Diploma Thesis: Modelling of Mercury (HG) Distribution in Natural Gas Mixtures

Diploma Thesis
Modelling of Mercury (Hg) Distribution in Natural

Gas Mixtures
Christos Mentzelos
Supervisor: Dr. Epaminondas C. Voutsas
1
Table of contents
Summary (Greek)......................... 5
Summary (English) ......................... 8
Acknowledgments ......................... 10
Chapter 1: Introduction .............. 11
Chapter 2: Theory .................... 15
2.1 Forms of mercury in natural gas ............ 15
2.2 Mercury distribution on gas processing plants ......... 16
Chapter 3: Evaluation of experimental data (critical properties, vapor pressure and

solubility) .................................... 20
3.1 Available Tc, Pc, for Hg in literature ............... 20
3.2 Creation of Ps database and Solubility database ................. 20
3.2.1 Evaluation of Ps experimental data ................. 20
3.2.2 Evaluation of DIPPRs equation for the prediction of Ps .. 22
3.2.3 Development of database for binary Vapor-Liquid-Equilibrium (VLE)

and Liquid-Liquid-Equilibrium (LLE) data of Hg with HC ........ 24
3.2.4 Experimental errors of the data assembled ...... 32
Chapter 4: Thermodynamic modeling ........................... 37
4.1 SRK and PR EOS ................................ 37
2
4.2 Evaluation of SRK and PR EOS with two different sets of T c, Pc, .. 38
4.3 Calculation of the Mathias-Copeman parameters for Hg for the PR EOS .. 42
4.4 Calculation of Twu parameters for Hg for the SRK EOS .. 44
4.5 Estimation of the binary interaction parameters (kij) for binary systems of Hg
with hydrocarbons ................................. 46
4.5.1 Calculation of the kij parameters .................. 46
4.6 The UMR-PRMC model ........................... 53
4.6.1 Main groups ACH and ACCH3 ..................... 55
4.6.2 Main group CH2 ...................... 58
4.6.3 Main group cy-CH2 ....................... 63
4.6.4 Main group CH4 ...................... 66
4.6.5 Main group C2H6 ....................... 68
4.6.6 Main group CO2 .............................. 70
4.6.7 Main group N2 ................................ 72
4.6.8 Main group H2O ............................. 73
4.7 Generalized correlations for the binary interaction parameters based on the
Tb and the CN of the HC .......................... 75
4.7.1 Evaluation of the generalized correlations for the binary interaction

parameters based on the Tb and the CN of the HC ............... 78
Chapter 5: Testing of PR-MC, SRK-Twu and UMR-PRMC model in binary and

multicomponent mixtures ............................... 82
5.1 Iso-butane .................................. 82
5.2 K variable ....................................... 84
5.2.1 K variable for ethane ................................. 84
5.2.2 K variable for propane ............................... 85
5.2.3 K variable for CO2 ...................................... 87
3
5.2.4 K variable for iso-butane ......................... 88
5.3 Multi-component mixtures .................................... 90
Chapter 6: Conclusions .................................... 95
Chapter 7: Future work ......................................... 98
Appendix A .......................................................... 99
Appendix B .......................................................... 109
Appendix C .......................................................... 117
Appendix D .......................................................... 128
Appendix E ........................................................... 130
Appendix F ........................................................... 160
Appendix G .......................................................... 175
Appendix H .......................................................... 179
Appendix I ............................................................ 183
Appendix J ........................................................... 185
*Note: At the end of each chapter, its respective references are being presented
(Hg)
quicksilver hydrargyrum. ,
,

.

.

,
, -
product streams- .
Hg
, , ,
.

Hg
.

Hg .

.
, Peng-Robinson (PR) Soave-Redlich-Kwong (SRK)

(EOS)
.
Hg, Mathias-Copeman
(MC) Twu-Coon (Twu) PR SRK EOS ,
Hg.

PR-MC SRK-Twu EOS.

Hg (HC),
Hg HC,
.
Hg , Hg
, .
Hg
5

.

Hg HCs
.
PR-MC SRK-Twu EOS.

(CN) HCs
(b).
Hg
.
, UMR-PRMC
.
PR EOS Mathias-Copeman
universal mixing
rules.
Unifac .

Hg

, ,
Hg ,
Hg ,

.
.

.

,
Hg ,
.

,
.

.
6
.
, PR-MC, SRK-Twu EOS UMR-PRMC,
Hg HCs
.

. T
,

-.
CN
.

,
.

,
HCs IUPAC
tentative, .
7
Summary
Mercury (Hg) is a toxic chemical element, commonly known as quicksilver or

hydrargyrum. Found in all hydrocarbon mixtures, mercury has been recorded in
natural gas and oil fields all over the world. Significantly high levels of mercury
concentration can be found in the natural gas with light hydrogen isotopes of
methane and at lacustrine dispositional environment. Mercury can cause significant
problems to processes of the gas industry such as aluminum heat exchanger failures,
poisoned catalysts, contaminated product streams and gas leaks.
Given the problems that Hg can cause to the industry and the human health, an
investigation on the matter is required. The scope of this thesis is the development
of thermodynamic models which are able to predict the elemental Hg distribution in
binary and multicomponent mixtures. This way the developed and evaluated models
can be used later on, for the prediction of the Hg distribution in real processes. That
will be the most efficient way of dealing with this component and the problems
caused from it.
To this purpose, the Peng-Robinson (PR) and the Soave-Redlich-Kwong (SRK) cubic
equations of state (EOS) were examined by employing different expressions for the
alpha parameter of the EOS. For this parameter, Mathias-Copeman (MC) and Twu-
Coon (Twu) pure components parameters were developed for Hg for PR and SRK EOS
respectively by fitting them to vapor pressure data of elemental Hg. These models
are symbolized as PR-MC and SRK-Twu EOS. Except for those parameters, binary
interaction parameters for Hg and hydracarbons (HC) were developed as well by
fitting them to experimental solubility data of Hg in binary mixtures with various
HCs, which are also evaluated.
For all binary mixtures of Hg with other components, except the one containing Hg
with water, constant binary interaction parameters have been developed. As far as
the binary mixture of Hg with water is concerned, temperature dependent
parameters have been developed due to the models weakness regarding the
description of the effect of the hydrogen bond.
Furthermore two predictive correlations have been developed, in order to estimate

the binary interaction parameters for binary mixtures of Hg with HC for which no
experimental data are available in the literature, for both the PR-MC and the SRK-
Twu EOS. These correlations have been based on the carbon number (CN) and also
the normal boiling temperature (Tb). In addition to that, the generalized correlations
have been tested against experimental data regarding a ternary and a
multicomponent mixture.
8
Another model developed is the UMR-PRMC proposed by Voutsas et al. This model
uses as its basis the Peng-Robinson EOS with the Mathias-Copeman parameters, but
instead of the classical mixing rules, it introduces the universal mixing rules. These
mixing rules use the Unifac contribution group parameters instead of binary
interaction parameters.
Finally the three models developed as well as their corresponding generalized

correlations have been evaluated against experimental data concerning two
multicomponent mixtures, as well as the experimental K variable, that is the
division of the Hg mole fraction in the vapor phase with the Hg mole fraction in the
liquid phase, where possible.
These evaluations have three objectives. The first one is to test how the developed
models cope against more complex systems than just binary mixtures. That way it
will be safer to conclude whether or not the models can be used in process
simulations at a later stage and provide reliable results, given the fact that there are
no experimental data regarding real processes. The second one is to explore the
capability of the models to predict the K variable, which is very important when it
comes down to evaluating models in a process. The last one is to find the optimum
type of the generalized correlations estimating the binary interaction parameters
that should be used for process simulations in the future.
The results can be summarized as follows. First of all every model, namely the PR-
MC, the SRK-Twu EOS and the UMR-PRMC are able to correlate elemental Hg in
binary mixtures and also accurately predict its mole fraction in multicomponent
mixtures. Secondly all models are capable of predicting the K variable of the
corresponding mixtures satisfyingly enough. Thirdly as far as the optimum type of
the generalized correlation is concerned, both types appear to provide very similar
results in all cases, except for the mixtures containing iso-butane, where the
correlations based on the CN seem to have the advantage. Therefore it is up to the
user to decide which type is appropriate according to the occasion, since both types
provide reliable results. Finally it can be concluded that all models and generalized
correlations can be used in process simulations at a future stage with some caution
regarding the branched alkanes and the aromatic HCs due to their tentative
experimental data, providing reliable results.
9
Acknowledgments
This master thesis was accomplished at the Norwegian University of Science and
Technology (NTNU) and the National Technical University of Athens (NTUA). My
primary supervisors have been professors Even Solbraa and Epaminondas C.
Voutsas, and my research advisors Mr. Efstathios Skouras and Ms. Andrea Carolina
Machado Miguens from Statoil.
I would like to express my sincere thankfulness to my supervisors and all my

research advisors for their helpfulness during my period of work at Statoil. To
emphasize my gratefulness I dedicate to them the words of my Greek ancestor
Alexander the Great:
I am indebted to my father for living, but to my teachers for living well.
I would also like to thank the Erasmus program and all people supporting it, for
giving me the opportunity to go to the NTNU in order to do my master thesis and
Statoil for helping me out during my stay at Trondheim and providing me with
various resources for the completion of my master thesis.
Finally I would like to thank my family, my closest friends Mitrou Michalis,

Dimopoulos Ilias, Kokkinos Dimitrios, Chatziloudis Charalampos, Petsagourakis
Giannis and Thomas George for always supporting me in my work.
10
Chapter 1: Introduction
Mercury is the chemical element symbolized as Hg and has the atomic number 80
and molecular weight of 200.59 g/mole. It is commonly known as quicksilver or
hydrargyrum. It is a heavy, silver colored, d-block element, and furthermore it is the
only liquid metal at standard temperature and pressure (STP) conditions. It has a
freezing point of -38.83 oC and boiling point 356.73 oC [1]. It is a poor conductor of
heat but a fair conductor of electricity. It exists in two main oxidation states, I and II.
The metal is relatively stable in dry air, but in moist air tarnishes slowly to form a
gray oxide coating.
The corrosive properties of mercury in the natural gas industry were firstly reported
at 1973, when an aluminum-heated transformer at a liquid natural gas plant of
Algeria was totally destroyed. After investigation was conducted on the matter it
was found out that corrosion caused by mercury in the natural gas, was the cause of
the damage [2].
Found in all hydrocarbon mixtures, mercury has been recorded in natural gas and oil
fields all over the world. Generally significantly high levels of mercury concentration
are found in the natural gas with light hydrogen isotopes of methane and at
lacustrine dispositional environment. Mercury can cause significant problems such as
aluminum heat exchanger failures, poisoned catalysts, contaminated product
streams, contaminated waste water and gas leaks [3].
Specifically about the equipment used in a natural gas process, mercury readily
forms alloys - amalgams- with a variety of metals such as aluminum, brass,
chromium, copper, zinc, iron, nickel, silver, gold and tin that are weaker than
mercury-free metals, causing mechanical failures and gas leaks. Because it does not
amalgamate with iron, mercury is often stored in containers made out of this
material.
Liquid metal embrittlement (LME) is a complex metal fracture mechanism that starts
without a warning in advance. Liquid mercury has a disastrous effect on aluminum. It
can be traced in natural gas feed stock, usually at very low levels, none the less it can
accumulate in quantities sufficient to cause failure of cryogenic aluminum heat
exchanger. Mercury can degrade the aluminum cold boxs materials -where the
cryogenic distillation happens- by three different mechanisms.
1) Amalgamation
This is the process by which mercury forms liquid solutions with various
metals, mainly with Al, Au, Ag and Zn. If the metal is Al, its concentration in
11
the amalgam is relatively low, therefore the depth of the corrosion is limited.
Furthermore aluminum is generally prevented from coming in contact with
mercury by the Al2O3 protective surface oxide -for the amalgam reaction to
begin, mercury must wet the aluminum metallic surface first-. The oxide
aluminum is not homogeneous and contains numerous defects, but in
general mercury will not migrate through these microscopic cracks and
defects to reach the underlying metal. However, if the extent or severity of
such defects is increased by thermal or mechanical stresses, abrasion or
chemical environments, the risk of mercury damage is increased.
2) Amalgam corrosion
This is the combined action of mercury and moisture producing a corrosion

