Documente Academic
Documente Profesional
Documente Cultură
net/publication/229130195
CITATIONS READS
135 2,278
2 authors, including:
D. Yogi Goswami
University of South Florida
504 PUBLICATIONS 10,139 CITATIONS
SEE PROFILE
Some of the authors of this publication are also working on these related projects:
All content following this page was uploaded by D. Yogi Goswami on 16 April 2019.
Abstract
Ammonia–water mixtures have been used as working fluids in absorption–refrigeration cycles for several
decades. Their use as multi-component working fluids for power cycles has been investigated recently.
The thermodynamic properties required are known or may be calculated at elevated temperatures and
pressures. We present a new method for these computations using Gibbs free energies and empirical equa-
tions for bubble and dew point temperature to calculate phase equilibria. Comparisons of calculated and
experimental data show excellent agreement. 1999 Published by Elsevier Science Ltd. All rights reserved.
1. Background
Many studies have been published on vapor–liquid equilibrium (VLE) and the thermodynamic
properties of ammonia–water mixtures, including p–t–x–y data and caloric properties. For enthalpy
data, see Refs. [1–3]. Ref. [4] published new values of enthalpy and entropy from ⫺ 70 to 370°F
and pressure up to 300 psia using experimental data from [2,3,5]. Ref. [6] created tables of VLE
and caloric properties that were used by other researchers to propose computational models [7–
9]. In Ref. [10], measured data from [11] were used to give correlations for pressures of 0.2 to
110 bar and temperatures of 230 to 600 K. Refs. [12–16] also presented models for calculating
the thermodynamic data at elevated temperatures and pressures.
In the present study, a method that combines the Gibbs free energy method for mixture proper-
ties and bubble and dew point temperature equations for phase equilibrium is used. This method
0360-5442/99/$ - see front matter 1999 Published by Elsevier Science Ltd. All rights reserved.
PII: S 0 3 6 0 - 5 4 4 2 ( 9 9 ) 0 0 0 0 7 - 9
526 F. Xu, D.Y. Goswami / Energy 24 (1999) 525–536
combines the advantages of the two and avoids the need for iterations for phase equilibrium by
the fugacity method.
冕 冕 冕
T P T
where h0, s0, T0 and P0 are the specific enthalpy, specific entropy, temperature and pressure at
the reference state. Use of empirical relations for v and Cp [9] leads to the following equations.
For the liquid phase:
GLr ⫽ hLr,o ⫺ TrsLr,o ⫹ B1(Tr ⫺ Tr,o) ⫹ (B2/2)(T 2r ⫺ T 2r,o) ⫹ (B3/3)(T 3r ⫺ T 3r,o)
⫺ B1Tr ln(Tr/Tr,o) ⫺ B2Tr(Tr ⫺ Tr,o) ⫺ (B3/2)(T 2r ⫺ T 2r,o) ⫹ (A1 ⫹ A3Tr ⫹ A4T 2r)(Pr (2)
⫺ Pr,o) ⫹ (A2/2)(P2r ⫺ P2r,o).
For the gas phase:
Ggr ⫽ hgr,o ⫺ Trsgr,o ⫹ D1(Tr ⫺ Tr,o) ⫹ (D2/2)(T 2r ⫺ T 2r,o) ⫹ (D3/3)(T 3r ⫺ T 3r,o)
⫺ D1Tr ln(Tr/Tr,o) ⫺ D2Tr(Tr ⫺ Tr,o) ⫺ (D3/2)(T 2r ⫺ T 2r,o) ⫹ Tr ln(Pr/Pr,o) ⫹ C1(Pr (3)
⫺ Pr,o) ⫹ C2(Pr/T 3r ⫺ 4Pr,o/T 3r,o ⫹ 3Pr,oTr/T 4r,o) ⫹ C3(Pr/T 11
r ⫺ 12Pr,o/T r,o
11
⫹ 11Pr,oTr/T 12
r,o) ⫹ (C4/3)(Pr /T r ⫺ 12Pr,o/T r,o ⫹ 11Pr,oTr/T r,o).
3 11 3 11 3 12
Here, the superscripts are L for liquid and g for gas, while subscript o is for the ideal gas state.
