Flow Measurement and Instrumentation 18 (2007) 18–26

Flow Measurement
and Instrumentation
www.elsevier.com/locate/flowmeasinst
A procedure for the calculation of the natural gas molar heat capacity, the
isentropic exponent, and the Joule–Thomson coefficient
Ivan Mari´ c

Ru ¯ der Boškovi´ c Institute, Division of Electronics, Laboratory for Information Systems, Bijeniˇ cka c. 54, P.O.B. 180, 10002 Zagreb, Croatia
Received 22 March 2006; received in revised form 19 September 2006; accepted 14 December 2006
Abstract
A numerical procedure for the computation of a natural gas molar heat capacity, the isentropic exponent, and the Joule–Thomson coefficient has
been derived using fundamental thermodynamic equations, DIPPR AIChE generic ideal heat capacity equations, and AGA-8 extended virial-type
equations of state. The procedure is implemented using the Object-Oriented Programming (OOP) approach. The results calculated are compared
with the corresponding measurement data. The flow-rate through the orifice plate with corner taps is simulated and the corresponding error due to
adiabatic expansion is calculated. The results are graphically illustrated and discussed.
c 2007 Elsevier Ltd. All rights reserved.
Keywords: Heat capacity; Joule–Thomson coefficient; Isentropic exponent; Natural gas; Flow measurement
1. Introduction
When a gas is forced to flow through a differential device
(see Fig. 1) it expands to a lower pressure and changes its
density. Flow-rate equations for differential pressure meters
assume a constant density of a fluid within the meter. This
assumption applies only to incompressible flows. In the case of
compressible flows, a correction must be made. This correction
is known as the adiabatic expansion factor, which depends
on several parameters including the differential pressure, the
absolute pressure, pipe diameter, the differential device bore
diameter, and the isentropic exponent. The isentropic exponent
has a limited effect on the adiabatic correction factor, but needs
to be calculated if accurate flow-rate measurements are needed.
When a gas expands through a restriction to a lower
pressure, it also changes its temperature. This process occurs
under the conditions of constant enthalpy and is known as
Joule–Thomson expansion [1]. It can also be considered as an
adiabatic effect, because the pressure change occurs too quickly
for significant heat transfer to take place. The temperature
change is related to pressure change and is characterized by
the Joule–Thomson coefficient. The temperature drop increases

Tel.: +385 1 4561191; fax: +385 1 4680114.
E-mail address: ivan.maric@irb.hr.
with the increase of the pressure drop and is proportional to
the Joule–Thomson coefficient. According to [2], the upstream
temperature is used for the calculation of flow-rates, but
the temperature is preferably measured downstream of the
differential device. The use of downstream instead of upstream
temperatures may cause a flow-rate measurement error due to
the difference in the gas density caused by the temperature
change. Our objective is to derive the numerical procedure for
the calculation of the natural gas specific heat capacity, the
isentropic exponent, and the Joule–Thomson coefficient that
can be used to compensate for the adiabatic expansion effects
in real-time flow-rate measurements.
2. Procedure
This section outlines the procedure for the calculation
of specific heat capacities at a constant pressure c
p
and
at a constant volume c
v
, the Joule–Thomson coefficient
µ
JT
, and the isentropic exponent κ of a natural gas based
on thermodynamic equations, AGA-8 extended virial type
characterization equations [3,4], and DIPPR generic ideal heat
capacity equations [5]. First, the relation of the molar heat
capacity at constant volume to the equation of state will be
derived. Then the relation will be used to calculate a molar
heat capacity at constant pressure, which will then be used
0955-5986/$ - see front matter c 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.flowmeasinst.2006.12.001
I. Mari´ c / Flow Measurement and Instrumentation 18 (2007) 18–26 19
Fig. 1. Schematic diagram of the gas flow-rate measurement using an orifice plate with corner taps.
for the calculation of the Joule–Thomson coefficient and the
isentropic exponent. The total differential for entropy, related
to temperature and molar volume [6], is:
ds =

∂s
∂T

v
m
dT +

∂s
∂v
m

T
dv
m
, (1)
where s denotes entropy, T denotes temperature, and v
m
is a
molar volume of a gas. By dividing the fundamental differential
for internal energy du = T · ds − p · dv
m
by dT while holding
v
m
constant, the coefficient of dT in Eq. (1) becomes c
m,v
/T
since the molar heat at constant volume is defined by c
m,v
=
(∂u/∂T)
v
m
. The Maxwell relation (∂s/∂v
m
)
T
= (∂p/∂T)
v
m
,
is used to substitute the coefficient of dv
m
. Finally, the Eq. (1)
becomes:
ds =
c
m,v
T
dT +

∂p
∂T

v
m
dv
m
. (2)
Similarly, starting from a total differential for entropy related to
temperature and pressure [6] ds = (∂s/∂T)
p
dT+(∂s/∂p)
T
dp
and by dividing the fundamental differential for enthalpy dh =
T · ds +v
m
· dp by dT while holding p constant, the coefficient
of dT in the total differential becomes c
m, p
/T, since the
molar heat capacity at constant pressure is defined by c
m, p
=
(∂h/∂T)
p
. The Maxwell relation (∂s/∂p)
T
= (∂v
m
/∂T)
p
is used to substitute the coefficient of dp and the following
relation is obtained:
ds =
c
m, p
T
dT +

∂v
m
∂T

p
dp. (3)
Subtracting Eq. (2) from Eq. (3), and then dividing the resulting
equation by dv
m
while holding p constant, and finally inverting
the partial derivative (∂T/∂v
m
)
p
, the following equation is
obtained:
c
m, p
− c
m,v
= T

∂v
m
∂T

p

∂p
∂T

v
m
. (4)
The total differential of the thermodynamic property
Eqs. (2) and (3) must be the exact differential, i.e. the order
of forming the mixed second derivative is irrelevant. The par-
tial derivative of the first coefficient with respect to the second
variable equals to the partial derivative of the second coefficient
with respect to the first variable. By applying this property to
Eq. (2) and by assuming T to be the first variable with the cor-
responding coefficient c
m,v
/T, and v
m
the second variable with
the corresponding coefficient (∂p/∂T)
v
m
, we obtain:

∂c
m,v
∂v
m

T
= T


2
p
∂T
2

v
m
. (5)
The Eq. (5) can be rewritten in the following integral form:
c
m,v
= c
m,vI
+ T

v
m
v
mI
→∞(T=const)


2
p
∂T
2

v
m
dv
m
(6)
where c
m,vI
, v
mI
, and ν
m
denote the ideal molar heat capacity
at constant volume, the corresponding molar volume of the
ideal, and the real gas at temperature T respectively. Real gases
behave more like ideal gases as the pressure approaches zero or
v
mI
→ ∞. After substituting v
m
= 1/ρ
m
, p = RT Zρ
m
and
c
m,vI
= c
m, pI
− R, the Eq. (6) transforms to:
c
m,v
= c
m, pI
− R − RT
×

ρ
m
ρ
mI
→0(T=const)
1
ρ
m

2

∂ Z
∂T

ρ
m
+ T


2
Z
∂T
2

ρ
m


m
(7)
where c
m, pI
denotes the temperature dependent molar heat
capacity of the ideal gas at constant pressure, R is the universal
gas constant, Z is the compression factor, and ρ
mI
and ρ
m
are
the corresponding molar densities of the ideal and the real gas
at temperature T. After substituting the first and the second
derivative of the AGA-8 compressibility equation [4],
Z = 1 + Bρ
m
− ρ
r
18
¸
n=13
C

n
+
58
¸
n=13
C

n

b
n
− c
n
k
n
ρ
k
n
r

ρ
b
n
r
e
−c
n
ρ
kn
r
(8)
into the Eq. (7) and after integration we obtain
c
m,v
= c
m, pI
− R + RTρ
r
(2C
0
+ TC
1
− C
2
) (9)
20 I. Mari´ c / Flow Measurement and Instrumentation 18 (2007) 18–26
with
C
0
=
18
¸
n=13
C

n

B

K
3
(10)
C
1
=
18
¸
n=13
C

n

B

K
3
(11)
C
2
=
58
¸
n=13

2C

n
+ TC

n

ρ
b
n
−1
r
e
−c
n
ρ
kn
r
(12)
where ρ
r
is the reduced density (ρ
r
= K
3
ρ
m
), B is the
second virial coefficient,
¸
C

n
¸
are the temperature dependent
coefficients, K is the mixture size parameter, while {b
n
}, {c
n
},
and {k
n
} are the parameters of the equation of state. The mixture
size parameter K is calculated using the following equation [4]:
K
5
=

