properties of the bulk
fluid cannot be regarded as equilibrium state properties
The energy balance takes the following form: of the system. The new mass, internal energy and volume
of the system are uniquely defined, but the entropy is not
v = + O, + w (l+) (see Appendix 2).
i C .]

Relationships for an ideal gas

If the local pressure that acts on the moving boundary
(the piston face) at each instant is approximately that of For an ideal gas it is convenient to evaluate the new
the bulk fluid, Puum, then temperature of the bulk fluid directly using Equations
(30)-(33), which are derived as follows:
W = --Pbulk ~" (19) hi - "bulk = l l i -- "bulk + PiVi = c , , ( T i - Tbulk ) + RTi
= %Ti - CeTbulk (28)
The specific internal energy of the system is equal to the
product of the mass within the system and the specific Similarly
internal energy of the bulk fluid, Ubum, and so he - Hbulk = c p T e - c,,Tbul k (29)
-- d(InUbulk ) Hence
dt - /Hhbulk -]- ////gbulk (20)

where, from the mass balance, /)bulk = I ~ i riti(cpTi - crTbulk) - Z e rhe(cpTe -

,i, = Z f i , , - Z fi''' (21)

i C

Therefore
/)bulk = [ Z rhi(cp ri - GTbulk)- Z rh,,(RTbulk)
Z rhi( hi -- '/hulk) -- Z t~le( he -- "bulk) q- ~ O/ -- Pbulk ~ k~
i e j

in
(22)
432 J. A. McGovern and S. Harte

J'bulk-- Ubulk (32)

Exergy destruction due to mixing
Cv Figure 7 illustrates the model used for the mixing of
incoming fluid with fluid in the system. The exergy
Tbulkne w = Tbulk + ]/'bulkAt (33) destruction rate due to mixing can be calculated using
Equation (10), which is derived as follows. For a small
Also temperature change 60 of the entering fluid, the heat
transfer rate from it is given by
V,~w = V + ~>At (25)
6Q = -ti~iCp60 (39)
Vnew
Vbulk.,~. -- (26) The corresponding exergy transfer rate from the stream is
mnew

Pbulk.~, - - RTbulkne~ (34)

~-l = -fiTicp60 ( 0 - 0 To) (40)

"Ubulk.~ and the exergy transfer rate to the system at temperature

T is

Appendix 2 Second law analysis 6~'2 = -Phict'60 ( T - T G) (41)

The entropy balance for the system shown in Figure 6 The rate of exergy destruction within the boundary layer
takes the following form: or thermal resistance where heat transfer occurs is given
by
S~--ZFhisi- ZtTleSeq-zQ-~i q-Scr (35) 5J = _-=~- ± - ,
i e .j •
= ti,,q,hOTo(~- 1 ) (42)
If the processes within the system are known to be
irreversible, the rate of entropy creation '~cr cannot be T
zero and must have a positive value. Given that the
specific entropy values of the exit streams can be
'mx: ;T~dJ : rhicpT° ,[Z' ( ~ - 1 ) d0
determined after each time step, there are two unknowns
in Equation (35): S~r and S. Given the first-law analysis
assumptions in Appendix 1, there is insufficient informa-
tion to evaluate both of the unknowns.
For a reversible system, where all streams enter at the =rhiTo((s-s;) (h~ hi)) (43)
same intensive thermodynamic state as the fluid in the
system and all heat transfer occurs at the temperature of For an ideal gas, this takes the form
the system, Equation (35) reduces to
]mx : riTTocp In Ti (44)

or Appendix 4 Derivation of the exergy destruction rate in a

fluid system from the individual irreversibility mechanism
l models
= srh + -~ Z Q (37)
The exergy destruction rate within a fluid system as
shown in Figure 6 is given in terms of the exergy
destruction mechanisms by
Appendix 3 Exergy analysis: further details ]: Z]Q~+ZJth +ZJmx, (45)
Expressions for the exergy destruction rates associated J i i
with fluid throttling, mixing and heat transfer have
already been given in the main paper. These apply for : To 9,. + -.;,.)
any real working fluid• T- i

Throttling of an ideal gas + ~//hi T; 0 T hi) (46)

For an ideal gas the throttling process causes no change
in the temperature or specific enthalpy. The increase in
specific entropy due to throttling can then be evaluated
directly from

-~-ZFhi(s--s:)--l~f Fhi(h--hi) (47)

 An exergy method for compressor performance analysis 433

1--(~i QI"}"ZQmx*)
i = S+ ZrrteS-
e i (51)

Hence, substituting Equation (51) into Equation (49),

1 ~ ri, i(h _ hi)] (48)
T i

=7" 1 e !

+ ~Im,/si - s,)l } (52)

(49)
i )
and as se = s
But as the equilibrium fluid system shown in Figure 8
contains no irreversibility,
(5)
S+~liles-Zt~,is-
¢' I
l(v0 ]q.-ZQmx
1
i" ) =0

(50)