The reduced (subscript r) thermodynamic properties are Tr ⫽ T/TB, Pr ⫽ P/PB, Gr ⫽ G/RTB, hr
⫽ h/RTB, sr ⫽ s/R and vr ⫽ vPB/RTB. The reference values for the reduced properties are R ⫽
8.314 kJ/kmol K, TB ⫽ 100 K and PB ⫽ 10 bar. The constants in Eqs. (2) and (3) are given in
Table 1.
The molar specific enthalpy, entropy and volume are related to Gibbs free energy, in terms of
reduced variables, by
h ⫽ ⫺ RTBT 2r 冋 ∂
(G /T )
∂Tr r r 册 Pr
, (4)
F. Xu, D.Y. Goswami / Energy 24 (1999) 525–536 527
Table 1
Coefficients of Eqs. (2) and (3)
s⫽⫺R 冋 册
∂Gr
∂Tr Pr
(5)
and
v⫽ 冋 册
RTB ∂Gr
PB ∂Pr Tr
. (6)
The Gibbs excess energy for liquid mixtures allows for deviation from ideal solution behavior.
The Gibbs excess energy of a liquid mixture is expressed by the relationship proposed in [9],
which is limited to three terms and is given by:
GEr ⫽ [F1 ⫹ F2(2x ⫺ 1) ⫹ F3(2x ⫺ 1)2](1 ⫺ x), (7)
where x is the ammonia mass fraction
F1 ⫽ E1 ⫹ E2Pr ⫹ (E3 ⫹ E4Pr)Tr ⫹ E5/T4 ⫹ E6/T 2r,
528 F. Xu, D.Y. Goswami / Energy 24 (1999) 525–536
hE ⫽ ⫺ RTBT 2r 冋∂
(GE/T )
∂Tr r r 册 ,
Pr, x
(8)
sE ⫽ ⫺ R 冋 册
∂GEr
∂Tr Pr, x
(9)
and
vE ⫽
PB ∂Pr冋 册
RTB ∂GEr
.
Tr, x
(10)
In addition, the enthalpy, entropy and volume of a liquid mixture are given by:
hLm ⫽ xfhLa ⫹ (1 ⫺ xf)hLw ⫹ hE, (11)
Table 2
Coefficients of Eq. (7)
E1 ⫺ 41.733398 E9 0.387983
E2 0.02414 E10 0.004772
E3 6.702285 E11 ⫺ 4.648107
E4 ⫺ 0.011475 E12 0.836376
E5 63.608967 E13 ⫺ 3.553627
E6 ⫺ 62.490768 E14 0.000904
E7 1.761064 E15 24.361723
E8 0.008626 E16 ⫺ 20.736547
F. Xu, D.Y. Goswami / Energy 24 (1999) 525–536 529
where subscripts a and w refer to ammonia and water, respectively and subscript f refers to the
saturated liquid condition.
Ammonia–water vapor mixtures are often assumed to be ideal solutions. The enthalpy, entropy
and volume of the vapor mixture are computed by:
hgm ⫽ xghga ⫹ (1 ⫺ xg)hgw, (15)
6. Vapor–liquid equilibrium
At equilibrium, binary mixtures must have the same temperature and pressure. Moreover, the
partial fugacity of each component in the liquid and gas mixtures must be equal:
f̂ La ⫽ f̂ga, (18)
f̂ Lw ⫽ f̂gw, (19)
where f̂ is the fugacity of each component in the mixture at equilibrium. The fugacities of
ammonia and water in liquid mixtures are given by [17]:
f̂ La ⫽ ␥af 0ax␦a (20)
and
f̂ Lw ⫽ ␥wf 0w(1 ⫺ x)␦w, (21)
where ␥ is the activity coefficient, f0 is the standard-state fugacity of the pure liquid component
corrected to zero pressure, ␦ is the Poynting correction factor from zero pressure to saturation
pressure of the mixture and x is the ammonia mass fraction in liquid phase.