N
¸
i =1
y
i
K
5/2
i

2
+2
N−1
¸
i =1
N
¸
j =i +1
y
i
y
j
(K
5
i j
− 1)(K
i
K
j
)
5/2
, (13)
where y
i
denotes the molar fraction of the component i , while
{K
i
} and
¸
K
i j
¸
are the corresponding size parameters and the
binary interaction parameters given in [4]. According to [4],
the second virial coefficient is calculated using the following
equation:
B =
18
¸
n=1
a
n
T
−u
n
N
¸
i =1
N
¸
j =1
y
i
y
j
B

ni j
E
u
n
i j
(K
i
K
j
)
3/2
(14)
and the coefficients
¸
B

ni j
¸
,
¸
E
i j
¸
and
¸
G
i j
¸
are defined by
B

ni j
= (G
i j
+ 1 − g
n
)
g
n
(Q
i
Q
j
+ 1 − q
n
)
q
n
×(F
1/2
i
F
1/2
j
+ 1 − f
n
)
f
n
· (S
i
S
j
+ 1 − s
n
)
s
n
(W
i
W
j
+ 1 − w
n
)
w
n
, (15)
E
i j
= E

i j
(E
i
E
j
)
1/2
(16)
and
G
i j
= G

i j
(G
i
+ G
j
)/2, (17)
where T is the temperature, N is the total number of gas
mixture components, y
i
is the molar fraction of the component
i , {a
n
}, { f
n
}, {g
n
}, {q
n
}, {s
n
}, {u
n
}, and {w
n
} are the parameters
of the equation of state, {E
i
}, {F
i
}, {G
i
}, {K
i
}, {Q
i
}, {S
i
}, and
{W
i
} are the corresponding characterization parameters, while
¸
E

i j
¸
and
¸
G

i j
¸
are the corresponding binary interaction
parameters. The main symbols and units are given in Table 1.
For additional symbols and units refer to ISO-12213-2 [4]. The
temperature dependent coefficients
¸
C

n
; n = 1, . . . , 58
¸
and
the mixture parameters U, G, Q, and F are calculated using
the equations [4]:
C

n
= a
n
(G + 1 − g
n
)
g
n
(Q
2
+ 1 − q
n
)
q
n
×(F + 1 − f
n
)
f
n
U
u
n
T
−u
n
, (18)
Table 1
Symbols and units (for additional symbols and units refer to ISO-12213-2 [4])
Symbols and units
Symbol Description Unit
B Second virial coefficient m
3
∗ kmol
−1
B

ni j
Mixture interaction coefficient –
C Coefficient of discharge –
c
m, p
Molar heat capacity at constant pressure J/(mol K)
c
m,v
Molar heat capacity at constant volume J/(mol K)
C

n
Temperature and composition dependent
coefficients

c
n
AGA-8 equation of state parameter –
c
p
Specific heat capacity at constant pressure J/(kg K)
c
m, pI
Ideal molar heat capacity of the natural gas mixture J/(mol K)
C
j
m, pi
Ideal molar heat capacity of the gas component j J/(mol K)
D Upstream internal pipe diameter m
d Diameter of orifice m
h Specific enthalpy J/kg
K Size parameter –
p Absolute pressure Pa
q
m
Mass flow-rate kg/s
R Molar gas constant 8314.51 J/(kmol K)
s Specific entropy J/(kg K)
T Absolute temperature K
v
m
Molar specific volume m
3
/kmol
v
mI
Molar specific volume of ideal gas m
3
/kmol
y
i
Molar fraction of i -th component in gas mixture –
Z Compression factor –
β Diameter ratio d/D –
p Differential pressure Pa
ω Pressure loss Pa
κ Isentropic exponent –
µ
JT
Joule–Thomson coefficient K/Pa
ρ
m
Molar density kmol/m
3
ρ
mI
Molar density of ideal gas kmol/m
3
ρ
r
Reduced density –
U
5
=

N
¸
i =1
y
i
E
5/2
i

2
+2
N−1
¸
i =1
N
¸
j =i +1
y
i
y
j
(U
5
i j
− 1)(E
i
E
j
)
5/2
, (19)
G =
N
¸
i =1
y
i
G
i
+ 2
N−1
¸
i =1
N
¸
j =i +1
y
i
y
j
(G

i j
− 1)(G
i
+ G
j
), (20)
Q =
N
¸
i =1
y
i
Q
i
, (21)
and
F =
N
¸
i =1
y
2
i
F
i
, (22)
where, U
i j
is the binary interaction parameter for mixture
energy. The first and the second derivatives of the coefficients
B and C

n
, with respect to temperature, are:
B

= −
18
¸
n=1
a
n
u
n
T
−u
n
−1
N
¸
i =1
N
¸
j =1
y
i
y
j
B

ni j
E
u
n
i j
(K
i
K
j
)
3/2
(23)
I. Mari´ c / Flow Measurement and Instrumentation 18 (2007) 18–26 21
Table 2
The DIPPR/AIChE gas heat capacity constants
Natural gas component DIPPR/AIChE ideal gas heat capacity constants
a b c d e
Methane — CH
4
33 298 79 933 2086.9 41 602 991.96
Ethane — C
2
H
6
40 326 134 220 1655.5 73 223 752.87
Propane — C
3
H
8
51 920 192 450 1626.5 116 800 723.6
B

=
18
¸
n=1
a
n
u
n
(u
n
+ 1)T
−u
n
−2
×
N
¸
i =1
N
¸
j =1
y
i
y
j
B

ni j
E
u
n
i j
(K
i
K
j
)
3/2
(24)
C

n
= −
u
n
T
C

n
(25)
C

n
= −
u
n
+ 1
T
C

n
. (26)
The ideal molar heat capacity c
pI
is calculated by
c
m, pI
=
N
¸
j =1
y
j
c
j
m, pi
(27)
where y
j
is the molar fraction of component j in the gas
mixture and C
j
m, pi
is the molar heat capacity of the same
component. The molar heat capacities of the ideal gas mixture
components can be approximated by the DIPPR/AIChE generic
equations [5], i.e.
c
j
m, pi
= a
j
+ b
j

c
j
/T
sinh(c
j
/T)

2
+ d
j

e
j
/T
cosh(e
j
/T)

2
,
(28)
where c
j
m, pi
is the molar heat capacity of the component j of the
ideal gas mixture, a
j
, b
j
, c
j
, d
j
, and e
j
are the corresponding
constants, and T is the temperature. The constants a, b, c, d,
and e for some gases are shown in Table 2.
The partial derivative of pressure with respect to temperature
at constant molar volume, and the partial derivative of molar
volume with respect to temperature at constant pressure, are
defined by the equations:

∂p
∂T

v
m
= Rρ
m
[Z + T(C
3
− ρ
r
C
0
)] (29)
and

∂v
m
∂T

p
=
R
p
¸
Z +

∂ Z
∂T

p
T

(30)
where,
C
3
=
58
¸
n=13

C

n
D

n

, (31)
D
n
= (b
n
− c
n
k
n
ρ
k
n
r

b
n
r
e
−c
n
ρ
kn
r
, (32)

∂ Z
∂T

p
=
R(T Z)
2
C
3
− pZ[T K
3
C
0
+ C
4
]
R(T Z)
2
+ pTC
4
, (33)
C
4
= C
5
+
58
¸
n=13
C

n
D
1n
(34)
C
5
= B − K
3
18
¸
n=13
C

n
(35)
and
D
1n
= K
3
[b
2
n
− c
n
k
n
(2b
n
+ k
n
− c
n
k
n
ρ
k
n
r

k
n
r
]
×ρ
b
n
−1
r
e
−c
n
ρ
kn
r
. (36)
The isentropic exponent is defined by the following relation
κ = −
c
m, p
c
m,v

∂p
∂v
m

T

v
m
p

= −
c
m, p
c
m,v
ρ
m
p

∂p
∂v
m

T
, (37)
where

∂p
∂v
m

T
=

∂p
∂ρ
m

T

∂ρ
m
∂v
m

T
= −RTρ
2
m
(Z + ρ
m
C
4
) . (38)
The Joule–Thomson coefficient is defined by the following
equation [2]:
µ
JT
=
RT
2
pc
m, p

∂ Z
∂T

p
. (39)
The derivation of the Eq. (39) is given in [6] and [7].
3. Implementation
The procedure for the calculation of the natural gas density,
compression, molar heat capacity, isentropic exponent, and the
Joule–Thomson coefficient is implemented in the OOP mode,
which enables an easy integration into new software projects.
The interface to the software object is shown in Fig. 2. The
input/output parameters and functions are accessible, while
the internal structure is hidden from the user. The function
“Calculate” maps the input parameters (pressure, temperature,
and the molar fractions of natural gas components) into the
output parameters (density, compression, molar heat capacity,
isentropic exponent, and Joule–Thomson coefficient). Table 3
depicts the calculation procedure. Prior to the calculation of
the molar heat capacities, the isentropic exponent, and the
Joule–Thomson coefficient, the density and the compression
factor of a natural gas must be calculated. The false position
method is combined with the successive bisection method
to calculate the roots of the equation of state [4]. Using
CORBA [8] or DCOM [9], the component can be accessed
remotely.
22 I. Mari´ c / Flow Measurement and Instrumentation 18 (2007) 18–26
Fig. 2. Interface to the software object that implements the calculation of the natural gas properties.
Table 3
The input/output parameters and the procedure for the computation of the
natural gas properties
Input parameters—constant:
• Molar gas constant (R = 8314.51 J/(kmol K))
• Natural gas equation of state parameters (a
n
, b
n
, c
n
, k
n
, u
n
, g
n
, q
n
, f
n
,
s
n
, w
n
; n = 1, 2, . . . , 58), characterization parameters (M
i
, E
i
, K
i
, G
i
,
Q
i
, F
i
, S
i
, W
i
; i = 1, . . . , 21) and binary interaction parameters (E

i, j
,
U
i, j
, K
i, j
, G

i, j
) (see ISO 12213-2)
• DIPPR/AIChE gas heat capacity constants (a
j
, b
j
, c
j
, d
j
, e
j
; j = 1, 2,
. . . , N)
Input parameters—time varying:
• Absolute pressure: p (MPa)
• Absolute temperature: T (K)
• Molar fractions of the natural gas mixture: y
i
; i = 1, 2, . . . , N
Calculation procedure:
1 Mixture size parameter K (Eq. (13)), second virial coefficient
B (Eq. (14)) and temperature dependent coefficients C

n
(Eq. (18))
2 Compression factor Z (Eq. (8)) (see ISO-12213-2 for details of
calculation)
3 Molar density ρ
m
= p/RT Z, reduced density ρ
r
= K
3
ρ
m
and molar
volume v
m
= 1/ρ
m
.
4 Coefficients D
n
and D
1n
(Eqs. (32) and (36))
5 1st and 2nd derivative of the second virial coefficient B: B