process that propagates with miniscule amounts of mercury. The reaction has
two steps and is as follows:
Hg + Al Hg(Al) amalgam (1)
Hg(Al) + 6 H2O 2AlO3 3H2O + 3H2 + Hg (2)
Small amounts of aluminum can be dissolved in liquid mercury, diffused to a

mercurymoist air interface, and then rapidly oxidized. Since, oxidation
removes aluminum from the mercury, further aluminum can dissolve, and
the process can continue until the aluminum is completely converted to
oxide. Rapid oxidation requires the presence of moisture. Reaction rates are
slow in its absence.
3) LME
LME is generally much more severe and therefore more important than other
embrittling processes, such as hydrogen-embrittlement or stress-corrosion
cracking. Once cracks have initiated, very rapid subcritical cracking can occur
even at low stresses. Cracking occurs preferentially along grain boundaries
for the Al:Hg couple (and for many other couples), but transgranular
fractures can also occur. Liquid metals are drawn into growing cracks so that
the crack tip is always in contact with embrittling metal atoms. The rate-
controlling process for cracking is still being debated, but the rate of flow of
liquid within cracks may control the rate of cracking in some circumstances.
Adsorption of embrittling atoms at crack tips weakens substrate interatomic

bonds, and facilitates crack growth by enabling interatomic bonds to break or
12
shear more easily than in inert environments. Thin films of liquid metal are
left behind the advancing crack tip and, hence, fracture surfaces are covered
with a film of liquid metal. For the Al:Hg system, de-wetting can occur so
that small globules of mercury are present on fracture surfaces. The presence
of mercury on fracture surfaces can also result in the growth of oxide
whiskers after fracture, a phenomenon peculiar to aluminum and discussed
in the preceding section.
For LME cracks to initiate there must be intimate contact between liquid and
solid metals, with no intervening oxide films to prevent wetting and
adsorption. Al alloys are covered by a thin, protective oxide film, and surfaces
can be covered by liquid mercury indefinitely without any reaction until the
oxide is damaged. Oxide films can be broken by mechanical processes, e.g. by
scribing or abrasion, by chemical processes, e.g. corrosion, or by plastic
deformation of the aluminum resulting in slip steps at the surface.
The aluminum amalgam in particular is extremely corrosive since it reacts with the
moisture to form a metal oxide, releasing the mercury to perpetuate the corrosive
process. This corrosion can lead to catastrophic failure of cryogenic heat exchanger
and wellhead valve stems and gates. [4, 5]
It is obvious that Hg can pose a major problem for the natural gas industry due to its
properties and its behavior. Therefore an investigation on the matter is required.
This way the most efficient way of dealing with this component and the problems
caused by it can be detected and solved.
The scope of this master thesis is the development of thermodynamic models that
will be able to predict the elemental Hg solubility in multicomponent mixtures. To
this purpose, the Peng-Robinson (PR) [6] and the Soave-Redlich-Kwong (SRK) [7]
cubic equations of state (EOS) will be examined by employing different expressions
for the alpha parameter of the EOS. For this parameter, Mathias-Copeman [8] and
Twu-Coon [9] pure components parameters will be developed for Hg for PR and SRK
EOS respectively by fitting them to vapor pressure data of elemental Hg. Except for
those parameters, binary interaction parameters for Hg and hydracarbons (HC) will
be developed as well by fitting them to solubility data of Hg in binary mixtures with
various HC. Furthermore the UMR-PRMC model proposed by Voutsas et al [10] will
be examined as well. This model belongs to the EOS/GE class of models and
combines the PR EOS [6] with Mathias-Copeman [8] parameters, with an original
Unifac-type GE model that employs temperature dependent group interaction
parameters which are estimated by fitting them to solubility data of Hg with HCs and
other components.
13
Two different types for the binary interaction parameters will be developed.
Constant ones for all binary mixtures except the one containing Hg with water. For
this particular mixture temperature dependent parameters will be introduced. In
addition there will also be presented two predictive correlations in order to estimate
them for binary mixtures of Hg with HCs for which no experimental data are
available in the literature. Finally the models will be tested with one binary, one
ternary and two multicomponent mixtures.
References:
[1] http://www.chemicool.com/elements/mercury.html
[2]http://www.americanlaboratory.com/913-Technical-Articles/114246-
Determination-of-Mercury-in-the-Environment-Petrochemical-Products-and-Fuels/
[3] Wilhelm M., Bloom N.(2000) , ''Mercury in petroleum'', Fuel Processing

technology, vol. 63, i.1, p. 1-27
[4] Coade R., Coldham D. (2006), ''The interaction of mercury and aluminium in heat
exchangers in a natural gas plants'' ,International Journal of Pressure Vessels and
Piping, Vol. 83, I. 5, p. 336342
[5] Wilhelm M. (2009), ''Risk Analysis for Operation of Aluminum Heat Exchangers
Contaminated by Mercury'', Process Safety Progress, Vol. 28, Issue 3, p. 259266
[6] Peng, D. Y., and Robinson, D. B. (1976). "A New Two-Constant Equation of State".
Industrial and Engineering Chemistry: Fundamentals 15, p. 5964
[7] Soave (1972), G. Equilibrium Constants from a Modified RedlichKwong Equation

of State, Chem. Eng. Sci., 27, p. 1197-12 3
[8] Otilio Hernndez-Garduza, Fernando Garcia-Snchez, David pam-Martnez,

Richart Vzquez-Romn (2002), Vapor pressures of pure compounds using the Peng-
Robinson equation of state with three different attractive terms, Fluid Phase
Equilibria, p. 195-228
[9] Twu C. et al (1995), ''A new generalized alpha function for a cubic equation of
state Part 2. Redlich-Kwong equation'', Fluid Phase Equilibria 105, p. 61-69
[10] V. Louli et al. (2012), ''Measurement and prediction of dew point curves of
natural gas mixtures '', Fluid Phase Equilibria 334, p. 2-4
14
Chapter 2: Theory
As far as the electron configuration of mercury is concerned, electrons fill up all the
available 1s, 2s, 2p, 3s, 3p, 3d, 4s, 4p, 4d, 4f, 5s, 5p, 5d and 6s subshells. Since this
configuration strongly resists removal of an electron, mercury behaves similarly to
noble gas elements, which form weak bonds and become solids which melt easily at
relatively low temperatures. The stability of the 6s shell is due to the presence of a
filled 4f shell. An f shell poorly screens the nuclear charge that increases the
attractive Coulomb interaction of the 6s shell and the nucleus. [1]
As far as the critical properties of mercury are concerned, meaning its critical
temperature and critical pressure, there has not been an agreement yet. The values
of these properties vary in the literature and the experiments that have been
conducted to determine them do not agree with each other. More details about
these properties will be discussed in the 3rd chapter. The values which are widely
used however are 1735.15 K and 1608 bar respectively.
2.1 Forms of mercury in natural gas

Mercury can be present in a natural gas mixture mainly in the elemental form (Hg 0)
due to its volatility, oxidized (Hg+1) and (Hg+2) form, organic or inorganic ionic forms.
Furthermore it is present at concentrations way below saturation. That fact indicates
that there is no liquid phase mercury in most reservoirs. All these types of mercury
bare unique species-dependent physical, physiological and chemical properties.
Mercury can form two kinds of compounds called:
1. Mercurous, when it uses two electrons in the bonding process
2. Mercuric, when it uses just one electron to bond with another element
Although mercury emissions from the natural gas sector have not been widely
calculated or reported, information and data about natural gas processing certainly
indicate that the natural gas sector could be a significant source of both global
mercury supply and emissions.
Organic compounds contain mercury at the +2 oxidation state. They include

organometallic compounds -with a Hg-C covalent bond-. Organic mercury forms
consist of two main groups:
1. R-Hg-X compounds -partly alkylated species-
2. R-Hg-R compounds -fully alkylated species-,
15
Where R stands for organic species, of which methyl (-CH3) is prominent, and X
stands for inorganic anions, such as chloride, nitrate or hydroxide. The R-Hg-X group
includes mainly monomethylmercury compounds. The most prominent R-Hg-R
compound is dimethylmercury (CH3HgCH3). Other examples of these categories are
C2H5HgC2H5, C3H7HgC3H7 and ClHgCH3 .
As far as the inorganic forms of mercury are concerned, there are two categories.
One contains the ionic mercury salts, which can be Hg2+X or Hg2+X2 , where X
symbolizes an inorganic ion for the first one and the other one contains mercury
salts (mostly halides) like mercuric chloride, which are soluble in gas condensates,
but they prefer to partition to the water phase in primary separations.
Gas processing can cause transformation of one chemical form of mercury to

another. A characteristic example is the reaction of elemental mercury with sulfur
compounds. The mixing of gas and/or condensate from sour and sweet wells allows
reaction of elemental mercury with S2 or ionic mercury with H2S to form HgS that can
settle out in tanks and deposit in equipment, or become attached to suspended
particles with small particle size. [2, 3, 4, 5]
2.2 Mercury distribution on gas processing plants

Elemental mercury and organic compounds of mercury can be found in many gas
fields. Table 1 shows some average concentrations of Hg in natural gas fields of
several places over the globe.
Table 1: Regional Average Level of Mercury in Natural Gas
Location Elemental Hg concentration [g/m3]

South America 69-119
East Asia 58-193
North Africa 0.3-130
North Europe 0.01-180
Middle East 1-9
Eastern US Pipeline 0.019-0.44
Midwest US Pipeline 0.001-0.1
North America 0.005-0.04
As already mentioned Hg0 has a normal boiling point of 356.73 oC and it would
therefore be expected to have a limited distribution in a gas processing plant.
However due to its mobility and its ability to be readily bonded to most of the
surfaces that it comes in contact with, this is not the case. Since mercury can be
bonded to gas gathering pipelines it is possible that it may take some time until it
arrives at the main plant. Therefore problems may be caused since it was not
16
expected and no precautionary measures were taken in advance. Table 2 shows the
change in the concentration of Hg in feed gas with time for two different plants.
Table 2: Change in the concentration of Hg in feed gas with respect to time (g/m 3)
Start up Year 7
Plant A 0.25 to 0.45 17 to 69
Plant B 0.01 0.253
Especially for Plant A the difference in the concentration of Hg is very significant. In

case the scale of this difference was not known beforehand, there could have been
some very serious damage to the pipelines and the equipment used.
This high mobility of Hg implies that, if not removed at source, it will distribute
throughout the whole process streams and in plant effluents. The course of mercury
in gas processing is easier to predict since the process is simpler than that of oil
processsing. Therefore no transformation of the elemental mercury to the species
that were mentioned above is caused. Gas processing does not include molecular
transformations to come up with final products, it is in simple terms put, a treatment
and separation process. The treatment part has to do with the removal of unwanted
constituents like CO2, H2S and H2O and contaminants -metals-. The separations are
typically cryogenic utilizing selective condensation of fractions (C2, C3, C4, C4+) by
removal of heat. It is a fact that some heavy condensate streams that are used in gas
separation processes can contain amounts of suspended and oxidized forms of
mercury as well as elemental mercury.
There have been a number of surveys by Johnson Matthey Catalysts on gas

processing plants in several continents around the world [6]. Table 3 presents a
typical example of the distribution of Hg on a 50 Million standard cubic feet per day
(mmscfd) gas plant in the Far East as a steady-state flow.
Table 3: Distribution of Hg on a 50 mmscfd Gas Plant
Process Stream Mercury [kg/year]