Assuming an ideal mixture in the vapor phase, the fugacities of the pure components in the
vapor mixtures are given by
f̂ ga ⫽ aPy (22)
and
530 F. Xu, D.Y. Goswami / Energy 24 (1999) 525–536
Eqs. (24) and (25), developed in [14], determine the start and end of the mixture phase change
and compute the mass fractions of ammonia and water in the liquid and vapor phases, respectively.
This avoids the complicated method of calculating the fugacity coefficient of a component in a
mixture to determine the bubble (Tb) and dew point (Td) temperatures.
冘 冘
7 10
and
冘 冘
6 4
where
冘
4
冘
8
8. Results
In this study, the Gibbs free energy method is used to calculate the properties of pure ammonia
and water [Eqs. (2)–(6)]. The properties of the ammonia–water mixture are also calculated from
the Gibbs free energy method using Eqs. (7)–(17). In order to determine the phase quilibrium,
bubble and dew points are calculated using the alternative method of Eqs. (24)–(27) instead of
the conventional method of equating the fugacities [Eqs. (18)–(23)]. Using the alternative method
F. Xu, D.Y. Goswami / Energy 24 (1999) 525–536 531
avoids the iterative solution necessary to solve Eqs. (18)–(23), thereby reducing the compu-
tational time.
The property data generated in this study have been compared with available experimental and
theoretical data in the literature.
Fig. 1 shows that the bubble and dew point temperatures generated by this study compare
favorably with the data from Ref. [6]. The differences between our computed values and the data
are less than 0.3%. Refs. [9,10] are reported to have differences of up to 2% from these data.
Figs. 2 and 3 show the saturation pressures of ammonia–water mixtures as compared with the
data from Ref. [11].
For temperatures less than 406 K, the computational results fit the experimental data well,
except at saturated liquid pressures. At higher temperatures, our computed values are within 5%
of the data even at pressures higher than 110 bar, while Ref. [9] has reported a difference of more
than 15%. Ref. [10] reported an error of less than 5% under 110 bar and higher errors over 110 bar.
1. Saturated liquid enthalpy. The saturated liquid enthalpy of this work is compared with the
data from Ref. [6], as shown in Fig. 4. The differences are less than 2% for all the data.
2. Saturated vapor enthalpy. The saturated vapor enthalpy at constant pressure is shown in Fig.
5. The agreement with the data is within 3%. Ref. [10] reported a 5% maximum difference.
The mass fraction of ammonia vapor shown in this figure is the ammonia liquid mass fraction
when the mixture reaches a saturated state. So, in order to compute the saturated vapor enthalpy,
the ammonia vapor mass fraction must be determined first.
The value of entropy is very important in predicting the performance of a turbine in a power
cycle. Entropy data are also essential to the second-law analysis of thermal systems. Ref. [4]
published saturated liquid and vapor entropy data based on experimental data from [2,3,5]. Ref.
[16] published calculated entropy. The entropy data from the present study are compared with
the experimental data in Ref. [4] and the computational data of Ref. [16].
1. Saturated liquid entropy. Fig. 6 shows saturated liquid entropy data compared with those of
Ref. [4]. Our data agree with the experimental data of [4] much better than the data generated
by the method of Ref. [16].
2. Saturated vapor entropy. Fig. 7 shows an excellent agreement of our computed values of
saturated vapor entropy with the data of Ref. [4]. Data computed by Ref. [16] are consistently
lower. Since it was very difficult to identify saturated vapor entropy data from Ref. [16], we
did not compare our results with them.
13. Conclusion
Different methods for calculating the properties of ammonia–water mixtures are studied. A
practical and accurate method is used in this study. This method uses Gibbs free energy equations
for pure ammonia and water properties, and empirical bubble and dew point temperature equations
for vapor–liquid equilibrium. The iterations necessary for calculating the bubble and dew point
temperatures by the fugacity method are avoided. Therefore, this method is much faster than
using the fugacity method. The computational results have been compared with accepted experi-
mental data in the literature and show very good agreement.
References
[1] Jennings BH, Shannon FP. Refrig Eng 1938;44:333.
[2] Zinner KZ, Gesamt Z. Kalte-Ind 1934;41:21.
536 F. Xu, D.Y. Goswami / Energy 24 (1999) 525–536