(Eq. (23)) and
B

(Eq. (24))
6 1st and 2nd derivative of the coefficient C

n
: C

n
(Eq. (25)) and
C

n
(Eq. (26))
7 1st derivative of the compression factor Z : (∂ Z/∂T)
p
(Eq. (33))
8 Partial derivatives of pressure: (∂p/∂T)
v
m
(Eq. (29)) and (∂p/∂v
m
)
T
(Eq. (38))
9 Ideal molar heat capacity of a gas mixture at constant pressure: c
m, pI
(Eq. (27))
10 Molar heat capacity of a gas mixture at constant volume: c
m,v
(Eq. (9))
11 Molar heat capacity of a gas mixture at constant pressure: c
m, p
(Eq. (4))
12 Isentropic exponent κ (Eq. (37))
13 Joule–Thomson coefficient µ
JT
(Eq. (39))
4. Application
We investigated the combined effect of the Joule–Thomson
coefficient and the isentropic exponent of a natural gas on
the accuracy of flow-rate measurements based on differential
devices. The measurement of a natural gas [4] flowing in a
pipeline through an orifice plate with corner taps (Fig. 1) is
assumed to be completely in accordance with the international
standard ISO-5167-2 [10]. The detailed description of the flow-
rate equation with the corresponding iterative computation
scheme is given in [2] and [10]. The flow-rate through the
orifice is proportional to the expansibility factor, which is
related to the isentropic exponent. According to [10], the
expansibility factor ε for the orifice plate with corner taps is
defined by:
ε = 1 − (0.351 + 0.256β
4
+ 0.93β
8
)[1 − ( p
2
/p
1
)
1/κ
] (40)
where β denotes the ratio of the diameter of the orifice to
the internal diameter of the pipe, while p
1
and p
2
are the
absolute pressures upstream and downstream of the orifice
plate, respectively. The corresponding temperature change T
of the gas for the orifice plate with corner taps is defined by
T = T
1
− T
2
≈ µ
JT
( p
1
, T
2
)ω (41)
where T
1
and T
2
indicate the corresponding temperatures
upstream and downstream of the orifice plate, µ
JT
( p
1
, T
2
) is
the Joule–Thomson coefficient at upstream pressure p
1
and
downstream temperature T
2
, and ω is the pressure loss across
the orifice plate [11], defined by
ω =

1 − β
4

1 − C
2

− Cβ
2

1 − β
4

1 − C
2

+ Cβ
2
p (42)
where C denotes the coefficient of discharge for the orifice plate
with corner taps [10], and P is the pressure drop across the
orifice plate. According to [2], the temperature of the fluid shall
preferably be measured downstream of the primary device, but
upstream temperature shall be used for the calculation of the
flow-rate. Within the limits of the application of ISO-5167 [2],
it is generally assumed that the temperature drop across the
differential device can be neglected, but it is also suggested that
it be taken into account if higher accuracies are required. It is
also assumed that the isentropic exponent can be approximated
by the ratio of the specific heat capacity at constant pressure
to the specific heat capacity at constant volume of the ideal
gas. The above approximations may produce a measurement
error. The relative flow measurement error E
r
is estimated by
comparing the approximate (q
m2
) and the corrected (q
m1
) mass
flow-rate i.e.
E
r
= (q
m2
− q
m1
) /q
m1
. (43)
I. Mari´ c / Flow Measurement and Instrumentation 18 (2007) 18–26 23
Table 4
The procedure for the correction of the mass flow-rate due to adiabatic
expansion effects
Step Description
1 Calculation of Joule–Thomson coefficient µ
JT
( p
1
, T
2
) by using the
procedure outlined in Table 3 with upstream pressure p
1
and
downstream temperature T
2
2 Estimation of upstream temperature T
1
using the Eq. (41)
3 Calculation of density ρ = Mρ
m
and isentropic exponent κ using
the procedure outlined in Table 3 with upstream pressure p
1
and
estimated upstream temperature T
1
4 Calculation of viscosity (residual viscosity function) [12]
5 Calculation of mass flow-rate in accordance with [10]
Fig. 3. Calculated and measured molar heat capacities at constant pressure of
the natural gas mixture.
Fig. 4. Calculated and measured Joule–Thomson coefficients of the natural gas
mixture.
The individual and the combined relative errors due to the
approximations of the temperature drop and the isentropic
exponent are calculated. The procedure for the correction of the
mass flow-rate due to the adiabatic expansion effects is shown
in Table 4. The calculation procedures are implemented in the
OOP mode. The results are presented in the following section.
Table 5
Difference between the calculated and measured specific heat capacity at
constant pressure of a natural gas
P (MPa): T (K) 250 275 300 350
(c
p calculated
− c
p measured
) (J/(g ∗ K))
0.5 −0.015 −0.018 −0.018 −0.012
1.0 −0.002 −0.014 −0.016 −0.011
2.0 −0.012 −0.019 −0.022 −0.020
3.0 −0.032 −0.020 −0.023 −0.026
4.0 −0.041 −0.023 −0.021 −0.027
5.0 −0.051 −0.022 −0.025 −0.029
7.5 −0.055 −0.032 – –
10.0 −0.077 −0.033 −0.048 −0.042
11.0 −0.075 – – –
12.5 −0.092 −0.030 – –
13.5 −0.097 −0.039 – –
15.0 −0.098 −0.033 −0.082 −0.069
16.0 – −0.036 – –
17.5 – −0.043 −0.075 –
20.0 −0.081 −0.048 −0.066 −0.134
25.0 −0.082 −0.033 −0.064 −0.171
30.0 −0.077 −0.025 −0.070 −0.194
Table 6
Difference between the calculated and measured Joule–Thomson coefficient of
a natural gas
P (MPa): T (K) 250 275 300 350

