Plant Engineering 2 Compressors
The Perfect Gas Model
Virtually all chemical engineering plant design is based on thermodynamic data. Sometimes
we have tables and charts of data and sometimes we have to use calculation methods. For
many bulk commodity chemicals, accurate data have been obtained in the laboratory for use
by producers, pipeline companies, end users and also the customs and excise. For the last
purpose, the properties of pipeline methane have to be known to within sixfigure accuracy.
If a process produces a commodity chemical at a price of a few tens of pence per kilogram, then
the costings for heat transfer processes etc will be required at the 0.01 pence/kg level. Only
experimental data provide such accuracy. For new processes and for evaluation studies, data
for organic substances are usually in very short supply and so we need calculation methods. For
gases, the initial starting point is the perfect gas model (also known as the ideal gas model).
What is a Model?
There are two approaches:
1. Make extensive experimental measurements and then find an equation to fit them.
eg PV = nRT fits the behaviour of thousands of gases and vapours.
2. Propose either a physical or purely mathematical model, derive performance equations
and then test them experimentally. Both the classical kinetic theory and the quantum theory
can be developed to provide the equations of thermodynamics. It is not obligatory to use a
physical
model and purely mathematical statements of the second law of thermodynamics, such
H
as dQ/T, can be used as the starting point.
In the PERFECT GAS MODEL, gas molecules undergo perfectly elastic collisions with
the walls of the enclosure. Gas pressure arises as an outward force due to the conservation
of momentum in the collisions. The further basis is that the gas is sufficiently dilute that the
molecules are far enough apart for us to neglect van der Waals forces (induced dipole effects). As
a consequence, we can expect that there will be no changes in u or h as the pressure changes.
This postulate of the model can be expressed mathematically as:
u
P
= 0 and
h
P
= 0 Applies for real gases as P 0 and v .
T
These equations can also be derived from the second law.
By definition, h = u + P v and since P v = RT (using pg model) then h = u + RT .
So differentiating with respect to T,
dh
dT
du
dT
+R
For perfect gases, h and u are both functions only of temperature and NOT pressure (or
volume), so
!
u
du
cv =
=
T v dT
and
h
T
cp =
Therefore
cp = cv + R
dh
dT
=
P
for a perfect gas.
An average constant specific heat capacity is often assumed for perfect gases. This gives for
monatomic gases cp = 52 R, for diatomic gases cp = 27 R and for other gases cp = 29 R. The ratio of
the heat capacities is defined as = cp /cv which has a value of about 1.67 for monatomic
gases, 1.4 for diatomic gases and 1.3 for most other perfect gases.
1.1
Isothermal Compression/expansion
T = 0 (isothermality) and thus u = cv dT = 0 (closed system) or h = cp dT = 0 (open
system)
Therefore,
u = 0 = q + w and q = w (closed system)
q = ws (open system)
R
Therefore q = w =
R2
R2
1
R2
dv
1 v
P dv = RT
R2
dP
1 P
= RT ln
= RT ln
or q = ws = 1 vdP = RT
removed in isothermal compression.
P2
P1
v1
v2
= RT ln
P2
P1
(closed system)
(open system). Note that heat must be
P2
w (or ws ) = RT ln
P1
The above is the isothermal work equation for gas compression in closed or open systems.
1.2
Adiabatic Compression/expansion (q = 0)
Under adiabatic conditions (q = 0), the first law (closed system) becomes
u = w
Z
cv dT =
P dv
Substituting P = RT/v (pg model) in the above and rearranging:
R dv
dT
=
T
cv v
Rearranging cp = cv + R as
cp
cv
=1+
R
cv
we get:
dT
dv
= ( 1)
T
v
Integrating both sides:
T2
ln
T1
v2
= ( 1) ln
v1
= ln
v1
v2
(1) !
Hence we get a T v relationship for q = 0.