Raw Gas 220
Acid Gas Removal Vent 22
Dryer Vent 3
Condensate 45
Sales Gas 150
Measurements on the raw gas of all plants of the survey revealed that the
concentration of Hg ranged from 10 up to 70000 ng/m3. The researchers Peter J. H.
Carnell and Paul J. Openshaw, Johnson Matthey Catalysts [6] mention that even in
the case of the plant with the lowest concentration of Hg in the initial feed of the
17
process, a small portion of it was present in both the sales gas and the condensate.
As far as the acid gas removal and sulphur recovery is concerned, Hg was found in
the rich amine liquor, in the stripper gas and in the elemental sulphur.
Finally as far as the product streams on all plants, meaning the sales gas and the
condensate streams, are concerned, all of them displayed Hg concentration. That
means that Hg is able to distribute throughout the whole process plant and not only
in the condensate streams. Naturally the highest concentrations of Hg were found
on the plants with no mercury removal units. However even in the presence of these
units a significant percentage of the total Hg reached the rest of the process [6].
The distribution of Hg on the gas processing plants that were examined on the
survey is shown in figure 1.
Figure 1: The distribution of Hg on the gas processing plants examined on the survey
References:
[1] Norrby, L.J. (1991). "Why is mercury liquid? Or, why do relativistic effects not get
into chemistry textbooks?". Journal of Chemical Education 68 (2), p. 110
[2] Frech W. et al (1996), ''Determination and Speciation of Mercury in Natural Gases

and Gas Condensates '', Analytical Communications, vol. 33 (7H-9H).
[3] Wilhelm M. (2000), Bloom N., ''Mercury in petroleum'', Fuel Processing

technology, vol. 63, i.1, p. 1-27
18
[4] Wilhelm M. (2001), ''MERCURY IN PETROLEUM AND NATURAL GAS: ESTIMATION
OF EMISSIONS FROM PRODUCTION, PROCESSING, AND COMBUSTION.'' U.S.
Environmental Protection Agency, Washington, D.C., EPA/600/R-01/066
[5] http://www.frontiergs.com/files/Measuring-Mercury-in-Natural-Gas.pdf
[6] Carnell P. et al (2004),'' MERCURY DISTRIBUTION ON GAS PROCESSING PLANTS '',

83rd annual GPA convention proceedings, New Orleans.
19
Chapter 3: Evaluation of experimental data (critical
properties, vapor pressure and solubility)
3.1 Available Tc, Pc, for Hg in literature

Throughout the literature search two different sets of critical temperature (Tc),
critical pressure (Pc) and acentric factor () for Hg have been encountered. One of
them was proposed by DIPPR [1] and the other one by a report from NIST [2], whose
purpose was to collect and evaluate all experimental data of P s of Hg as well as
create an equation capable of describing them. Table 4 presents the two sets.
Table 4: DIPPRs and NISTs sets of Tc, Pc and
DIPPRs set NISTs set

Tc [K] 1735.15 1764.15
Pc [bar] 1608 1670
-0.1645 -0.1911
NISTs set of Tc and Pc was chosen from experimental data regarding those
properties. In the report is stated that there has been a literature research on the
matter and that the values considered the most reliable among the existing
experimental data, were those of Kozhevnikov [3], who estimated that Tc = 1764 K
and Pc = 1670 bar.
The was calculated afterwards using equation 1. The vapor pressure (P s) of Hg at Tr

= 0.7 was calculated using DIPPRs equation. Tr is defined as T divided by the Tc. The
choice of this equation will be analyzed thoroughly later on.
Given the fact that there still is a debate about the value of these properties it was
decided that these two sets should be checked, at least for the effect that they have
on the calculation of the Ps while using the SRK and PR EOS.
( Tr ) (eq.1)
3.2 Creation of Ps database and Solubility database
3.2.1 Evaluation of Ps experimental data
Any thermodynamic model should be able to predict the behavior of a given mixture
if it is to be considered suitable for implementation in an industrial process. This
means that it has to be able to predict the phases of the mixture as well as their
20
composition given the temperature (T), the pressure (P) and an initial composition.
Therefore it is essential that the models can predict, on a first level, accurately the P s
of each component of the mixture in order to be able to predict correctly the phase
equilibrium. For the models of this thesis it is of interest that they will be able to
predict the Ps of Hg, which is the element in question.
The Ps of Hg is very important to the thermodynamic modeling that will be

performed in this master thesis, because it will be used later on for the fitting of the
Mathias-Copeman and the Twu-Coon parameters which are used in the alpha
parameter of the EOSs as will be explained in chapter 4.
There have been several attempts to measure the P s of Hg for several temperatures
over the previous centuries from many researchers like Ernsberger, Menzies,
Ambrose and Douglas. All these researchers are mentioned in the report from NIST
[2]. However taking into consideration the chronicle periods that these experiments
were conducted it is obvious that the results alter depending on the means available.
Therefore the uncertainty level of the measurements poses a crucial factor in
deciding which data will be used for the creation of the database.
Most of the experimental data that are chosen to build a database for the Ps of Hg,
deal with the temperature range of 285 K up to 900 K. For that temperature range
these data have an uncertainty level of less than 1%.
It is noted that at lower temperatures the uncertainty level of the measurements is

higher than that of the higher temperatures. This happens because the P s of Hg is
getting lower as the temperature declines. However since the purpose of this thesis
is to predict the distribution of Hg throughout a process plant for the natural gas
industry it is essential that the temperature range of the P s also includes
temperatures below 285 K, which is the lowest temperature for which the
uncertainty level of the experimental data is less than 1%.
Therefore Ps experimental data for temperatures lower than 285 K and for
temperatures higher than 900 K are added to the database. As far as the higher
temperature region is concerned (1052 K 1735 K) due to the lack of many
experimental data, the data of Schonherr [2] are chosen. These data have a
maximum uncertainty level of 3% for the highest temperature. As far as the lower
temperature region is concerned (234 K 275 K) the data of Douglas et al. [2] are
chosen. These display a maximum uncertainty level of 1.5% at the lowest
temperature.
Given these facts the database that was created has experimental P s for the
temperature range from the freezing temperature up to the critical temperature of
Hg. The database is presented analytically at appendix A, at table A1.
21
3.2.2 Evaluation of DIPPRs equation for the prediction of Ps
As it can be seen at the database at table A1, the experimental data concerning the
Ps of Hg are not evenly spread through the whole temperature range. Therefore it is
extremely important to have a way -via an equation- to predict Ps almost equal to
the available experimental data. Thus one will be able to use this equation in order
to reliably estimate Ps for temperatures where no data are available. That way the P s
for the whole temperature range will be evenly described.
The equation from DIPPR (eq.2) for calculating the P s of elemental Hg was tested for
that purpose.
) )
eq. (2)
Where:
A = 30.951, B = -7717, C = -1.1296, D = 2.6938*10-7, E = 2
and P = [Pa] T = [K]
DIPPR equations temperature range is mentioned to be from the triple point

(234.31 K) up to the critical temperature of Hg (1735 K). At figure 2 it can be seen
that the results from eq. 2 are in very good agreement with the experimental data.
The name of each researcher as well as his/her experimental data are presented as
points on the figure.
The greatest deviation noted was that for temperature 1051 K and was equal to
3.69%. One must keep in mind however that for these high temperatures the
uncertainty level of the experimental data is around 3%, therefore the results of eq.
2 for the whole temperature range can be characterized as acceptable. In general
the deviations that the equation displays towards the experimental data are less
than 1% except for the temperatures above 1000 K. Table 5 presents some
characteristic results of DIPPRs equation and their deviation from the experimental
Ps. More details concerning the deviations of eq. 2 and the experimental data can be
found at the appendix A at table A1, where the full version of table 5 is located.
22
1.00E+07
1.00E+06
1.00E+05
1.00E+04
1.00E+03
1.00E+02 Dippr
1.00E+01 Ambrose
P [kPa]
1.00E+00 Beattie
1.00E-01 Ernsberger
1.00E-02 Menzies
1.00E-03 Schonherr
1.00E-04 Shpil'rain
1.00E-05 Spedding
1.00E-06 Douglas
1.00E-07
1.00E-08
200 350 500 650 800 950 1100 1250 1400 1550 1700
T [K]
Figure 2: DIPPRs Ps compared to experimental Ps
Given figure 2, as well as the results from table 5 and table A1, it can be concluded
that eq. 2 can provide reliable estimations for the Ps of Hg for the whole temperature
range examined. Thus hereafter the term experimental Ps will be referring to the
values of eq.2 for the given temperature range.
Table 5: Comparison of DIPPRs estimated Ps with experimental Ps for some

characteristic temperatures
Exp. Ps DIPPRs Ps % Deviation

Temperature [K] Pressure [kPa] Pressure [kPa] ((Psexp-PsDIPPR)/Psexp)*100
285.22 8.45E-05 8.464E-05 -0.13%
294.11 0.000185 0.0001855 -0.15%
300.25 0.000312 0.0003102 0.45%
309.29 0.000636 0.0006368 -0.15%
321.15 0.001539 0.0015366 0.16%
417.095 0.293 0.293 -0.07%
447.681 0.964 0.965 -0.10%
462.634 1.627 1.629 -0.13%
497.53 4.882 4.879 0.07%
513.69 7.708 7.699 0.11%
546.934 18.080 18.045 0.19%
587.994 45.215 45.103 0.25%
624.85 92.662 92.429 0.25%
666.82 190.120 189.604 0.27%
23
Continuation of table 5
683.43 246.290 245.784 0.21%
702.724 327.808 327.089 0.22%
749.788 617.883 616.913 0.16%
796.86 1079.800 1078.732 0.10%
814.46 1305.700 1307.372 -0.13%
854.54 1967.280 1966.343 0.05%
882.13 2553.520 2548.829 0.18%
1051.44 9006.415 9338.783 -3.69%
1510.37 78204.030 79191.073 -1.26%
1735.51 157499.800 159904.817 -1.53%
3.2.3 Development of database for binary Vapor-Liquid-Equilibrium (VLE)

and Liquid-Liquid-Equilibrium (LLE) data of Hg with HC
As already mentioned in chapter 1 at the scope of this master thesis, solubility data
for binary mixtures of Hg with various hydrocarbons (HCs) -as solvents- are
necessary for the calculation of binary interaction parameters -kij- for all the models
that are being developed.
Unfortunately only a few papers with such data are available in the literature [4, 5, 7,
8+ and other than that the only available source is the IUPACs book concerning the
solubility of Hg in liquids and compressed gases [6]. Furthermore only [7] and [8]
contained data concerning the solubility of Hg in light HC like methane and ethane,
as well as solubility of Hg in other gases like CO2 and N2. The last four components
mentioned are of great importance to the natural gas industry and this is the reason
why they are specifically pointed out.
The available solubility data of Hg with HC have to do with the mole fraction of Hg in
each binary system. Table 6 displays some information regarding all the binary
mixtures found in literature as well as their source. All the data are presented
analytically on tables C1-C7 at the appendix C from each reference respectively. All
the data presented at table 6 concern liquid-liquid equilibrium (LLE) data of Hg with
HCs except for the data concerning Hg in methane and N2 that are vapor phased
data and also Hg in ethane, propane, iso-butane and CO2, which are vapor-liquid-
equilibrium (VLE) data. The data of IUPAC[6] concerning Hg in propane and n.butane
were excluded from the database due to the fact that they were at very high
temperatures, thus making it impossible to compare with the rest available data, and
therefore in a way unsafe to use.
24
The binary systems found at papers [4] and [5] concern Hg in n.C6, cy-C6, toluene,
benzene and n.C8 and toluene respectively. Except for the data concerning the
binary mixtures of Hg in n.C6, cy-C6, toluene and benzene, the other data cannot be
compared to the data from the IUPAC because they are at different temperatures.
Table 6: Basic informations about binary mixtures of Hg with various HCs, water, CO2
and N2
HC T range[K] P [atm] Number of data Type of data