JT calculated
− µ
JT measured
) (K/MPa)
0.5 −0.014 −0.023 −0.075 −0.059
1.0 −0.032 −0.024 −0.068 −0.053
2.0 – – – −0.051
3.0 −0.092 −0.032 −0.069 −0.049
5.0 −0.022 −0.036 −0.044 −0.026
7.5 0.043 – – –
10.0 0.060 0.096 0.019 0.030
12.5 0.034 – – –
15.0 0.113 0.093 0.050 0.061
20.0 0.029 0.084 0.009 0.047
25.0 0.025 0.059 0.002 0.043
30.0 0.031 0.052 0.005 0.012
5. Measurement results
In order to compare the calculation results, for the specific
heat capacity c
p
and the Joule–Thomson coefficient µ
JT
, with
the corresponding high accuracy measurement data [13] (Ernst
et al.), we assume identical artificial natural gas mixtures with
the following mole fractions: x
CH
4
= 0.79942, x
C
2
H
6
=
0.05029, x
C
3
H
8
= 0.03000, x
CO
2
= 0.02090, and x
N
2
=
0.09939. The results of the measurements [13] and the
results of the calculation of the specific heat capacities c
p
and the Joule–Thomson coefficients µ
JT
of the natural gas
mixtures, for absolute pressures ranging from 0 to 30 MPa
in 0.5 MPa steps and for four upstream temperatures (250,
275, 300, and 350 K), are shown in Fig. 3 and Fig. 4,
respectively. The differences between the calculated values and
the corresponding measurement results [13], for c
p
and µ
JT
,
are shown in Table 5 and Table 6, respectively. From Table 5,
it can be seen that the calculated values of c
p
are within
±0.08 J/(g ∗ K) of the measurement results for pressures up
24 I. Mari´ c / Flow Measurement and Instrumentation 18 (2007) 18–26
to 12 MPa. At higher pressures, up to 30 MPa, the difference
increases, but never exceeds ±0.2 J/(g ∗ K). For pressures
up to 12 MPa, the relative difference between the calculated
and experimentally obtained c
p
never exceeds ±2.00%. From
Table 6, it can be seen that the calculated values of µ
JT
are
within ±0.113 K/MPa, with the experimental results for the
pressures up to 30 MPa. The relative difference increases with
the increase in pressure, but never exceeds ±2.5% for pressures
up to 12 MPa. At higher pressures, when the values of µ
JT
are
close to zero, the relative difference may increase significantly.
The results of the calculations obtained for pure methane and
methane–ethane mixtures are in considerably better agreement
with the corresponding experimental data [13] than those for
the natural gas mixture shown above. We estimate that the
relative uncertainties of the calculated c
p
and µ
JT
of the
AGA-8 natural gas mixtures in common industrial operating
conditions (pressure range 0–12 MPa and temperature range
250–350 K) are unlikely to exceed ±3.00% and ±4.00%,
respectively. Fig. 5 shows the results of the calculation of
the isentropic exponent. Since the isentropic exponent is a
theoretical parameter, there exist no experimental data for its
verification.
In order to simulate a flow-rate measurement error caused
by an inappropriate compensation for the adiabatic expansion,
a natural gas mixture (Gas 3) form Annex C of [4] is
assumed to flow through the orifice plate with corner taps [9]
shown in Fig. 1. Following the recommendations [2], the
absolute pressure is assumed to be measured upstream ( p
1
),
and the temperature downstream (T
2
), of the primary device.
A natural gas analysis in mole fractions is the following:
methane 0.859, ethane 0.085, propane 0.023, carbon dioxide
0.015, nitrogen 0.010, i-butane 0.0035, n-butane 0.0035, i-
pentane 0.0005, and n-pentane 0.0005. Fig. 6 illustrates the
temperature drop due to the Joule–Thomson effect calculated
in accordance with Eq. (41). The results calculated are given
for two discrete differential pressures, p (20 and 100 kPa),
for absolute pressures p
1
ranging from 1 to 60 MPa in 1 MPa
steps, and for six equidistant upstream temperatures T
1
in the
range from 245 to 345 K. From Fig. 6 it can be seen that
for each temperature there exists the corresponding pressure
where the Joule–Thomson coefficient changes its sign, and
consequently alters the sign of the temperature change. A
relative error in the flow-rate measurements (Fig. 1) due to the
Joule–Thomson effect is shown in Fig. 7. The error is calculated
in accordance with Eq. (43) by comparing the approximate
mass flow-rate (q
m2
) with the corrected mass flow-rate (q
m1
).
The procedure for the precise correction of the mass flow-
rate is shown in Table 4. The approximate flow-rate and the
corresponding natural gas properties (density, viscosity, and
isentropic exponent) are calculated at upstream pressure p
1
,
downstream temperature T
2
, and differential pressure p, by
neglecting the temperature drop due to the Joule–Thomson
effect (T
2
= T
1
). The results are shown for two discrete
differential pressures p (20 and 100 kPa), for absolute
upstream pressures p
1
ranging from 1 to 60 MPa in 1 MPa
steps, and for four equidistant downstream temperatures T
2
in
the range from 245 to 305 K.
Fig. 5. Calculated isentropic exponent of the natural gas mixture.
Fig. 6. Temperature drop due to the Joule–Thomson effect T = µ
JT
ω
when measuring flow-rate of the natural gas mixture through the orifice plate
with corner taps (ISO-5167-2). The upstream pressure varies from 1 to 60 MPa
in 1 MPa steps and upstream temperature from 245 to 305 K in 20 K steps for
each of two differential pressures p (20 and 100 kPa). The internal diameters
of the orifice and pipe are: d = 120 mm and D = 200 mm.
Fig. 7. Relative error E
r
= (q
m2
− q
m1
) /q
m1
in the flow-rate of the natural
gas mixture measured by the orifice plate with corner taps (ISO-5167-2) when
using downstream temperature with no compensation for the Joule–Thomson
effect (q
m2
) instead of upstream temperature (q
m1
). The upstream pressure
varies from 1 to 60 MPa in 1 MPa steps, and downstream temperature from
245 to 305 K in 20 K steps for each of two differential pressures p (20 and
100 kPa). The internal diameters of the orifice and pipe are: d = 120 mm and
D = 200 mm.
I. Mari´ c / Flow Measurement and Instrumentation 18 (2007) 18–26 25
Fig. 8. Relative error E
r
= (q
m2
− q
m1
) /q
m1
in the flow-rate of the natural
gas mixture measured by the orifice plate with corner taps (ISO-5167-2) when
using the isentropic exponent of the ideal gas (q
m2
) instead of the real gas
(q
m1
). The upstream pressure varies from 1 to 60 MPa in 1 MPa steps, and
downstream temperature from 245 to 305 K in 20 K steps, for each of two
differential pressures p (20 and 100 kPa). The internal diameters of the orifice
and pipe are: d = 120 mm and D = 200 mm.
Fig. 8 illustrates the relative error in the flow-rate
measurements due to the approximation of the isentropic
exponent by the ratio of the ideal molar heat capacities.
The error is calculated by comparing the approximate mass
flow-rate (q
m2
) with the corrected mass flow-rate (q
m1
) in
accordance with Eq. (43). The procedure for the precise
correction of the mass flow-rate is shown in Table 4. The
approximate flow-rate calculation is carried out in the same
way, with the exception of the isentropic exponent, which
equals the ratio of the ideal molar heat capacities (κ =
c
m, pI
/(c
m, pI
− R)). The results are shown for two discrete
differential pressures p (20 and 100 kPa), for absolute
upstream pressures p
1
ranging from 1 to 60 MPa in 1 MPa
steps, and for four equidistant downstream temperatures T
2
in
the range from 245 to 305 K.
Fig. 9 shows the flow-rate measurement error produced
by the combined effect of the Joule–Thomson and the
isentropic expansions. The error is calculated by comparing the
approximate mass flow-rate (q
m2
) with the corrected mass flow-
rate (q
m1
) in accordance with Eq. (43). The procedure for the
precise correction of the mass flow-rate is shown in Table 4.
The approximate flow-rate and the corresponding natural gas
properties are calculated at upstream pressure p
1
, downstream
temperature T
2
, and differential pressure p, by neglecting the
temperature drop due to the Joule–Thomson effect (T
2
= T
1
)
and by substituting the isentropic exponent with the ratio of the
ideal molar heat capacities, κ = c
m, pI
/(c
m, pI
−R). The results
are shown for two discrete differential pressures p (20 and
100 kPa), for absolute upstream pressures p
1
ranging from 1 to
60 MPa in 1 MPa steps, and for four equidistant downstream
temperatures T
2
in the range from 245 to 305 K.
The results obtained for the Joule–Thomson coefficient
and isentropic exponent are in complete agreement with
the results obtained when using the procedures described
in [7] and [14], which use natural gas fugacities to derive
the molar heat capacities and are, therefore, considerably
more computationally intensive and time consuming. The
Fig. 9. Relative error E
r
= (q
m2
− q
m1
) /q
m1
in the flow-rate of the natural
gas mixture measured by an orifice plate with corner taps (ISO-5167-2) when
using downstream temperature, with no compensation for the Joule–Thomson
effect and the isentropic exponent of the ideal gas at downstream temperature
(q
m2
) instead of upstream temperature, and the corresponding real gas
isentropic exponent (q
m1
). The upstream pressure varies from 1 to 60 MPa in
1 MPa steps and downstream temperature from 245 to 305 K in 20 K steps for
each of two differential pressures p (20 and 100 kPa). The internal diameters
of the orifice and pipe are: d = 120 mm and D = 200 mm.
calculation results are shown up to a pressure of 60 MPa,
which lies within the wider ranges of application given
in [4], of 0–65 MPa. However, the lowest uncertainty for
compressibility is for pressures up to 12 MPa, and no
uncertainty is quoted in Reference [4] for pressures above
30 MPa. It would therefore seem sensible for the results of
the Joule–Thomson and the isentropic exponent calculations
to be used with caution above this pressure. From Fig. 9 it
can be seen that the maximum combined error is lower than
the maximum individual errors, because the Joule–Thomson
coefficient (Fig. 7) and the isentropic exponent (Fig. 8) show
the countereffects on the flow-rate error. The error always
increases by decreasing the operating temperature. The total
measurement error is still considerable, especially at lower
temperatures and higher differential pressures, and cannot be
overlooked. The measurement error is also dependent on the
natural gas mixture. For certain mixtures, like natural gases
with a high carbon dioxide content, the relative error in
the flow-rate may increase up to 0.5% at lower operating
temperatures (245 K), and up to 1.0% at very low operating
temperatures (225 K). Whilst modern flow computers have a
provision for applying a Joule–Thomson coefficient and an
isentropic exponent correction to measured temperatures, this
usually takes the form of a fixed value supplied by the
user. The calculations in this paper show that any initial error
in choosing this value, or subsequent operational changes
in temperature, pressure or gas composition, could lead to
significant systematic metering errors. Our further work will be
directed towards the improvement of the model, with the aim
of simplifying the calculation procedure and further decreasing
the uncertainty of the calculation results.
6. Conclusion
This paper describes the numerical procedure for the
calculation of the natural gas molar heat capacity, the
26 I. Mari´ c / Flow Measurement and Instrumentation 18 (2007) 18–26
Joule–Thomson coefficient and the isentropic exponent. The
corresponding equations have been derived by applying the
fundamental thermodynamic relations to AGA-8 extended
virial-type equations of state. The DIPPR AIChE generic
ideal heat capacity equations have been used to calculate the
ideal molar heat capacities of a natural gas mixture. The
implementation of the procedure in an OOP mode enables
its easy integration into new software developments. An
example of a possible application of the procedure in the
flow-rate measurements has been given. The procedure can be
efficiently applied in both off-line calculations and real time
measurements.
Acknowledgements
This work has been supported by the Croatian Ministry of
Science, Education, and Sport with the projects “Computational
Intelligence Methods in Measurement Systems” and “Real-Life
Data Measurement and Characterization”. The author would
like to express his thanks to Professor Donald R. Olander,
University of California, Berkeley, USA, for his valuable
suggestions, and to Dr. Eric W. Lemmon, NIST, USA, for his
help in obtaining the relevant measurement data.
References
[1] Shoemaker DP, Garland CW, Nibler JW. Experiments in physical
chemistry. 6th ed. New York: McGraw-Hill; 1996.
[2] ISO-51671-1. Measurement of fluid flowby means of pressure differential
devices inserted in circular-cross section conduits running full — Part 2:
General principles and requirements. Ref. No. ISO-51671-1:2003(E).
ISO; 2003.
[3] AGA 8. Compressibility and supercompressibility for natural gas and
other hydrocarbon gases. Transmission measurement committee report
no. 8, AGA catalog no. XQ 1285. Arlington (Va.); 1992.
[4] ISO-12213-2. Natural gas – calculation of compression factor – Part
2: Calculation using molar-composition analysis. Ref. No. ISO-12213-
2:1997(E). ISO; 1997.
[5] DIPPR
R
801. Evaluated standard thermophysical property values, Design
Institute for Physical Properties. Sponsored by AIChE. 2004.
[6] Olander DR. Thermodynamic relations. In: Engineering thermodynam-
ics, Retrieved September 21, 2005, from Department of Nuclear En-
gineering, University of California: Berkeley [Chapter 7]. WEB site:
http://www.nuc.berkeley.edu/courses/classes/E-115/Reader/index.htm.
[7] Mari´ c I. The Joule–Thomson effect in natural gas flow-rate measurements.
Flow Measurement and Instrumentation 2005;16:387–95.
[8] Vinoski S. CORBA: Integrating diverse applications within distributed
heterogeneous environments. IEEE Communications Magazine 1997;
35(2):46–55.
[9] Microsoft: DCOM Architecture. White Paper. Microsoft Corporation.
1998.
[10] ISO-51671-2. Measurement of fluid flowby means of pressure differential
devices inserted in circular-cross section conduits running full — Part 2:
Orifice plates. Ref. No. ISO-51671-2:2003(E). ISO; 2003.
[11] Urner G. Pressure loss of orifice plates according to ISO-5167. Flow
Measurement and Instrumentation 1997;8:39–41.
[12] Miller RW. The flow measurement engineering book. 3rd ed. New York:
McGraw-Hill; 1996.
[13] Ernst G, Keil B, Wirbser H, Jaeschke M. Flow calorimetric results for
the massic heat capacity cp and Joule–Thomson coefficient of CH
4
, of
(0.85CH
4
+ 0.16C
2
H
6
), and of a mixture similar to natural gas. Journal
of Chemical Thermodynamics 2001;33:601–13.
[14] Mari´ c I, Galovi´ c A, Šmuc T. Calculation of natural gas isentropic
exponent. Flow Measurement and Instrumentation 2005;16(1):13–20.