T2
v1
=
T1
v2
(1)
Equ(1)
Rearranging equation (1) to give
1
1
( 1
)
v1
T2 ( 1 )
T1
=
=
T1
v2
T2
and substituting into
P1 v1
= PT22v2 (pg model) we obtain the required T P relationship:
T1
( 1
)
T2
P2
=
T1
P1
From equations (1) and (2),
power
,
(1)
So, finally:
v1
v2
P2
P1
v1
v2
(1)
1
P2
P1
P v = constant
Equ(2)
. Raising both sides of this equation to the
Equ(3)
The equations (1) (3) above apply only to adiabatic changes with Q = 0
1.3
Work of Gas Compression
When considering the compression of a flowing gas, in order to define enthalpy, we found:
w = ws + P1 v1 P2 v2
(All work done on an element)
But for a closed system, dW = F d = (P A)d = P dV and hence w = P dv
R
Therefore ws = P dv + P2 v2 P1 v1 = P dv + (P v) = P dv + vdP + P dv
R
ws =
w=
vdP
(Open system)
P dV
(Closed system)
In a flow compressor, the work is positive (in) and P is positive.
In a turbine, which is an expander, power is taken out of the system (W ve) and there is a
pressure drop (P2 < P1 ).
Integration of the above gas equations may not be too easy but for a liquid the volume can be
assumed constant during compression and:
ws =
vdP = v
dP = v(P2 P1 )
From the above we can now derive the work equations for adiabatic compression.
3
With q = 0, from the first law we get:
w = u = cv T =
Therefore
R
RT2 RT1
T =
( 1)
( 1)
"
P1 v1 (P2 v2 )
P2 v2 P1 v1
=
1
w=
( 1)
( 1) (P1 v1 )
But P1 v1 = RT1 and from Equ (3),
v2
v1
P1
P2
( 1 )
P2
P1
( 1 )
, so, finally
(1)
RT1 P2 ( )
w=
1
( 1)
P1
This is the equation for the work of adiabatic compression for a closed system as P1 P2 ,
starting with a temperature T1 . T2 must be found from the appropriate change equation, (2),
if required.
To find the work of adiabatic compression for an open system we need to integrate ws =
using P v = constant.
h = ws = cp T =
Therefore
The result is:
Hence
(RT2 RT1 )
R
T =
( 1)
( 1)
"
(P2 v2 P1 v1 )
P1 v1 (P2 v2 )
ws =
=
1
( 1)
( 1) (P1 v1 )
(1)
RT1 P2 ( )
1
ws =
( 1)
P1
ws (Flow) = w (Closed)
vdP
Compressors and Turbines
Compressors and turbines fall into two broad classes: positive displacement and dynamic, as
shown in the Figure below (from Bett, Rowlinson and Saville). From the thermodynamic point
of view these differ in one important respect. Positive displacement devices operate by gas at
the suction pressure entering an enclosed space which then undergoes expansion/contraction
as a closed system followed by release of the gas at the discharge pressure. Dynamic equipment
has the expansion/contraction occurring in an open system.
Compressors
Positive Displacement
Dynamic
Jet
Rotary
Turbo
Radial flow
(centrifugal)
Axial flow
Single
Rotor
Diaphragm
Reciprocating
Twin
Rotor
Turbines
Positive Displacement
Dynamic
Radial flow
Inward
Outward
Axial flow
Impulse
Reaction
Fig 5.10 Classification of compressors and expanders, based on operating principles
2.1
Positive Displacement Compressors/Expanders
It should be noted that although we confine our attention here to one example of a compressor,
the same result will apply for all compressors acting on similar principles and to positive
displacement expanders, except that in the case of expanders the flow of energy is reversed
(work is done by the system).
2.1.1
Reciprocating Piston Compressors
This is essentially a piston-in-cylinder device with non-return valves in the inlet and outlet
pipes from the cylinder. The piston is driven by an electric motor via an eccentrically mounted
rod.
P
P
2
2
adiabatic
expansion
P
1
P
2
Vclearance
adiabatic
compression
/ expansion
adiabatic
compression
P
1
4
Volume
2 (3)
1 (4)
Molar Volume
The figure above shows the pressurevolume relation for an ideal compressor in which all
operations are carried out reversibly. The Pv diagram is also shown. It should be noted that
the thermodynamic state of the gas during suction and discharge remains constant so that the
adiabatic expansion of the gas remaining in the compressor after discharge follows the same
path as the adiabatic compression of the gas. However, the amount of gas is different for these
two adiabatic changes.