Robert R. n.C6 298.15-336.15 1 3 LLE
Kuntz and
Gilbert J. cy-C6 298.15 1 1 LLE
Mains [4] toluene 298.15 1 1 LLE
benzene 298.15 1 1 LLE
M. M. n.C8 383.15-473.15 1 3 LLE
MIEDANER, A.
A. MIGDISOV,
and A. E. toluene 393.15-473.15 1 4 LLE
WILLIAMS-
JONES [5]
n.C5 278.15-313.15 1 8 LLE
n.C6 273.15-338.15 1 14 LLE
n.C7 273.15-313.15 1 9 LLE
n.C8 273.15-313.15 1 9 LLE
n.C10 273.15-318.15 1 10 LLE
2.2-dm-C4 273.15-308.15 1 6 LLE
cy-C6 288.15-313.15 1 6 LLE
water 273.15-393.15 1 22 LLE
benzene 288.15-308.15 1 6 LLE
o-xylene 273.15-308.15 1 5 LLE
IUPAC [6] toluene 273.15-308.15 1 6 LLE
2.2.4-tm-C5 273.15-308.15 1 7 LLE
m-cy-C6 273.15-308.15 1 6 LLE
cis-1.2-dm-cy-C6 289.15-308.15 1 5 LLE
cis-1.4-dm-cy-C6 288.15-308.15 1 5 LLE
trans-1.2-dm-cy-C6 288.15-308.15 1 5 LLE
trans-1.4-dm-cy-C6 288.15-308.15 1 5 LLE
Confidential i-C4 263.15-278.15 1.09-1.875 4 VLE
CH4 248.15-293.15 27.58-68.95 19 VE
C2H6 273.15-293.15 23.92-37.65 5 VLE
Confidential C3 273.15-293.15 4.76-8.34 5 VLE
CO2 273.15-293.15 34.82-57.30 5 VLE
N2 273.15 6.85-69.29 6 VE
25
The experimental data from references [4, 5, 6] will now be evaluated since they are
all at the same temperatures, pressures, and also in the liquid phase. The rest of the
data will be analytically evaluated later on because of their ranging in pressure. At
tables C1 and C3 at appendix C, it can be seen that the data concerning benzene,
toluene and n.C6 from IUPAC and Robert et al. [4], are close to each other, except for
the mole fraction of Hg in n.C6 at 336.15 K. At that temperature the mole fraction is
significantly bigger than the one of IUPAC for T = 338.15 K. This is not possible
however because as figures later on show as well, the solubility of Hg in the solvent
is increased monotonically as the temperature rises.
As far as cy-C6 is concerned the common experimental point at 298.15 K is

somewhat different. IUPAC state that the mole fraction of Hg in cy-C6 is equal to
1.32E-06 whereas paper [4] states that it equal to 1.2E-06. Both values will be added
to the database however because the mole fraction is too small and a difference of
such magnitude could be well attributed to experimental errors.
The pressure of the binary systems found in IUPACs book is considered to be equal
to 1 atm because there is no evidence to suggest otherwise. Finally in IUPACs book
[6] all the binary systems that have as solvent any HC after benzene -as they are
shown at table 6- are classed as tentative. However since there is no way to decide
their validity, they will also be added to the solubility database as experimental data,
keeping however that fact in mind.
After the solubility data have been collected, they are evaluated. The evaluation
concerns the solubility of Hg in HCs and consists of a comparison between these data
based on several criteria. The observations that are made from those comparisons
are of great interest because they can be considered as a first indication of how Hg is
expected to be distributed throughout the plant. This evaluation does not contain
the solubility of Hg in propane and n.butane because these experimental data are at
different temperatures and pressures compared to the rest of the data therefore
there can be no comparison between them. However some indirect conclusions can
be reached, with every precaution, about them as well.
Figures 3, 4 and 5 show the effect of the carbon number on the solubility of Hg in all
types of HC. That means that the HC can be either n.alkane, cyclo-alkane, branched-
alkane or aromatic HC.
26
7.0E-06
mole fraction of Hg 6.0E-06
5.0E-06
4.0E-06
n.C6
3.0E-06 2,2-dm-C4
cy-c6
2.0E-06 benzene
1.0E-06
0.0E+00
270 280 290 300 310 320 330 340
T[K]
Figure 3: Solubility of Hg in liquid HC with carbon number 6
As far as the HC with 6 carbon atoms are concerned it is clear that Hg dissolves the
least in branched alkanes, more in n. alkanes, even more in aromatic HC and the
most in cyclo-alkanes.
At 273 K the branched alkanes dissolve almost the same amount of Hg as n.C6 does.
As the temperature rises however the solubility of Hg in n.C6 is greater than that of
Hg in all other branched alkanes.
As far as the solubility of Hg in benzene is concerned it is greater than that of Hg in

n.C6. It is however less than the solubility of Hg in cy-C6 except for the highest
temperature, where it is almost the same in both HCs. It is possible that in higher
temperatures Hg prefers -dissolves more, in terms of quantity- more benzene than
cy-C6.
2.50E-06
2.00E-06
mole fraction of Hg
1.50E-06
1.00E-06 n.C7
m-cy-c6
toluene
5.00E-07
0.00E+00
270 275 280 285 290 295 300 305 310 315
T[K]
27
As far as the HC with 7 carbon atoms are concerned it is clear that Hg dissolves the
least in n.C7, even more in m-cy-C6 and the most in toluene.
3.0E-06 n.C8
2,2,4-tm-C5
2.5E-06
cis-1,2-dm-cy-c6
trans-1,2-dm-cy-c6
mole fraction of Hg
2.0E-06
cis-1,4-dm-cy-c6
trans-1,4-dm-cy-c6
1.5E-06
o-xylene
1.0E-06
5.0E-07
0.0E+00
270 275 280 285 290 295 300 305 310 315
T[K]
Finally as far as the HC with 8 carbon atoms are concerned, Hg prefers the least the
2.2.4-tm-C5, more than that it prefers n.C8, more than that the cyclo-alkanes and it
is solubilized the most in the aromatic HC o-xylene.
One observation that can be made is that almost the same order of preference that
applies to the solubility of Hg in HC with carbon number 6 applies to the HC with
carbon number 7 and 8. The only difference can be located at the order of the cyclo-
alkanes and the aromatic HC. The data of aromatic HC and cyclo-alkanes for 7 and 8
atoms of carbon number respectively suggest that Hg is more soluble in aromatic HC
than in cyclo-alkanes.
Figure 6 shows the effect of the carbon number for straight-chained HC on the
solubility of Hg in n.C5-10 except for n.C9 because there are no experimental data
available for it.
Figure 6 shows that at the lower temperatures where solubility data are available,
that all n.alkanes dissolve almost the same amount of Hg except for n.C10 that
solubilizes a bigger amount. As the temperature rises figure 5 shows that the greater
the number of C atoms of the n.alkane is, the more soluble Hg is in it. This
observation is valid for the liquid data of propane as well because the mole fractions
of Hg in it fluctuate from 10-8 to 10-7 magnitude of order as it will be seen at the next
chapter [8].
Therefore a conclusion to which one can arrive from this figure is that the heavier
the n.alkane is, the more amount of Hg it dissolves. This trend may with every
28
precaution be valid for n.butane as well, meaning that n.alkanes with carbon number
less than 5 may indeed solubilize less amount of Hg than what the rest of the
n.alkanes do in the liquid phase. This is an important observation because after
further analyzing it one may suggest that in a process plant the biggest part of the
total Hg in the feed stream is expected to be found in the streams with the heavier
HC and in the liquid phase.
7.0E-06
6.0E-06
mole fraction of Hg
5.0E-06
4.0E-06
n.C5
3.0E-06 n.C6
2.0E-06 n.C7
n.C8
1.0E-06 n.C10
0.0E+00
270 280 290 300 310 320 330 340
T[K]
Figure 6: Solubility of Hg in liquid n. alkanes
The next case examines the effect of the branches that a HC may have on regardless
of their carbon number. Figure 7 shows the solubility of Hg in all branched alkanes
for which there are available experimental data.
As far as 2.2-dm-C4 and 2.2.4-tm-C5 are concerned, the data suggest that at
temperatures lower than 288.15 K, they dissolve the same amount of Hg, and for
higher temperatures, the heavier one can dissolve more Hg. This is in agreement
with the conclusion deducted from the n.alkanes that Hg prefers the heavier HC.
1.2E-06
1.0E-06
mole fraction of Hg
8.0E-07
6.0E-07
2,2-dm-c4
4.0E-07
2,2,4-tm-c5
2.0E-07
0.0E+00
270 275 280 285 290 295 300 305 310 315
T[K]
Figure 7: Solubility of Hg in liquid branched alkanes
29
The next figure presents the effect of the aromacity on the solubility of Hg in HC
regardless of the number of C atoms of the HC. Figure 8 shows the solubility of Hg in
all aromatic HC for which there are available experimental data. It is reminded that
toluenes and o-xylenes data are classed as tentative.
3.0E-06
2.5E-06
mole fraction of Hg
2.0E-06
1.5E-06
1.0E-06 benzene
toluene
5.0E-07 o-xylene
0.0E+00
270 275 280 285 290 295 300 305 310
T[K]
Figure 8: Solubility of Hg in liquid aromatic HC
At 273.15 K o-xylene and toluene dissolve almost the same amount of Hg. As the
temperature rises however, it is made clear that Hg is more soluble in o-xylene than
in toluene and more soluble in toluene than in benzene. In addition as the
temperature rises figure 8 suggest that the difference among the solubility of Hg in
heavier and lighter aromatic HC is constantly increasing. That means that for
temperatures higher than 308.15 K one may well anticipate that this difference will
keep increasing judging from figure 8.
Therefore it is observed that the aromatic HC with (n+1) C atoms can dissolve more
Hg than the aromatic HC with (n) C atoms. The same thing is valid for n.alkanes as
well. So one may suggest that there is a general rule for the solubility of Hg in HC,
that it is always more soluble in the HC with higher molecular weight when it comes
to the same type of HC.
The next case examines the solubility of Hg in all cyclo-alkanes. Figure 9 shows the
solubility of Hg in every cyclo-alkane for which there are experimental data available.
It has to be reminded beforehand again that all the experimental data, except for
those concerning the cy-C6, are classed as tentative in the IUPACs book [6].
30
2.5E-06 cy-c6
m-cy-c6
2.0E-06 cis-1,2-dm-cy-c6
trans-1,2-dm-cy-c6
mole fraction of Hg
cis-1,4-dm-cy-c6
1.5E-06
trans-1,4-dm-cy-c6
1.0E-06
5.0E-07
0.0E+00
285 290 295 300 305 310
T[K]
Figure 9: Solubility of Hg in liquid cyclo-alkanes
At 288.15 K cy-C6 is the component that dissolves the biggest amount of Hg, cis-1.4-
dm-cy-C6 is the next best solvent and all the other solvents dissolve almost the same
amount of Hg.
Figure 9 also suggest that m-cy-C6 is the solvent that solubilizes the least amount of
Hg. That doesnt apply for the temperature of 3 8.15 K where it can dissolve almost
the same amount of Hg as all the other solvents except for cis-1.2-dm-cy-C6 and cis-
1.4-dm-cy-C6.
Cy-C6 is the best solvent at 288.15 K and among the bests up to 298.15 K. Above that
temperature Hg prefers the cis-1.2-dm-cy-C6, and cis-1.4-dm-cy-C6. Trans-1.2 and
1.4-dm-cy-C6 present almost the same solubility ability regardless of the
temperature. Although these components have the same number of C atoms as their
respective cis structures, Hg is more soluble in the cis ones. This suggests that the
trans structure is prohibiting Hg from dissolving in these solvents as much as it does
in the cis structure.
As far as the cis-1.2 and 1.4-dm-cy-C6 are concerned, they appear to be among the
best solvents along with cy-C6. Above 300 K they dissolve a bigger amount of Hg
than cy-C6 does. An interesting fact is that at 303.15 K cis-1.2-dm-cy-C6 is better
solvent than the 1.4 respective one, but at 308.15 K the experimental data suggest
the exact opposite. This fact can probably be attributed to the position of their
second methyl-group.
A further analysis of these data can be found at the appendix C at figures C1-10
where 4 more cases are being investigated. After this analysis there are four general
conclusions can be summed up as shown below:
31
1) Hg is more soluble in heavier HC compared to the lighter ones of the same
type, meaning n.alkanes, cyclo-alkanes, aromatic HC, and branched alkanes
2) The branched alkanes dissolve the least amount of Hg
3) For all HC the order of their ability to dissolve Hg is as follows: branched