v = cm. (1) becomes cm. for the calculation of the Joule–Thomson coefficient and the isentropic exponent. 1. After substituting the first and the second derivative of the AGA-8 compressibility equation [4]. the Eq. (2) ∂2 p ∂T 2 vm . T (1) variable equals to the partial derivative of the second coefficient with respect to the first variable. the order of forming the mixed second derivative is irrelevant. p − cm. After substituting vm = 1/ρm .v /T since the molar heat at constant volume is defined by cm. Mari´ / Flow Measurement and Instrumentation 18 (2007) 18–26 c 19 Fig. (4) where cm. and vm the second variable with the corresponding coefficient (∂ p/∂ T )vm . i. The Maxwell relation (∂s/∂vm )T = (∂ p/∂ T )vm . Real gases behave more like ideal gases as the pressure approaches zero or vm I → ∞.v I = cm. and then dividing the resulting equation by dvm while holding p constant.v dT + T ∂p ∂T dvm .v = cm. p I − R. (7) and after integration we obtain cm. By dividing the fundamental differential for internal energy du = T · ds − p · dvm by dT while holding vm constant.v = cm.e. related to temperature and molar volume [6]. By applying this property to Eq. and finally inverting the partial derivative (∂ T /∂vm ) p . The total differential for entropy. p /T . R is the universal gas constant. T denotes temperature. (5) can be rewritten in the following integral form: vm cm. (2) and (3) must be the exact differential. Z is the compression factor.v I + T vm I →∞(T =const) ∂2 p ∂T 2 vm dvm (6) vm Similarly. Finally.v = (∂u/∂ T )vm . (6) transforms to: cm.v = T ∂vm ∂T ∂p ∂T . the following equation is obtained: cm. we obtain: ∂cm.I. p I denotes the temperature dependent molar heat capacity of the ideal gas at constant pressure.v I . The Maxwell relation (∂s/∂ p)T = (∂vm /∂ T ) p is used to substitute the coefficient of d p and the following relation is obtained: ds = cm. the coefficient of dT in the total differential becomes cm. (2) and by assuming T to be the first variable with the corresponding coefficient cm. The partial derivative of the first coefficient with respect to the second into the Eq. starting from a total differential for entropy related to temperature and pressure [6] ds = (∂s/∂ T ) p dT +(∂s/∂ p)T d p and by dividing the fundamental differential for enthalpy dh = T · ds + vm · d p by dT while holding p constant. and ρm I and ρm are the corresponding molar densities of the ideal and the real gas at temperature T . the corresponding molar volume of the ideal. since the molar heat capacity at constant pressure is defined by cm. the coefficient of dT in Eq. vm I . (5) The Eq. Schematic diagram of the gas flow-rate measurement using an orifice plate with corner taps. (3). and the real gas at temperature T respectively. the Eq. 18 Z = 1 + Bρm − ρr n=13 58 ∗ Cn p vm + n=13 ∗ k b Cn bn − cn kn ρr n ρr n e−cn ρr kn (8) The total differential of the thermodynamic property Eqs. (2) from Eq. and νm denote the ideal molar heat capacity at constant volume. p = RT Zρm and cm. is used to substitute the coefficient of dvm . p I − R − RT ρm × ρm I →0(T =const) 1 ρm 2 ∂Z ∂T +T ρm ∂2 Z ∂T 2 ρm dρm (7) (3) Subtracting Eq.v ∂vm =T T where s denotes entropy. and vm is a molar volume of a gas. p where cm. p = (∂h/∂ T ) p .v /T . (1) becomes: ds = cm. p dT + T ∂vm ∂T d p. is: ds = ∂s ∂T dT + vm ∂s ∂vm dvm . p I − R + RTρr (2C0 + T C1 − C2 ) (9) .