Stage 4 1 n2 n1 moles of gas are drawn into the cylinder through the suction valve at
constant thermodynamic state (state 1) where they mix with the n1 moles remaining from
the previous cycle.
Stage 1 2 the n2 moles of gas in the cylinder is compressed adiabatically from state 1 to 2
with both valves closed.
Stage 2 3 n2 n1 moles of gas (state 2) are expelled through the discharge valve.
Stage 3 4 the remaining n1 moles of gas trapped in the clearance volume expand
adiabatically from state 2 to 1 with both valves closed.
The work done is the sum of the work done in each stage, ie
Wcycle = W12 + W23 + W34 + W41
where
W12 =
P dV = n2
P dv
W23 = P2 (V3 V2 ) = P2 (n1 v2 n2 v2 )
Z
W34 =
P dV = n1
P dv
W41 = P1 (V1 V4 ) = P1 (n2 v1 n1 v1 )
so
Wcycle = n2
P dv P2 (n1 v2 n2 v2 ) n1
P dv P1 (n2 v1 n1 v1 )
Since points 2 and 3 are are Rat state 2R and 1 and R4 are at state 1 and both the compression
and expansion are adiabatic 34 P dv = 21 P dv = 12 P dv and
Wcycle = n2
Wcycle = (n2 n1 )
P dv P2 (n1 v2 n2 v2 ) + n1
P dv P1 (n2 v1 n1 v1 )
P dv + (n2 n1 )(P2 v2 P1 v1 ) = (n2 n1 )((P v)12
Z
vdP +
Wcycle = (n2 n1 )
Wcycle = (n2 n1 )(
P dv
P dv)
P dv)
vdP
This is the same as for a flow device with (n2 n1 ) moles processed per cycle. The power
requirement is given by the product of Wcycle and the frequency of the cycles,
P ower = Wcycle = (n2 n1 )
vdP = (n2 n1 )
P1 v1
( 1)
( (1)
)
P2
P1
But, (n2 n1 )v1 is the volume of gas taken in at the intake conditions, so (n2 n1 ) = f is
the molar flowrate of the gas and f v1 is the volumetric flowrate of gas at the inlet conditions
VI and
(1)
( (1)
)
P
P1 v1 P2 ( )
P
2
1
P ower = f
1 = VI
1
( 1)
P1
( 1)
P1
or
P ower =
( (1)
)
P2
f cp T1
P1
2.2
Dynamic Compressors/Expanders
These are generally rotary devices for momentum transfer. The most elementary domestic
rotary compressor is the hair dryer. Commercially, very large compressors operate at up to 250
atmospheres. A large turbine consists of a rotating disk on a shaft carrying the turbine blades.
High pressure gas undergoes expansion in a ring of nozzles.
The analysis of these devices is very simple if we apply the first law. We can ignore changes
in potential energy for horizontal flow. Considering an element of expanding gas in a nozzle,
then both Q and W are zero, because the element is isolated during its passage through the
nozzle. So:
E = H + (KE) + (P E) = Q + W = 0
The increase in kinetic energy is equal to the decrease in enthalpy. Subsequently the high
velocity gas impinges on the turbine blades and transfers momentum from the gas to the
P
blades. ( mu = constant for the whole turbine). According to the first law, the gas kinetic
energy is now lost with the development of shaft work in the turbine. Overall we have very
simply:
Z
h2 h1 = ws =
or
vdP
(1)
P1 v1 P2 ( )
f h = P ower = f
1
( 1)
P1
as before. Clearly (h2 h1 ) is negative and so the negative value of Ws represents the export
of work. A gas compressor works in reverse.
In power generation, the high pressure fluid is superheated steam. It must expand to low
pressure without forming liquid water, which would damage the turbine blades. The design of
these machines is often very easily performed on a property chart as we shall see later.
The following diagram compiled from figures 1.10 and 1.11 in Bett, Rowlinson and Saville,
shows a schematic view of the workings of a turbine.