alkanes < normal alkanes < cyclo-alkanes < aromatic HC. For the HCs with 6 C
atoms there is a difference at the order of the aromatic HC and the cyclo-
alkanes
4) At some cases the aromatic HC can dissolve more Hg than n.alkanes even if
the n.alkanes have more C atoms than the aromatic HC. The same conclusion
applies for the cyclo-alkanes as well.
Generally these conclusions present a good indication of how Hg is expected to

distribute in a real process. One last comment about these data is that Hg is not very
soluble in HCs and the deviations between its solubility in various types of HCs are
not very big. Therefore at a real process one may anticipate that the bigger part of
Hg will end up in the streams with the heavier HC and with the bigger
concentration in aromatic HC and cyclo-alkanes, however this does not necessarily
mean that almost no Hg will end up in other streams as well. This is said because of
the results from Carnells P. et al survey *9+, which indicated that Hg distributes
throughout entire gas processing plants. Thus it is not possible to provide a safe
estimation beforehand about its exact distribution throughout a process.
3.2.4 Experimental errors of the data assembled
It is of great importance for one to know the error margins of the experimental data
at hand. Unfortunately all data available for the solubility of Hg in other components
of interest to the natural gas industry do not include their experimental errors. It is
obvious that for the data assembled from IUPAC [6], one cannot even estimate that
error since the solubility of Hg is in the liquid phase. As far as the data assembled
from [8] goes however, it is possible to have an estimation of it and thus get a feeling
about the accuracy for the models that will be developed.
[8] has solubility data of Hg both in the liquid and in the vapor phase for some
components as seen in section 3.2.3, at table 6. The vapor phased data can be used
in order to get a feeling of the experimental error for this reports data.
It is known [10+ that for low pressures, which are within the margin of 1 to 2 bar,
the following formulas are valid:
32
1) ) (eq. 3) if T, x constant
2) (eq. 4) if T, x constant
3) By combining eq. 3, 4 it occurs: ) (eq. 5)
4) (eq.6)
5) From eq. 6 it occurs : ) ) (eq. 7)
6) By putting eq. 7 to eq. 5 and if it is: ) (eq. 8)
7) ) ) (eq. 9)
8) Assumptions:
i) Vi is the liquid molar volume, independent of p and equal to V is
ii) Vi is negligible, therefore = )
iii) The gas phase is considered to be an ideal one, thus:
iv) For ideal gasses it is considered that: and also
v) Therefore )
vi) The liquid phase is considered to be an ideal solution and also pure Hg.
Thus it is deducted that and
9) By using the assumptions mentioned and eq. 9 one can conclude that:
(eq. 10)
Where:
a) dG is the derivative of the Gibbs energy
b) R is the global constant of gasses
c) f is the fugacity
33
d) V is the liquid grammomolar volume
e) is the fugacity coefficient
f) is the activity coefficient
g) is the fugacity coefficient of a gas i in a mixture consisted of many

gasses
h) T is the absolute temperature [K]
i) p is the pressure
j) is the fugacity of component i in the vapor phase
k) is the fugacity of component i in the liquid phase
l) is the liquid mole fraction of component i
m) is the vapor mole fraction of component i
n) The superscript s indicates saturated property
By using the DIPPRs equation it is possible to estimate very accurately the ps of Hg at

the temperatures of the experimental data that fit within the low pressure margin.
Tables 7 and 8 present the yideal of Hg estimated as , the experimental yi of Hg and
their deviations for the binary systems of Hg with propane and N2 respectively.
Table 7: Deviations between the experimental yi and the yideal for Hg in propane
T [K] Dy%
273.15 16.54
278.15 18.10
283.15 16.39
288.15 15.75
293.15 16.09
Overall dev. 16.57
Table 8: Deviations between the experimental yi and the yideal for Hg in N2
T [K] Dy%
273.15 3.96
273.15 3.59
273.15 8.21
Overall dev. 5.26
34
s
Hg
Table 9: Deviations of the models from i for the binary mixture of Hg with
C3
T [K] PR-MC dev. % SRK-Twu dev. % UMR-PRMC dev.%

273.15 4.54 4.36 2.35
278.15 4.98 3.24 2.54
283.15 5.23 3.61 2.53
288.15 6.19 4.67 3.15
293.15 6.63 5.21 3.22
Overall dev. for each model % 5.51 4.22 2.76

Total overall dev. for all models % 4.16
s
Hg
Table 10: Deviations of the models from i for the binary mixture of Hg
with N2
T [K] PR-MC dev. % SRK-Twu dev. % UMR-PRMC dev.%

273.15 2.70 1.44 2.64
273.15 2.76 1.50 2.69
273.15 7.77 6.77 7.64
Overall dev. for each model % 4.41 3.24 4.32

Total overall dev. for all models % 3.99
The analytical results of tables 9 and 10 will be presented at their corresponding

subchapters explicitly. The deviations of these tables imply that as the pressure
rises, so does the deviation between the models and the yideal. This is expected since
the raise of pressure means also the movement away from the area where the vapor
mole fractions of Hg behave as an ideal gas.
It can be seen that the overall average experimental error is 16.57% for Hg in
propane and 5.26% for the system of Hg with N2 compared to the yideal. These
percentages set an ideal experimental error area. However no system in nature
can be considered as totally ideal, therefore there will always be a deviation from
that ideal consideration. The three models that will be developed in the following
chapter can provide an estimation of this deviation from the ideal state without of
course being totally accurate. For that reason if the results from tables 7-10 are
combined, it can be concluded that the approximate deviation of the experimental
error of these data in the vapor phase compared to the more realistic non-ideal
systems, is higher than 10% -around 13%- regarding the mole fractions of Hg in
propane, and around 1.3% regarding the mole fractions of Hg in N2. These
experimental errors are extremely good, given the fact that the Hg solubility is
35
measured in ppb. Furthermore this area set by these percentages will be called for
the rest of the master thesis as non-ideal experimental error area.
Given the experimental error for Hg in propane, this can be generalized to include all
the vapor phased data in the database concerning Hg in HCs from [8], for which the
respective parameters of each model will be estimated. This is important to know
because that way one has an indication about whether the results of the models
need to be further improved in order for them to be more accurate or not,
depending on whether their results are within both experimental error areas and
especially the non-ideal one.
References:
[1] T.E. Daubert et al.(1994), Physical and Thermodynamic Properties of
Pure Chemicals, Data Compilation, Hemisphere, New York
[2] Marcia L. Huber et al.(2 6), The vapor pressure of mercury, NIST
[3] Kozhevnikov et al.(1996), S. Phase transitions and critical phenomena in mercury

fluid probed by sound, Fluid Phase Equilib., 125: 149-157
[4] Robert R. Kuntz et al.(1963), The Solubility of Mercury in Hydrocarbons, The

Journal of Physical Chemistry
[5] M. M. Miedaner et al. (2005), Solubility of metallic mercury in octane, dodecane

and toluene at temperatures between 100 oC and 200 oC
[6] IUPAC, Solubility Data Series, Mercury in liquids, compressed gases molten salts
and other elements, Volume 29, p. 101-160
[7] Confidential reference
[9] Carnell P. et al (2004),''MERCURY DISTRIBUTION ON GAS PROCESSING PLANTS '',

83rd annual GPA convention proceedings, New Orleans
[10] Michael M.Abbott, Hendrick C. Van Ness (1983), '''', Translated

by Konstantinos . Pattas, Nikolaos A. Kiriakis, Schaum outline series, Mc Graw-Hill,
New York, p. 238-239, 262-263
36
Chapter 4: Thermodynamic modeling
4.1 SRK and PR EOS