{Fi }. {G i }. E i j = E i∗j (E i E j )1/2 and ∗ G i j = G i j (G i + G j )/2. {cn }. i=1 j=i+1 (19) N G= (17) Q= i=1 i=1 N yi G i + 2 (20) where T is the temperature. pi D d h K p qm R s T vm vm I yi Z β p ω κ µJT ρm ρm I ρr where ρr is the reduced density (ρr = K 3 ρm ). (21) and N F= i=1 yi2 Fi . and F are calculated using the equations [4]: ∗ Cn = an (G + 1 − gn )gn (Q 2 + 1 − qn )qn × (F + 1 − f n ) fn U u n T −u n . {qn }. . { f n }. p cm. For additional symbols and units refer to ISO-12213-2 [4]. while {bn }. .51 Specific entropy Absolute temperature Molar specific volume Molar specific volume of ideal gas Molar fraction of i-th component in gas mixture Compression factor Diameter ratio d/D Differential pressure Pressure loss Isentropic exponent Joule–Thomson coefficient Molar density Molar density of ideal gas Reduced density ∗ and the coefficients Bni j . {Q i }. (13) where yi denotes the molar fraction of the component i. {u n }. while {K i } and K i j are the corresponding size parameters and the binary interaction parameters given in [4]. the second virial coefficient is calculated using the following equation: 18 N N ∗ yi y j Bni j E iujn (K i K j )3/2 i=1 j=1 B= n=1 an T −u n (14) Second virial coefficient Mixture interaction coefficient Coefficient of discharge Molar heat capacity at constant pressure Molar heat capacity at constant volume Temperature and composition dependent coefficients AGA-8 equation of state parameter Specific heat capacity at constant pressure Ideal molar heat capacity of the natural gas mixture Ideal molar heat capacity of the gas component j Upstream internal pipe diameter Diameter of orifice Specific enthalpy Size parameter Absolute pressure Mass flow-rate Molar gas constant 8314. . E i j and G i j are defined by ∗ Bni j = (G i j + 1 − gn )gn (Q i Q j + 1 − qn )qn N 2 5/2 yi E i N × (Fi F j + 1 − f n ) fn · (Si S j + 1 − sn )sn (Wi W j + 1 − wn )wn . p I j Cm. and {wn } are the parameters of the equation of state. {E i }. {an }. Mari´ / Flow Measurement and Instrumentation 18 (2007) 18–26 c Table 1 Symbols and units (for additional symbols and units refer to ISO-12213-2 [4]) ∗ Cn − n=13 18 with 18 C0 = B K3 B K3 b ρr n −1 e−cn ρr kn (10) Symbols and units Symbol Description Unit m3 ∗ kmol−1 – – J/(mol K) J/(mol K) – – J/(kg K) J/(mol K) J/(mol K) m m J/kg – Pa kg/s J/(kmol K) J/(kg K) K m3 /kmol m3 /kmol – – – Pa Pa – K/Pa kmol/m3 kmol/m3 – C1 = n=13 58 ∗ Cn − (11) C2 = n=13 ∗ ∗ 2Cn + T Cn (12) B ∗ Bni j C cm. N is the total number of gas mixture components. The ∗ temperature dependent coefficients Cn .20 I. Ui j is the binary interaction parameter for mixture energy. and {kn } are the parameters of the equation of state. {sn }. (22) where. {K i }. are: 18 N N ∗ yi y j Bni j E iujn (K i K j )3/2 (23) i=1 j=1 B =− (18) n=1 an u n T −u n −1 . . yi is the molar fraction of the component i. The mixture size parameter K is calculated using the following equation [4]: N 2 5/2 yi K i N −1 N K = i=1 5 +2 i=1 j=i+1 yi y j (K i5j − 1)(K i K j )5/2 . Cn are the temperature dependent coefficients. {Si }. {gn }. The main symbols and units are given in Table 1. Q. yi Q i . with respect to temperature. and {Wi } are the corresponding characterization parameters. i=1 j=i+1 N −1 N ∗ yi y j (G i j − 1)(G i + G j ). n = 1. 1/2 1/2 U (15) (16) 5 = +2 i=1 N −1 yi y j (Ui5j − 1)(E i E j )5/2 .v ∗ Cn cn cp cm. The first and the second derivatives of the coefficients ∗ B and Cn . while ∗ E i∗j and G i j are the corresponding binary interaction parameters. B is the ∗ second virial coefficient. According to [4]. G. K is the mixture size parameter. 58 and the mixture parameters U .

and the partial derivative of molar volume with respect to temperature at constant pressure.I.5 1626. b. T The ideal molar heat capacity c p I is calculated by N and 2 k k D1n = K 3 [bn − cn kn (2bn + kn − cn kn ρr n )ρr n ] b × ρr n −1 e−cn ρr . 58 j where cm.6 B = n=1 an u n (u n + 1)T −u n −2 N N ∗ yi y j Bni j E iujn (K i K j )3/2 i=1 j=1 C4 = C5 + n=13 ∗ Cn D1n 18 (34) × ∗ Cn = − (24) (25) (26) C5 = B − K 3 n=13 ∗ Cn (35) un ∗ C T n un + 1 ∗ ∗ Cn = − Cn .e. Table 3 depicts the calculation procedure. p cm. and the molar fractions of natural gas components) into the output parameters (density.87 723.9 1655. The constants a.5 58 d 41 602 73 223 116 800 e 991. Implementation The procedure for the calculation of the natural gas density. pi is the molar heat capacity of the same component. The partial derivative of pressure with respect to temperature at constant molar volume. d j . and e for some gases are shown in Table 2.v ρm p ∂p ∂vm . and the Joule–Thomson coefficient. R(T Z )2 + pT C4 . isentropic exponent.v ∂p ∂vm ∂p ∂ρm T vm p =− cm. Mari´ / Flow Measurement and Instrumentation 18 (2007) 18–26 c Table 2 The DIPPR/AIChE gas heat capacity constants Natural gas component Methane — CH4 Ethane — C2 H6 Propane — C3 H8 18 21 DIPPR/AIChE ideal gas heat capacity constants a b 33 298 40 326 51 920 79 933 134 220 192 450 c 2086. (38) c j /T sinh(c j /T ) 2 + dj e j /T cosh(e j /T ) 2 . pi j (27) κ=− where ∂p ∂vm cm. T (37) where y j is the molar fraction of component j in the gas j mixture and Cm. the component can be accessed remotely. (28) The Joule–Thomson coefficient is defined by the following equation [2]: µJT = RT 2 pcm. molar heat capacity. Using CORBA [8] or DCOM [9]. The function “Calculate” maps the input parameters (pressure. compression. 3. The false position method is combined with the successive bisection method to calculate the roots of the equation of state [4]. ∂Z ∂T = p R(T Z )2 C3 − p Z [T K 3 C0 + C4 ] . p I = j=1 y j cm. p is the molar heat capacity of the component j of the ideal gas mixture. The input/output parameters and functions are accessible. a j . temperature. which enables an easy integration into new software projects. and T is the temperature. pi = a j + b j j = T = T 2 −RTρm (Z ∂ρm ∂vm T + ρm C4 ) . Prior to the calculation of the molar heat capacities. (39) is given in [6] and [7]. while the internal structure is hidden from the user. The interface to the software object is shown in Fig. c j . 2. p ∂Z ∂T . isentropic exponent. pi (39) The derivation of the Eq. kn (31) (32) (33) k b Dn = (bn − cn kn ρr n )ρr n e−cn ρr . c. molar heat capacity. and Joule–Thomson coefficient). cm. vm = Rρm [Z + T (C3 − ρr C0 )] (29) = p R p Z+ ∂Z ∂T T p (30) C3 = n=13 ∗ ∗ C n Dn . The molar heat capacities of the ideal gas mixture components can be approximated by the DIPPR/AIChE generic equations [5]. and e j are the corresponding constants. d. b j . the isentropic exponent. p cm. are defined by the equations: ∂p ∂T and ∂vm ∂T where.96 752. kn (36) The isentropic exponent is defined by the following relation cm. i. and the Joule–Thomson coefficient is implemented in the OOP mode. compression. the density and the compression factor of a natural gas must be calculated.

. According to [2]. reduced density ρr = K 3 ρm and molar volume vm = 1/ρm . the expansibility factor ε for the orifice plate with corner taps is defined by: ε = 1 − (0. 4 Coefficients Dn and D1n (Eqs. j . f n . i = 1. (14)) and temperature dependent coefficients Cn (Eq.22 I. The corresponding temperature change T of the gas for the orifice plate with corner taps is defined by T = T1 − T2 ≈ µJT ( p1 . j = 1. but upstream temperature shall be used for the calculation of the flow-rate.256β 4 + 0. and P is the pressure drop across the orifice plate. p (Eq. it is generally assumed that the temperature drop across the differential device can be neglected. The above approximations may produce a measurement error. n = 1. (33)) 8 Partial derivatives of pressure: (∂ p/∂ T )vm (Eq. defined by ω= 1 − β 4 1 − C 2 − Cβ 2 1 − β 4 1 − C 2 + Cβ 2 p (42) 4. ∗ Ui. u n . Mari´ / Flow Measurement and Instrumentation 18 (2007) 18–26 c Fig. (38)) 9 Ideal molar heat capacity of a gas mixture at constant pressure: cm. Table 3 The input/output parameters and the procedure for the computation of the natural gas properties Input parameters—constant: • Molar gas constant (R = 8314. which is related to the isentropic exponent. d j . The relative flow measurement error Er is estimated by comparing the approximate (qm2 ) and the corrected (qm1 ) mass flow-rate i. . 2. (25)) and ∗ Cn (Eq. gn . j .51 J/(kmol K)) • Natural gas equation of state parameters (an . Si .v (Eq. Within the limits of the application of ISO-5167 [2]. . p I (Eq. (27)) 10 Molar heat capacity of a gas mixture at constant volume: cm. T2 ) ω (41) where T1 and T2 indicate the corresponding temperatures upstream and downstream of the orifice plate. . Interface to the software object that implements the calculation of the natural gas properties. (37)) 13 Joule–Thomson coefficient µJT (Eq. qn . (43) . but it is also suggested that it be taken into account if higher accuracies are required. b j . i = 1. G i . G i.e. According to [10]. Application We investigated the combined effect of the Joule–Thomson coefficient and the isentropic exponent of a natural gas on the accuracy of flow-rate measurements based on differential devices. The flow-rate through the orifice is proportional to the expansibility factor. respectively. . . second virial coefficient ∗ B (Eq. characterization parameters (Mi . 2. Wi . µJT ( p1 . . and ω is the pressure loss across the orifice plate [11]. (24)) ∗ ∗ 6 1st and 2nd derivative of the coefficient Cn : Cn (Eq. 1) is assumed to be completely in accordance with the international standard ISO-5167-2 [10]. 2. It is also assumed that the isentropic exponent can be approximated by the ratio of the specific heat capacity at constant pressure to the specific heat capacity at constant volume of the ideal gas. (26)) 7 1st derivative of the compression factor Z : (∂ Z /∂ T ) p (Eq. . wn . E i . (18)) 2 Compression factor Z (Eq. (23)) and B (Eq. the temperature of the fluid shall preferably be measured downstream of the primary device. K i . while p1 and p2 are the absolute pressures upstream and downstream of the orifice plate.351 + 0.93β 8 )[1 − ( p2 / p1 )1/κ ] (40) where β denotes the ratio of the diameter of the orifice to the internal diameter of the pipe. 58). . The measurement of a natural gas [4] flowing in a pipeline through an orifice plate with corner taps (Fig. N) Input parameters—time varying: • Absolute pressure: p (MPa) • Absolute temperature: T (K) • Molar fractions of the natural gas mixture: yi . (39)) scheme is given in [2] and [10]. Fi . cn . (8)) (see ISO-12213-2 for details of calculation) 3 Molar density ρm = p/RT Z . (29)) and (∂ p/∂vm )T (Eq. . sn . Er = (qm2 − qm1 ) /qm1 . kn . T2 ) is the Joule–Thomson coefficient at upstream pressure p1 and downstream temperature T2 . . (13)). K i. ∗ Q i . 21) and binary interaction parameters (E i. j . . . . e j . j ) (see ISO 12213-2) • DIPPR/AIChE gas heat capacity constants (a j . (32) and (36)) 5 1st and 2nd derivative of the second virial coefficient B: B (Eq. bn . . . c j . (9)) 11 Molar heat capacity of a gas mixture at constant pressure: cm. (4)) 12 Isentropic exponent κ (Eq. 2. N Calculation procedure: 1 Mixture size parameter K (Eq. The detailed description of the flowrate equation with the corresponding iterative computation where C denotes the coefficient of discharge for the orifice plate with corner taps [10].