2.3
Polytropic Compression/expansion
In general the process path is not known (neither isothermal nor adiabatic) with both heat and
work being transferred. We consider, for the moment a reversible change. A path along which
both heat and work is transferred with constant heat capacity is termed a polytropic path. One
representation of a reversible polytropic path is to put P v n = constant. Here the polytropic
coefficient n must be found from experiment. n = ccvp c
where c is the apparent heat capacity
c
along the path (ie Q = cT ). c is negative.
2.4
Compressor/Turbine Efficiencies
In practice the compression/expansion operation is not carried out reversibly. Mismatches
between pressures at the point of valve opening, leakage and fluid friction all lead to internal
irreversibility. In addition there are external mechanical irreversibilities in the moving parts
of the turbine, pistons or motor. These all reduce the available useful work from turbines and
increase the input work requirement of compressors.
We can define the efficiency of a turbine or compressor in relation to a variety of ideal or
reversible processes.
Compressor
Turbine
Isothermal Efficiency
isothermal work requirement
actual work requirement
actual work output
isothermal work output
Isentropic Efficiency
isentropic work requirement
actual work requirement
actual work output
isentropic work output
Polytropic Efficiency
polytropic work requirement
actual work requirement
actual work output
polytropic work output
2.5
Multistage Compression
The work of compression of a gas is easily calculated from the change in properties. Starting
from any given initial state, it is evident that the work of adiabatic compression is greater than
the work of isothermal compression because in the latter the work is inevitably diminished
by heat removal to maintain isothermality. For the case of isothermal compression, the work
equations are easily integrated, and the same result is obtained for both open and closed
systems.
w=
P dv = RT1
dv
v1
= +RT ln
v
v2
P2
= RT1 ln
P1
(closed system)
dP
P2
(open system)
= RT1 ln
P
P1
In the diagram below, isothermal compression is compared with adiabatic compression.
Remember that the work of isothermal compression is represented by the area between the
v = 0 axis and the compression curve. For an adiabatic flow process, we found the compression
work to be:
!
( 1
)
P
R
2
1
T1
ws =
( 1)
P1
ws =
vdP = RT1
All of the above work quantities relate to one mole of gas; in general they must be multiplied
by the amount of the flow, f mol/s. Isothermal compression is an impractical ideal, but the
10
isothermal work can be approached by using staged compression with interstage cooling.
Adiabatic
compression
6
Intercoolers
5
Isothermal
compression
T1
Adiabatic
compression
stages
1
v
v
Work of isothermal compression
Consider a series of reversible and adiabatic compression stages with a constant suction
temperature, T1 .
T2
2
P
2
T1
P2
Discharge
6
T4 P
4
Suction
1
P T
1 1
T3
T1
P
3
P
3
Intercoolers
WT otal =
X
i
1
( 1
( 1
)
)
P2
n
P3
PN +1 ( )
(RT1 )
Wsi =
+
+ ...
N
( 1)
P1
P2
PN
To find the optimal interstage pressure ratio, the work must be minimised by:
Ws
dPi
=0
all other Pj
11
!( 1 )
!
!
1
1
1
1
Ws
1
1
n
1
)
(
(Pi )
(Pi+1 )( ) (Pi )( +1)
RT1
=
Pi Pj
( 1)
Pi1
All other terms vanish. Thus the required minimum is :
! 1
1
Pi+1 ( )
RT1 Pi ( )
Ws
=0=n
Pi Pj
Pi
Pi1
Pi
i
This can only be zero when PPi1
= PPi+1
. Hence the optimal condition requires that the ratios
i
be the same for all stages (any value of i).
P2
P1
P3
P2
P4
PN +1
......
P3
PN
PN +1
= (Pressure Ratio)N
P1
Use a pressure ratio for all stages given by:
1
PN +1 ( N )
Pressure Ratio =
P1
This result for the design of multistage compressors, which are the norm, is easy to apply.
i.e. For a multistage compressor with constant inlet temperature, Tin , to each stage, constant
pressure ratio, P r, per stage and with adiabatic compression in each stage, if N is the number
of stages and f is the gas throughput, then
Power = f
RTin N (P r)( ) 1
( 1)
The design can also be done graphically on a Ts chart. We shall soon show that
R
R
dS = dQ/T . Therefore if Q=0 for a compression then the process is also isentropic. h2 h1 =
ws which can be plotted as a vertical line on the chart.
12