In this thesis parameters for three models will be developed. These models are
based on the SRK [1] and the PR EOS [2]. The reason why these two EOS are chosen
is because they are widely used for process simulations at the industrial sector.
There are of course other models, more advanced as well, that differ from these
EOS. One example is the UMR-MCPRU model developed by Voutsas et.al[3] that uses
more advanced mixing rules than the Van der Waals one fluid [4] used by classical
EOS SRK and PR. One other example is the PC-SAFT equation [5] that is not a cubic
equation of state. Such models could very well be examined as a future work in
order to compare their results with the ones from the models that will be developed
in this master thesis. It is mentioned that the b parameter is estimated for each
model as described at [4] and is not analyzed in this master thesis since no adjusting
for it has occurred.
The SRK EOS is presented as equation 11:
RT a
P (eq. 11), where
Vm b Vm (Vm b)
a xi x j (ai a j ) 0.5 (1 kij ) (eq. 12) and

i j
R 2Tci2
ai i 0.42747 , i alpha (eq. 13)
Pci
The PR EOS is presented as equation 14:
RT a
P (eq. 14), where
Vm b Vm (Vm b) b(Vm b)
a xi x j (ai a j ) 0.5 (1 kij ) (eq. 15) and

i j
R 2Tci2
ai i 0.45724 , i alpha (eq. 16)
Pci
37
4.2 Evaluation of SRK and PR EOS with two different sets of Tc, Pc,
The Neqsim tool [6] was used for the calculation of the Ps of Hg with SRK and PR EOS.
It has to be mentioned that Neqsim was able to provide P s from 238.15 K up to
almost 1728 K (depending on the set of Tc, Pc, that was examined and on the EOS
that was used). This temperature is close to the critical one and perhaps this is the
reason why it was not able to provide Ps for any higher temperatures.
Figure 10 illustrates the deviations of the Ps acquired from Neqsim tool using the SRK
with both sets of Tc, Pc and for temperatures from 238.15 K up to 64 .15 K. It can
be seen that the deviations are quite big and that SRK under-predicts the Ps of Hg at
those temperatures.
NISTs set of Tc, Pc, predicts better the Ps than DIPPRs set does for these
temperatures. The maximum deviation for DIPPRs set was observed at the lowest
temperature and was 87.5%. The maximum deviation for NISTs set was also
observed at the lowest temperature and was 66.7%.
For temperatures higher than 640 K, as figure 10 implies as well, the deviations
become significantly smaller. Table A2 at the appendix A presents the deviations of
the Ps calculated by the Neqsim tool using the SRK EOS and eq. 2 analytically.
1.00E+01
1.00E+00
1.00E-01
1.00E-02
1.00E-03
P [bar]
1.00E-04
1.00E-05
1.00E-06 Dippr
1.00E-07 Neqsim-SRK (DIPPR's set)
1.00E-08 Neqsim-SRK (NIST's set)
1.00E-09
1.00E-10
230 280 330 380 430 480 530 580 630
T [K]
Figure 10: Comparison of results from DIPPRs eq. and Neqsim tool using the SRK
EOS for both set of Tc, Pc and
Table 11: Absolute average deviation of each case examined
SRK (DIPPRs set) SRK (NISTs set)

Absolute average deviation % 18.53 12.02
38
Table 11 indicates that the set of Tc, Pc, from NISTs report improves the prediction
of the Ps compared to the set proposed by DIPPR for the whole temperature range
examined. However the absolute average deviation is considerably smaller -about
6.5%- than the deviations at the temperatures shown at figure 10, because as the
temperature rises the deviations are getting significantly smaller. Thus the absolute
average deviation is significantly reduced.
Table 12 presents some indicative Ps estimated from the SRK EOS and their
respective deviations from the experimental Ps of eq.17. It can be seen that as the
temperature rises the deviations of the estimated and the experimental P s are
declining and above 1500 K are rising again. Still the deviations are not that high -less
than 2%-. Furthermore the SRK EOS using DIPPRs set is providing less accurate
estimations of the Ps than the one using NISTs set for the whole temperature range.
)
(eq. 17)
Set 1 = DIPPRs set of Tc, Pc,
Set 2 = NISTs set of Tc, Pc,
Table 12: Indicative Ps estimated from Neqsim tool using SRK EOS and their
respective deviations from the experimental Ps estimated by eq.17
T [K] DIPPR Ps Neqsim- Ps % Neqsim- Ps %

[bar] SRK (set 1) SRK(set 2)
238.15 4.90E-09 6.13E-10 87.49 1.63E-09 66.65
368.14 2.85E-04 1.01E-04 64.65 1.64E-04 42.53
478.14 0.027 1.44E-02 46.71 0.0194 28.27
588.14 0.452 3.06E-01 32.43 0.369 18.37
678.14 2.266 1.734 23.47 1.978 12.71
798.14 10.942 9.334 14.70 10.113 7.58
878.14 24.574 22.024 10.38 23.291 5.22
938.14 41.200 38.007 7.75 39.617 3.84
1058.14 97.501 93.908 3.69 95.712 1.83
1188.14 205.531 204.177 0.66 204.471 0.52
1298.14 347.101 350.607 -1.01 347.152 -0.01
1308.14 362.641 366.719 -1.12 362.776 -0.04
1428.14 588.989 601.037 -2.05 588.825 0.03
1538.14 870.370 889.198 -2.16 864.756 0.64
1648.14 1236.647 1256.377 -1.60 1214.210 1.81
Figure 11 shows the deviations of the Ps acquired from Neqsim tool while using the
PR EOS this time for both sets of Tc, Pc and at temperatures from 238.15 K up to
640.15 K. It can be seen that the deviations are bigger than those for the SRK EOS.
39
Also PR seems to under-predict the Ps of elemental Hg at those temperatures while
using the DIPPRs set, whereas it over-predicts it while using the NISTs set.
The use of DIPPRs set of Tc, Pc, seems to lead to better results Ps than NISTs set
does at low temperatures. The maximum absolute deviation noted was 89.5% for
NISTs set, whereas DIPPRs set had a maximum deviation of 32.7%.
For higher temperatures, as figure 11 implies as well, the deviations become

significantly smaller. Table A2 at the appendix A presents the deviations of the Ps
calculated by Neqsim tool using the PR EOS and analytically.
1.00E+01
1.00E+00
1.00E-01
1.00E-02
1.00E-03
P [bar]
1.00E-04
Dippr
1.00E-05
Neqsim-PR (DIPPR's set)
1.00E-06
Neqsim-PR (NIST's set)
1.00E-07
1.00E-08
1.00E-09
230 280 330 380 430 480 530 580 630
T [K]
Figure 11: Comparison of results from DIPPRs eq. and Neqsim tool using the PR EOS
for both set of Tc, Pc and
Table 13 presents some indicative Ps estimated from the PR EOS and their respective
deviations from the experimental Ps of eq. 2. Like the SRK EOS and the PR EOS show,
as the temperature rises, the deviations of the estimated and the experimental P s
are declining and above 1500 K are rising again. In addition the PR EOS using NISTs
set is providing less accurate estimations of the Ps than the one using DIPPRs set for
the whole temperature range.
)
(eq. 18)
40
Table 13: Indicative Ps estimated from Neqsim tool using PR EOS and their respective
deviations from the experimental Ps of eq.18
T [K] DIPPR Ps Neqsim- Ps % Neqsim- Ps %

[bar] PR(set 1) PR(set 2)
238.15 4.90E-09 3.30E-09 32.70 9.29E-09 -89.50
368.14 2.85E-04 2.298E-04 19.47 0.00039 -35.84
478.14 0.027 0.0238 11.95 0.033 -21.89
588.14 0.452 0.4216 6.81 0.521 -15.07
678.14 2.266 2.177 3.95 2.528 -11.58
798.14 10.942 10.785 1.43 11.857 -8.36
878.14 24.574 24.504 0.28 26.242 -6.79
938.14 41.200 41.364 -0.40 43.602 -5.83
1058.14 97.501 98.913 -1.45 101.721 -4.33
1188.14 205.531 210.101 -2.22 211.860 -3.08
1298.14 347.101 356.113 -2.60 354.502 -2.13
1308.14 362.641 372.128 -2.62 370.0579 -2.05
1428.14 588.989 604.594 -2.65 594.509 -0.94
1538.14 870.370 890.477 -2.31 868.048 0.27
1648.14 1236.647 1256.077 -1.57 1215.157 1.74
Table 14: Absolute average deviation of each case examined for the whole
temperature range
PR (DIPPRs set) PR (NISTs set)

To sum up, table 14 indicates that the set of Tc, Pc, from DIPPR improves the
prediction of the Ps compared to the set proposed by NIST for the whole
temperature range examined when used by the PR EOS. This is the exact opposite
conclusion compared to the one from the SRK EOS. Furthermore the PR EOS seems
to be able to predict more accurately the Ps of Hg than the SRK EOS does. Thus there
is no clear advantage for NISTs set over the one from DIPPR.
DIPPRs set in addition has the advantage that is used in official software like Hysys,
PROII and PVT-Sim. Therefore it has been decided that this set will be used for the
calculations and the fitting of all the parameters necessary for the thermodynamic
models that will be developed for this master thesis.
41
4.3 Calculation of the Mathias-Copeman parameters for Hg for the PR
EOS
The parameters of Mathias-Copeman (MC) [7] will be introduced in the PR EOS.
These parameters are fitted to experimental Ps data of pure components and
provide a better description of the Ps than the correlation of Soave for the alpha
parameter. The MC parameters are associated with the calculation of the alpha
parameter for the EOS. The equation with these parameters is equation 19 and the
alpha parameter of Soave is equation 20. The objective function that is used for the
calculation of the MC parameters is equation 21.
) ( ) ( ) ( )3 (eq.19)
) ( ) (eq.20)
)
) (eq.21)
The results from eq. 2 at table A1 were used as experimental data for the fitting of
the MC parameters because they cover Hgs temperature range from the freezing
point up to the critical point. Table 15 presents the parameters calculated for
DIPPRs set of Tc, Pc and as well as the temperature range of the fitting in terms of
Tr.
Table 15: The estimated MC parameters
MC Parameters Tr *0.137, 0.875+

MC1 0.1491
MC2 -0.1652
MC3 0.1447
As table 15 illustrates, the temperature fitting range is from Tr = 0.137 (that means T
= 238.15 K) up to Tr = 0.875 (which means T = 1518.15 K). Given the fact that the
freezing point of Hg is at 234.15 K, the lower limit of the Tr range can be
characterized as satisfying.
After the MC parameters were determined, the PR EOS was employed while
calculating the new alpha parameter as it stands at eq.19, for the prediction of the Ps
anew (hereafter this EOS shall be referred as PR-MC). Figure 12 presents the
deviation between the results from PR and PR-MC EOS with the results from eq. 2
respectively, for Tr = 0.137 up to Tr = 0.97.
42
35%
30%
25%
20%
deviation
15% PR
PR-MC
10%
5%
0%
-5%
0.12 0.22 0.32 0.42 0.52 0.62 0.72 0.82 0.92
Tr
Figure 12: Deviation of Ps estimated from eq.2 and Neqsim tool using the PR and PR-
MC EOS
As figure 12 indicates the results from PR-MC are much closer to those from eq.2
compared to those from PR EOS. This fact can be attributed to the effect of the MC
parameters, something that also reveals the importance of these parameters to the
model under development.
It can also be observed that for Tr = 0.88 and above that the deviation becomes
significantly larger compared to what it was below that reference temperature. This
could be explained up to a point by the fact that the temperature range that was
used for the fitting was up to Tr = 0.875. However these deviations are still less than
1.2% and furthermore they will have no effect on the testing of the models because
the processes in the natural gas industry take place at much lower temperatures.
At the appendix B at table B1 analytical Ps for all temperatures and their respective
deviations can be found for PR-MC.
Table 16: Absolute average deviation of PR and PR-MC EOS from the experimental Ps
for temperatures from 238.15 K up to 608.15 K
PR PR-MC
As an overall comment as table 16 shows, the Ps calculated from PR-MC present

small deviation to the Ps calculated from eq.2 from 238.15 K up to 608.15 K. This is
43
extremely important because this temperature range also includes the temperatures
where the processes of natural gas take place.
4.4 Calculation of Twu parameters for Hg for the SRK EOS

The L, M, N parameters introduced by Twu [8] also apply to the calculation of the
alpha parameter for the SRK-Twu EOS. Like the MC parameters, these are also fitted
to experimental Ps data of pure components and assist in a better description of the
Ps for pure Hg. The equation with these parameters is eq.22. The objective function
for the calculation of these parameters is again equation 21.
-
) ) (eq.22)
The temperature range at which the author did the fitting of the L, M, N parameters
was that of 238.15 K up to 1508.15 K. As mentioned for the MC parameters as well,
given the fact that the freezing point of Hg is at 234.15 K, the lower limit of the Tr
range can be characterized as satisfying. Table 17 presents the parameters
estimated for the SRK EOS.
Table 17: Estimated L, M, N parameters
Twu-Coon Parameters Tr *0.137, 0.875+