08 J/(g ∗ K) of the measurement results for pressures up Fig.060 0.014 −0.043 0.064 −0.044 – 0.077 −0.025 – −0. Calculated and measured molar heat capacities at constant pressure of the natural gas mixture.029 – −0.075 −0.0 5. The procedure for the correction of the mass flow-rate due to the adiabatic expansion effects is shown in Table 4.0 12. and 350 K).018 −0.012 Fig. 275.055 −0.113 0.026 −0.077 −0.031 −0.082 −0.042 – – – −0.0 250 275 300 (µJT calculated − µJT measured ) (K/MPa) −0. Mari´ / Flow Measurement and Instrumentation 18 (2007) 18–26 c Table 4 The procedure for the correction of the mass flow-rate due to adiabatic expansion effects Step 1 Description Calculation of Joule–Thomson coefficient µJT ( p1 .036 −0.030 – 0.0 2.002 −0. From Table 5.048 −0. .032 −0.0 25. T2 ) by using the procedure outlined in Table 3 with upstream pressure p1 and downstream temperature T2 Estimation of upstream temperature T1 using the Eq. are shown in Fig.0 3.0 2.023 −0.012 −0.I.0 11. The differences between the calculated values and the corresponding measurement results [13]. The calculation procedures are implemented in the OOP mode.075 −0.092 −0.082 – −0.050 0.5 10.0 3.059 0.5 20.032 – −0.041 −0.171 −0.012 −0.034 0.039 −0. The individual and the combined relative errors due to the approximations of the temperature drop and the isentropic exponent are calculated.5 10.194 2 3 4 5 Table 6 Difference between the calculated and measured Joule–Thomson coefficient of a natural gas P (MPa): T (K) 0.032 −0.068 – −0.092 −0. The results of the measurements [13] and the results of the calculation of the specific heat capacities c p and the Joule–Thomson coefficients µJT of the natural gas mixtures.79942. with the corresponding high accuracy measurement data [13] (Ernst et al.0 16.5 15.066 −0.009 0.023 −0.061 0. Measurement results In order to compare the calculation results.030 −0.070 350 −0.023 −0.033 −0. The results are presented in the following section.0 5.5 15.096 – 0.024 – −0.0 4.059 −0.025 −0.0 20. xCO2 = 0. Calculated and measured Joule–Thomson coefficients of the natural gas mixture.098 – – −0.051 −0.043 0.134 −0.047 0.019 – 0.0 30.03000. for absolute pressures ranging from 0 to 30 MPa in 0.018 −0.016 −0.015 −0.033 −0. respectively. and xN2 = 0. xC2 H6 = 0.05029.069 −0. 3.026 – 0.022 0. 3 and Fig.002 0. respectively.093 0.033 – −0.048 – – – −0.075 −0.051 −0. 5. 4.022 −0.043 −0.0 7.0 17.069 – – −0.052 −0.081 −0.5 1.053 −0.036 – 0.0 25.02090.5 1.084 0.021 −0. are shown in Table 5 and Table 6.0 12.032 −0.020 −0.029 0. 4.025 0.0 30. for c p and µJT .020 −0.049 −0. we assume identical artificial natural gas mixtures with the following mole fractions: xCH4 = 0.0 7.022 −0. 300. for the specific heat capacity c p and the Joule–Thomson coefficient µJT .09939.5 13.097 −0.011 −0.). xC3 H8 = 0.005 350 −0. it can be seen that the calculated values of c p are within ±0. (41) Calculation of density ρ = Mρm and isentropic exponent κ using the procedure outlined in Table 3 with upstream pressure p1 and estimated upstream temperature T1 Calculation of viscosity (residual viscosity function) [12] Calculation of mass flow-rate in accordance with [10] 23 Table 5 Difference between the calculated and measured specific heat capacity at constant pressure of a natural gas P (MPa): T (K) 0.019 −0.027 −0.0 250 275 300 (c p calculated − c p measured ) (J/(g ∗ K)) −0.5 MPa steps and for four upstream temperatures (250.014 −0.

Temperature drop due to the Joule–Thomson effect T = µJT ω when measuring flow-rate of the natural gas mixture through the orifice plate with corner taps (ISO-5167-2). and n-pentane 0. the relative difference between the calculated and experimentally obtained c p never exceeds ±2. for absolute upstream pressures p1 ranging from 1 to 60 MPa in 1 MPa steps. but never exceeds ±0. Fig.00% and ±4. for absolute pressures p1 ranging from 1 to 60 MPa in 1 MPa steps. p (20 and 100 kPa). For pressures up to 12 MPa. downstream temperature T2 . Fig. The relative difference increases with the increase in pressure. viscosity. Calculated isentropic exponent of the natural gas mixture. Following the recommendations [2].0005. 7. 5. At higher pressures. The internal diameters of the orifice and pipe are: d = 120 mm and D = 200 mm. of the primary device. The error is calculated in accordance with Eq. and for six equidistant upstream temperatures T1 in the range from 245 to 345 K.085. 1. a natural gas mixture (Gas 3) form Annex C of [4] is assumed to flow through the orifice plate with corner taps [9] shown in Fig. In order to simulate a flow-rate measurement error caused by an inappropriate compensation for the adiabatic expansion.023. . ipentane 0. Fig. Fig. The approximate flow-rate and the corresponding natural gas properties (density. propane 0. At higher pressures. carbon dioxide 0. it can be seen that the calculated values of µJT are within ±0. the relative difference may increase significantly. Since the isentropic exponent is a theoretical parameter.24 I. and differential pressure p. 1) due to the Joule–Thomson effect is shown in Fig. The internal diameters of the orifice and pipe are: d = 120 mm and D = 200 mm. 6 it can be seen that for each temperature there exists the corresponding pressure where the Joule–Thomson coefficient changes its sign.015.0005. and isentropic exponent) are calculated at upstream pressure p1 .5% for pressures up to 12 MPa.113 K/MPa.2 J/(g ∗ K). The upstream pressure varies from 1 to 60 MPa in 1 MPa steps and upstream temperature from 245 to 305 K in 20 K steps for each of two differential pressures p (20 and 100 kPa). but never exceeds ±2. by neglecting the temperature drop due to the Joule–Thomson effect (T2 = T1 ).00%. The results of the calculations obtained for pure methane and methane–ethane mixtures are in considerably better agreement with the corresponding experimental data [13] than those for the natural gas mixture shown above. and downstream temperature from 245 to 305 K in 20 K steps for each of two differential pressures p (20 and 100 kPa). 6 illustrates the temperature drop due to the Joule–Thomson effect calculated in accordance with Eq. i-butane 0. (43) by comparing the approximate mass flow-rate (qm2 ) with the corrected mass flow-rate (qm1 ). Mari´ / Flow Measurement and Instrumentation 18 (2007) 18–26 c to 12 MPa. and the temperature downstream (T2 ).0035. 5 shows the results of the calculation of the isentropic exponent.0035. Fig.859. and consequently alters the sign of the temperature change. We estimate that the relative uncertainties of the calculated c p and µJT of the AGA-8 natural gas mixtures in common industrial operating conditions (pressure range 0–12 MPa and temperature range 250–350 K) are unlikely to exceed ±3. The upstream pressure varies from 1 to 60 MPa in 1 MPa steps. (41). From Fig. A relative error in the flow-rate measurements (Fig. up to 30 MPa. 6.00%. ethane 0.010. nitrogen 0. with the experimental results for the pressures up to 30 MPa. when the values of µJT are close to zero. respectively. The results are shown for two discrete differential pressures p (20 and 100 kPa). The results calculated are given for two discrete differential pressures. n-butane 0. From Table 6. 7. the difference increases. and for four equidistant downstream temperatures T2 in the range from 245 to 305 K. Relative error Er = (qm2 − qm1 ) /qm1 in the flow-rate of the natural gas mixture measured by the orifice plate with corner taps (ISO-5167-2) when using downstream temperature with no compensation for the Joule–Thomson effect (qm2 ) instead of upstream temperature (qm1 ). The procedure for the precise correction of the mass flowrate is shown in Table 4. A natural gas analysis in mole fractions is the following: methane 0. the absolute pressure is assumed to be measured upstream ( p1 ). there exist no experimental data for its verification.