L 0.09245
M 0.9784
N 2.244
After the determination of the L, M, N parameters, SRK EOS was used while taking
into consideration the parameters for the calculation of the new alpha, for the
prediction of the Ps of pure Hg. The SRK EOS when combined with these parameters
will be hereafter referred as SRK-Twu EOS. Analytical results and deviations of SRK-
Twu with the experimental Ps can be found at the appendix B at table B2.
Figure 13 shows the deviation between the experimental Ps and the results from SRK
and SRK-Twu EOS respectively for Tr = 0.137 up to Tr = 0.978. As figure 13 indicates
the Ps acquired from the SRK-Twu EOS are much closer to the experimental data
than the ones acquired from the SRK EOS. This fact can be attributed to the effect of
the estimated parameters.
44
87%
77%
67%
57%
deviation
47%
SRK
37% SRK-Twu
27%
17%
7%
-3%
0.12 0.22 0.32 0.42 0.52 0.62 0.72 0.82 0.92
Tr
Figure 13: Deviation of Ps estimated from eq.2 and Neqsim tool using the SRK and
SRK-Twu EOS
One more comment that can be made concerning those parameters and the MC
parameters respectively is the fact that PR-MC EOS at the temperature range of
238.15 K up to 348.15 K seems to give a better prediction for the P s of Hg compared
to SRK-Twu EOS. The maximum deviation of the PR-MC at that temperature range is
0.86% whereas the respective deviation of the SRK-Twu using the authors L, M, N
parameters is 3.82% as table 18 presents.
This fact is mentioned because this is the temperature range of major interest to the
gas industry. Therefore it is a critical observation for the evaluation of the results
that the models will provide once tested for a real process.
Table 18: Ps calculated using SRK-Twu and PR-MC and their deviations from the Ps of
eq.2 for the temperature range of 238.15 K up to 378.15 K.
DIPPRs eq. PR-MC SRK-Twu

T [K] Ps [bar] s
P [bar] s
P % s
P [bar] Ps %
238.15 4.90E-09 4.86E-09 0.86 4.71E-09 3.82
248.15 1.73E-08 1.72E-08 0.72 1.67E-08 3.11
258.15 5.52E-08 5.49E-08 0.62 5.38E-08 2.53
268.15 1.62E-07 1.61E-07 0.53 1.58E-07 2.05
278.15 4.37E-07 4.35E-07 0.47 4.30E-07 1.65
288.15 1.10E-06 1.10E-06 0.41 1.09E-06 1.32
298.15 2.61E-06 2.60E-06 0.37 2.58E-06 1.05
308.15 5.82E-06 5.81E-06 0.33 5.78E-06 0.82
318.15 1.24E-05 1.23E-05 0.30 1.23E-05 0.63
45
328.15 2.51E-05 2.50E-05 0.28 2.49E-05 0.48
338.15 4.86E-05 4.85E-05 0.25 4.85E-05 0.35
348.15 9.08E-05 9.06E-05 0.23 9.06E-05 0.25
358.15 1.64E-04 1.63E-04 0.21 1.63E-04 0.16
368.15 2.85E-04 2.85E-04 0.19 2.85E-04 0.09
378.15 4.83E-04 4.82E-04 0.17 4.83E-04 0.03
4.5 Estimation of the binary interaction parameters (kij) for binary

systems of Hg with hydrocarbons
4.5.1 Calculation of the kij parameters
The kij parameter [9] is an important parameter that needs to be estimated for the
thermodynamic models under development. The kij parameter is an empirical factor
introduced to increase the accuracy of the cubic equations of states through a better
representation of different pair interactions. Its implementation will help the model
give a better estimation for the phase equilibrium of systems containing Hg with HCs
and other components as well.
This parameter is calculated by fitting experimental solubility data of binary mixtures

consisting of Hg and other hydrocarbons or other components. All the experimental
data of the binary mixtures used for the fitting of the k ij parameters that will be
presented are in the liquid phase, except for the parameters concerning Hg in
methane and N2, which were fitted to vapor phased data.
The Neqsim-tool was used for the fitting. The objective function that the program
has to minimize can be seen at equation 23.
)
) (eq.23),
Where x represents the mole fraction of Hg.
As mentioned before, this parameter is calculated by fitting solubility data of binary

mixtures of Hg with other hydrocarbons. The corresponding database has been
presented in Chapter 3. For this thesis kij parameters for two different models are
calculated.
The first one is the PR EOS which takes into consideration the Mathias Copeman
parameters for the calculation of the alpha parameter and the second one is the
SRK-Twu EOS which uses the L, M, N parameters of all components for the
calculation of the same parameter. The L, M, N parameters for all components were
46
taken from the official program PROII and the MC parameters from Neqsims
database. At the appendix D at tables D1 and D2 one can find the L, M, N and MC1,
MC2, MC3 parameters of each component respectively.
Table 19 presents the kij parameters that were calculated for each binary system and
each EOS respectively.
Table 19: kij parameters calculated for each binary system
System kij
SRK-Twu PR-MC
Hg-CH4
Hg-C2H6
Hg-C3
Hg-n.C5
Hg-n.C6
Hg-n.C7
Hg-n.C8
Hg-n.C10
Hg-2.2-dm-C4
Hg-2.2.4-tm-C5
Hg-cy-C6 Confidential data
Hg-toluene
Hg-m-cy-C6
Hg-benzene
Hg-o-xylene
Hg-cis-1.2-dm-cy-C6
Hg-cis-1.4-dm-cy-C6
Hg-trans-1.4-dm-cy-C6
Hg-trans-1.2-dm-cy-C6
Hg-CO2
Hg-N2
Hg-water - -
It is interesting to note that as table 19 shows, the kij parameters for both models
when it comes to the same binary mixture are very similar to one another. In
addition to that although for n.alkanes with CN equal to or greater than 3 the k ij
parameters are constantly declining, the respective parameters of methane and
ethane do not exactly fit in this pattern. Therefore given the fact that their
respective parameters have been safely derived from experimental data, these
components will not be taken into consideration later on when generalized
correlations for these parameters will be developed.
47
It is important to mention that for the binary mixtures of Hg in ethane, propane and
CO2, there are experimental data available both in the liquid phase, as well as in the
vapor phase. For the fitting of the parameters required for each model, the data of
the liquid phase have been used as already mentioned. Afterwards these parameters
were used with each model respectively in order to predict the experimental data in
the vapor phase.
The Neqsim-tool was used for the fitting of the kij parameters, as well as the
prediction of the mole fractions of Hg in the vapor phase for these mixtures as well.
For the prediction of the mole fractions of Hg in the vapor phase the Neqsim-tool
was employed once again, doing a Bubble Point Pressure (B.P.P) calculation this
time. The data given to the tool were the liquid mole fraction estimated from the
fitting of the kij parameters, the kij parameters, and the pressures. This happened in
order for the results of the B.P.P calculations to be in agreement, in terms of
consistency, with the fitted data. This method was adopted for the B.P.P calculations
of all models.
Finally a component that should be treated with extra caution is water. As table 16
indicates the binary mixture of Hg with water [13] has no constant kij parameters.
For the kij parameters it was observed and concluded that they are strongly
dependent on the temperature. Therefore temperature dependent parameters were
developed. In order to do that the parameters were first fitted to the experimental
data at 293.15 K and 363.15 K respectively and exclusively. This temperature range
was decided upon the fact that at that particular area, the experimental data
presented a somewhat linear behavior. After that the two parameters resulted from
the fitting were combined into a linear equation.
It is stated beforehand however that due to the hydrogen bond of H2O, the last two
models are not the best option for conducting simulations of processes, since they
fail to take these bonds into account.
Table 20 presents the equations of the kij parameters for the PR-MC and the SRK-
Twu models, as well as the overall absolute deviations of their predictions from the
experimental data. The analytical results can be found at appendix E at tables E21
and E42.
Table 20: Equations of temperature dependent kij parameters for Hg and H2O
Model kij = a T[K] + b Abs. Deviation (%)

SRK-Twu kij = 0.00254 T 0.11382 12.86
PR-MC kij = 0.00246 T 0.09719 13.06
48
1.00E-07
9.00E-08
8.00E-08
mole fractions of Hg
7.00E-08
6.00E-08
5.00E-08 exp
4.00E-08 SRK-Twu
3.00E-08 PR-MC
2.00E-08
1.00E-08
0.00E+00
250 300 350 400
T[K]
Figure 14: Experimental and predicted mole fractions of Hg in H2O with PR-MC and
SRK-Twu models
As figure 14 displays the PR-MC and SRK-Twu models fail to provide accurate
predictions of the mole fraction of Hg at temperatures higher than 373.15 K. It is
important however that up to that temperature the maximum deviation of both
models is around 28% at 273.15 K and less than 20% for all other temperatures.
Furthermore the results of both models are extremely close to one another as figure
14 illustrates and table 20 suggests as well.
Table 21 presents the absolute average deviation of the mole fraction of Hg for all
systems and each EOS examined. It is noted that water was excluded from this
calculation since it is considered to be a special component as mentioned. The
kij=con. means that the kij parameter is constant with temperature and different
than zero. The absolute average deviation is estimated as eq. 24 shows.
In principle one may say that once the kij is calculated for each binary mixture
respectively and then taken into account for the calculations of the mole fraction of
Hg, as table 21 indicates as well, the accuracy of the results is improving significantly
for most of the mixtures.
| |)
(eq.24),
where x stands for the mole fraction of Hg
49
Table 21: The overall absolute average deviation (%) of the mole fraction of Hg in the
liquid phase for each EOS
Absolute average deviation of mole fraction of Hg for each EOS(%)

SRK-Twu PR-MC
kij = 0 32.83 29.18
kij = con. 4.55 4.57
Tables 22 and 23 present the absolute overall deviation of each model -with the kij
parameters- regarding the mole fractions of Hg from the experimental data of the
liquid and the vapor phase respectively. Table 23 also presents the total average
deviation of the models compared to the experimental data for the vapor phase.
Table 22: Absolute overall deviation (%) of the mole fractions of Hg in each binary
mixture for each model concerning the liquid phase
Model SRK-Twu PR-MC
System Absolute Overall Deviation %

Hg-C2H6 5.82 5.66
Hg-C3 0.40 0.42
Hg-n.C5 3.33 3.62
Hg-n.C6 2.29 1.71
Hg-n.C7 2.04 2.21
Hg-n.C8 3.71 3.07
Hg-n.C10 12.60 12.27
Hg-2.2-dm-C4 9.19 8.71
Hg-2.2.4-tm-C5 5.88 5.86
Hg-cy-C6 7.39 7.15
Hg-toluene 4.01 3.83
Hg-m-cy-C6 3.31 2.94
Hg-benzene 1.99 2.11
Hg-o-xylene 2.91 3.52
Hg-cis-1.2-dm-cy-C6 4.59 4.47
Hg-cis-1.4-dm-cy-C6 3.44 3.71
Hg-trans-1.4-dm-cy-C6 3.97 4.09
Hg-trans-1.2-dm-cy-C6 4.35 4.05
Hg-CO2 5.24 7.49
50
Table 23: Absolute overall deviation (%) of the mole fractions of Hg in each binary
mixture for each model concerning the vapor phase
Model SRK-Twu PR-MC
System Absolute Overall Deviation %