It would therefore seem sensible for the results of the Joule–Thomson and the isentropic exponent calculations to be used with caution above this pressure. 6. or subsequent operational changes in temperature. 7) and the isentropic exponent (Fig. and differential pressure p. which use natural gas fugacities to derive the molar heat capacities and are. The procedure for the precise correction of the mass flow-rate is shown in Table 4. The error always increases by decreasing the operating temperature. The internal diameters of the orifice and pipe are: d = 120 mm and D = 200 mm. p I /(cm. this usually takes the form of a fixed value supplied by the user. Our further work will be directed towards the improvement of the model. could lead to significant systematic metering errors. The approximate flow-rate calculation is carried out in the same way. The upstream pressure varies from 1 to 60 MPa in 1 MPa steps and downstream temperature from 245 to 305 K in 20 K steps for each of two differential pressures p (20 and 100 kPa). The results are shown for two discrete differential pressures p (20 and 100 kPa). and up to 1. the . 9. therefore. The procedure for the precise correction of the mass flow-rate is shown in Table 4.5% at lower operating temperatures (245 K). 8 illustrates the relative error in the flow-rate measurements due to the approximation of the isentropic exponent by the ratio of the ideal molar heat capacities. The Fig. by neglecting the temperature drop due to the Joule–Thomson effect (T2 = T1 ) and by substituting the isentropic exponent with the ratio of the ideal molar heat capacities. p I − R). especially at lower temperatures and higher differential pressures. The internal diameters of the orifice and pipe are: d = 120 mm and D = 200 mm. Fig. However. The upstream pressure varies from 1 to 60 MPa in 1 MPa steps. which lies within the wider ranges of application given in [4]. like natural gases with a high carbon dioxide content.I. and for four equidistant downstream temperatures T2 in the range from 245 to 305 K. Relative error Er = (qm2 − qm1 ) /qm1 in the flow-rate of the natural gas mixture measured by an orifice plate with corner taps (ISO-5167-2) when using downstream temperature. (43). with no compensation for the Joule–Thomson effect and the isentropic exponent of the ideal gas at downstream temperature (qm2 ) instead of upstream temperature. The measurement error is also dependent on the natural gas mixture. and the corresponding real gas isentropic exponent (qm1 ). (43). pressure or gas composition. p I − R)). The total measurement error is still considerable. 9 it can be seen that the maximum combined error is lower than the maximum individual errors. The calculations in this paper show that any initial error in choosing this value. for absolute upstream pressures p1 ranging from 1 to 60 MPa in 1 MPa steps. p I /(cm. The results are shown for two discrete differential pressures p (20 and 100 kPa). 9 shows the flow-rate measurement error produced by the combined effect of the Joule–Thomson and the isentropic expansions. calculation results are shown up to a pressure of 60 MPa. Conclusion This paper describes the numerical procedure for the calculation of the natural gas molar heat capacity. because the Joule–Thomson coefficient (Fig. and no uncertainty is quoted in Reference [4] for pressures above 30 MPa. 8. For certain mixtures. The results obtained for the Joule–Thomson coefficient and isentropic exponent are in complete agreement with the results obtained when using the procedures described in [7] and [14]. and for four equidistant downstream temperatures T2 in the range from 245 to 305 K. downstream temperature T2 . κ = cm. the relative error in the flow-rate may increase up to 0. which equals the ratio of the ideal molar heat capacities (κ = cm. with the aim of simplifying the calculation procedure and further decreasing the uncertainty of the calculation results. the lowest uncertainty for compressibility is for pressures up to 12 MPa. for absolute upstream pressures p1 ranging from 1 to 60 MPa in 1 MPa steps. Whilst modern flow computers have a provision for applying a Joule–Thomson coefficient and an isentropic exponent correction to measured temperatures. The error is calculated by comparing the approximate mass flow-rate (qm2 ) with the corrected mass flow-rate (qm1 ) in accordance with Eq. with the exception of the isentropic exponent.0% at very low operating temperatures (225 K). Fig. for each of two differential pressures p (20 and 100 kPa). and downstream temperature from 245 to 305 K in 20 K steps. of 0–65 MPa. 8) show the countereffects on the flow-rate error. The error is calculated by comparing the approximate mass flow-rate (qm2 ) with the corrected mass flowrate (qm1 ) in accordance with Eq. and cannot be overlooked. considerably more computationally intensive and time consuming. From Fig. Relative error Er = (qm2 − qm1 ) /qm1 in the flow-rate of the natural gas mixture measured by the orifice plate with corner taps (ISO-5167-2) when using the isentropic exponent of the ideal gas (qm2 ) instead of the real gas (qm1 ). The approximate flow-rate and the corresponding natural gas properties are calculated at upstream pressure p1 . Mari´ / Flow Measurement and Instrumentation 18 (2007) 18–26 c 25 Fig.

2005. ISO-51671-2. Joule–Thomson coefficient and the isentropic exponent. Ref. 8. 1998. CORBA: Integrating diverse applications within distributed heterogeneous environments. The flow measurement engineering book. White Paper. Microsoft Corporation.nuc. Ref. Ref. Flow calorimetric results for the massic heat capacity cp and Joule–Thomson coefficient of CH4 . Calculation of natural gas isentropic c c exponent. Olander DR. Berkeley. Transmission measurement committee report no. [2] ISO-51671-1. of (0. Mari´ I. No. DIPPR R 801. Experiments in physical chemistry. ISO. ISO-51671-2:2003(E). ISO-51671-1:2003(E). Flow Measurement and Instrumentation 1997. from Department of Nuclear Engineering. Mari´ I. Sponsored by AIChE. and to Dr.edu/courses/classes/E-115/Reader/index. 2003. ISO-12213-2. 1992. New York: McGraw-Hill. Natural gas – calculation of compression factor – Part 2: Calculation using molar-composition analysis.26 I.16C2 H6 ). Urner G. 35(2):46–55. XQ 1285. University of California: Berkeley [Chapter 7]. The Joule–Thomson effect in natural gas flow-rate measurements. Design Institute for Physical Properties.85CH4 + 0. Microsoft: DCOM Architecture. Journal of Chemical Thermodynamics 2001. The author would like to express his thanks to Professor Donald R. Pressure loss of orifice plates according to ISO-5167. References [1] Shoemaker DP. Galovi´ A. Evaluated standard thermophysical property values. The corresponding equations have been derived by applying the fundamental thermodynamic relations to AGA-8 extended virial-type equations of state. Jaeschke M. Lemmon. USA. Measurement of fluid flow by means of pressure differential devices inserted in circular-cross section conduits running full — Part 2: Orifice plates. NIST. ISO. Ernst G. c Flow Measurement and Instrumentation 2005.16(1):13–20. University of California. No. Nibler JW. 2003. and of a mixture similar to natural gas. IEEE Communications Magazine 1997. for his valuable suggestions. Mari´ / Flow Measurement and Instrumentation 18 (2007) 18–26 c General principles and requirements.8:39–41. Miller RW. Education. 3rd ed. ISO-122132:1997(E). Šmuc T. No. The implementation of the procedure in an OOP mode enables its easy integration into new software developments. Arlington (Va. 1996. for his help in obtaining the relevant measurement data. Wirbser H. AGA catalog no. USA. Acknowledgements [3] [4] [5] [6] [7] [8] This work has been supported by the Croatian Ministry of Science. WEB site: http://www. 6th ed.htm. Vinoski S. AGA 8. Olander. ISO. Keil B. Garland CW.berkeley.). Retrieved September 21. The DIPPR AIChE generic ideal heat capacity equations have been used to calculate the ideal molar heat capacities of a natural gas mixture. and Sport with the projects “Computational Intelligence Methods in Measurement Systems” and “Real-Life Data Measurement and Characterization”. Eric W. An example of a possible application of the procedure in the flow-rate measurements has been given. Flow Measurement and Instrumentation 2005. New York: McGraw-Hill. In: Engineering thermodynamics. Measurement of fluid flow by means of pressure differential devices inserted in circular-cross section conduits running full — Part 2: [9] [10] [11] [12] [13] [14] . Thermodynamic relations.16:387–95. 2004. 1996. The procedure can be efficiently applied in both off-line calculations and real time measurements. 1997. Compressibility and supercompressibility for natural gas and other hydrocarbon gases.33:601–13.