Hg-CH4 2.35 2.22
Hg-C2H6 16.33 15.37
Hg-C3 13.04 11.96
Hg-CO2 2.28 12.36
Hg-N2 1.34 0.90
Total deviation % 7.07 8.56
The overall deviations illustrated at table 22 reveal that both the SRK-Twu and the
PR-MC models are able to predict accurately enough Hgs mole fractions in the liquid
phase. The biggest deviations from the experimental data are presented in the
binary mixtures of Hg with n.C10 and 2.2-dm-C4, namely around 12% and 9% for both
models.
It is also notable that the overall deviations of both models for each binary mixture
are very similar, as verified from table 19 as well since their kij parameters are also
very close to each other. The only binary mixture that presents a difference of about
2% regarding these deviations is that of Hg in CO2. Given however the fact that these
deviations are less than 10% for all binary mixtures except for the one of Hg with
n.C10 which is around 12.4%, which is also a very small percentage given the
magnitude of order of the experimental data, it can be concluded that the models
are reliable.
The overall deviations illustrated at table 23 reveal that the SRK-Twu and the PR-MC
models are able to predict Hgs mole fractions in the vapor phase satisfyingly enough
as well. It is very important that both models are able to predict extremely
accurately the vapor phase of Hg in methane, since this is the main component of
interest when it comes to natural gas processes.
The biggest deviations from the experimental data are presented in the binary
mixtures of Hg with ethane and propane, namely around 15% and 12% respectively.
These deviations are on the verge of the non-ideal experimental error area and
within the ideal experimental error area, as defined at subchapter 3.2.4. However
since they are so close to the non-ideal area and the solubility of Hg is measured in
ppb, therefore there is also always the case of a computation error by the algorithm
due to the extremely small magnitude of order, the results of the models are
acceptable for both cases.
51
The binary mixture that once again stands out is that of Hg in CO 2. This is the first
time that the PR-MC and the SRK-Twu models display such a difference in their
predictions either regarding the liquid or the vapor phase. Tables 24 - 25 and figure
15 illustrate the mole fractions of Hg in the vapor phase, as estimated from the
models.
Table 24: Mole fractions of Hg in CO2 with the PR-MC model in the vapor phase
T[K] Pexp[bar] yexp ycalc y%

273.15 12.27
278.15 10.30
283.15 12.19
288.15 11.90
293.15 Confidential data 15.15
Overall dev% 12.36
Table 25: Mole fractions of Hg in CO2 with the SRK-Twu model in the vapor phase
T[K] Pexp[bar] yexp ycalc y%

273.15 3.34
278.15 3.36
283.15 1.40
288.15 2.07
Overall dev% 2.28
6.00E-08
5.00E-08
mole fraction of Hg
4.00E-08
3.00E-08 PR-MC
SRK-Twu
2.00E-08
exp
1.00E-08
0.00E+00
270 275 280 285 290 295
T[K]
Figure 15: Experimental and estimated mole fractions of Hg in CO2 in the vapor
phase from the SRK-Twu and the PR-MC model
52
As tables 24 and 25 show, the mole fractions of Hg in the vapor phase, as estimated
from PR-MC vary from the experimental data by 10.30% up to 15.15%, whereas the
ones predicted by SRK-Twu vary by 1.21% up to 3.36%. The 15.15% appears only at
the experimental point with the highest temperature and pressure and is very close
to the non-ideal experimental error margin, therefore it can be accepted.
Apparently the SRK-Twu model can describe better the experimental data, none the
less both models predictions have to be considered valid ones.
The analytical results for all binary mixtures in both the liquid and the vapor phase -
where available- and both models can be found at appendices E and G, at tables E1-
42 and G1-9. Figures E1-42 and G1-3 at the same appendices illustrate these results
graphically.
Conclusively both the PR-MC and the SRK-Twu models are able to provide
satisfactory predictions regarding the mole fractions of Hg in both the vapor and the
liquid phase. There are of course some deviations in certain binary mixtures,
however by still being within or very close to the acceptable limits, set by the non-
ideal experimental error margin and within the ideal experimental error area,
concerning the vapor phase, the models estimations for all mixtures are considered
to be credible and can be used for process simulations at a later stage.
4.6 The UMR-PRMC model

The next task of this master thesis is to estimate the Unifac group interaction
parameters for the UMR-PRMC model. This is a predictive model belonging to the
category of the EOS/GE models. It is based on the PR-MC model and instead of the
classical mixing rules with the kij parameters, it combines the PR-MC with a Unifac-
type GE model via the universal mixing rules developed by Voutsas et al [3].
Equations 25 to 35 describe the corresponding mixing rules.
(eq.25)
(eq.26)
( )2 (eq.27)
) (eq.28)
) (eq.29)
53
) (eq.30)

For component i

(eq.31)

(eq.32)
For Unifac group m

(eq.33)
)
(eq.34)
) )
(eq.35)
Where:
a) Anm, Bnm, Cnm are the Unifac interaction parameters between groups n
and m
b) b is the co-volume parameter of an EOS
c) is the molar volume
d) ri is the relative Van der Waals volume of compound i
e) qi is the relative Van der Waals surface area of compound i
f) Qk is the relative Van der Waals surface area of sub-group k
g) x is the mole fraction
h) Xm is the group mole fraction of group m
i) is the Unifac parameter
j) is the Staverman-Guggenheim term of the combinatorial part
of the excess Gibbs energy
k) is the Staverman-Guggenheim term of the residual part of the
excess Gibbs energy
l) is the global constant for gasses
m) is the residual activity coefficient of group k in a solution
n) is the surface area fraction of component i
o) is the segment fraction of component i
p) T is the absolute temperature [K]
q) The parameter A (eq.25) is equal to -0.53
The advantage of this model compared to the previous ones is the fact that it
considers that all components are comprised of the unifac groups and describes
them as a combination of these groups. This allows the model to use the
54
corresponding unifac parameters for components where no experimental solubility
data are available and provide a better approximation of the composition of any
mixture that has two different phases in equilibrium.
A good example to underline this importance is n.dodecane. For the binary mixture
of this component with Hg there are no experimental data regarding the solubility of
Hg in it. The PR-MC and SRK-Twu models using the classical mixing rules, if not for
the generalized correlations -that will be developed later on-, would be in no
position to estimate a kij parameter, thus providing questionable results. The UMR-
PRMC model however is able to construct n.dodecane as a combination of its
respective unifac groups and use the groups respective parameters in the universal
mixing rules, thus resulting in more credible results. It is exactly for this reason that
the UMR-PRMC model can be characterized as a predictive model.
To this purpose, Hg is considered to be a separate Unifac group and based on its

solubility data with other HC, the parameters in question are estimated using the
Neqsim-tool.
The objective function used by Neqsim for the fitting is:

)
) (eq.36), where x symbolizes the mole fraction of Hg.
For the usage of the UMR-PRMC model except for the group interaction parameters,
two more parameters for Hg are essential. One of them is the relative Van der Waals
volume parameter (r), which is equal to 10.598 and the other one is the relative Van
der Waals surface area parameter (q), which is equal to 8.739 [14].
4.6.1 Main groups ACH and ACCH3
As it can be seen at table 26, the experimental data of the aromatic HCs are divided
into two main groups. Table 26 also presents the structure of each HC. This structure
is used for the fitting of the UMR-PRMC parameters. Thus the interaction
parameters between main groups of Hg, ACH and ACCH3 can be estimated given the
available data from the solubility database.
Table 26: Unifacs group structure of each aromatic HC at the solubility database
Unifac group ACH ACCH3

Components
benzene 6 -
toluene 5 1
o-xylene 4 2
55
First of all for the parameters of Hg with groups ACH and ACCH3 two approaches
were adopted. As table 26 shows only benzene [13], toluene [13] and o-xylene [13]
are composed out of unifac main groups ACH and ACCH3. Therefore there is lack of
many experimental solubility data which is the reason leading to the two
approaches.
The first approach is to simultaneously fit the experimental data of all three HCs into
two sets of Aij and Bij parameters. Thus Aij and Bij sets of parameters will be
calculated simultaneously for interactions between groups Hg-ACH and Hg-ACCH3.
The ij index stands for interaction between groups i and j. The second one is to fit
the Aij and Bij parameters first to benzene and then while keeping these as constants
for group interaction Hg-ACH, to fit the set Aij and Bij parameters (the interaction of
groups Hg-ACCH3) for toluene and o-xylene.
Table 27: Estimated unifac group interaction parameters between ACH, ACCH3 and
Hg estimated with both approaches for the UMR-PRMC model
i j Aij [K] Bij [-] Cij [K-1] Aji [K] Bji [-] Cji [K-1] Type of
fitting
ACH Hg Confidential data Simultaneous
ACCH3 Hg fit
ACH Hg Confidential data Separate fit

ACCH3 Hg
As table 27 presents the parameters estimated for both cases are similar to each
other. In addition to that, as tables F17-19 and F20-22 at appendix F show, the
results of the UMR-PRMC model using the estimated parameters are close to the
experimental data for both cases. It is mentioned that only parameters AACH-Hg, BACH-
Hg, AACCH3-Hg and BACCH3-Hg parameters estimated so far. The parameters AHg-ACH, BHg-
ACH, AHg-ACCH3 and BHg-ACCH3 had no effect on the results and were thus set equal to
2000 each Aji and 0 each Bji respectively.
Figures 16-18 graphically present the experimental data and the results of the UMR-
PRMC model using the group interaction parameters for each aromatic HC. The sep
means that these are the results from UMR-PRMC using the parameters estimated
from the separate fitting and the sim from the simultaneous fitting.
Table 28 shows that for the separate fitting the absolute average deviations of
toluene and o-xylene are slightly worse than those of the simultaneous fitting,
however there is a notable difference when it comes to benzene. As expected the
absolute average deviation of the separate fitting for benzene is better compared to
that of the simultaneous fit because the interaction parameter of Hg and ACH are
56
estimated based solely on experimental data concerning these groups. From these
results it can be concluded that the parameters from the simultaneous fitting should
be used because the data used for this fitting are used all together and not
separately, like the other case, thus making the parameters estimated more reliable.
Table 28: Absolute average deviation of the mole fraction of Hg for each approach in
the liquid phase
Simultaneous fit
HC toluene o-xylene benzene
Abs.av.dev. % 4.67 2.81 2.41
Separate fit
HC toluene o-xylene benzene
Abs.av.dev. % 4.27 2.49 1.59
Finally table 29 presents the overall absolute deviations for each binay mixture with
the SRK-Twu and the PR-MC models. It can be seen that all three models have a
similar absolute average deviation for these mixtures. This is very promising for the
UMR-PRMC model given the fact that the fitting of the unifac group parameters has
been performed simultaneously for all three binary mixtures.
Table 29: Overall absolute deviations of SRK-Twu and PR-MC models for Hg with
benzene, toluene and o-xylene respectively in the liquid phase
Model SRK-Twu PR-MC
System Overall Abs. Deviation (%)

Hg-benzene 1.99 2.11
Hg-toluene 4.01 3.83
Hg-o-xylene 2.91 3.52
2.50E-06
mole fraction of Hg
2.00E-06
1.50E-06
1.00E-06 exp
benzene (sep)
5.00E-07 benzene (sim)
0.00E+00
285 290 295 300 305 310
T[K]
Figure 16: Experimental and calculated mole fractions of Hg estimated with the
UMR-PRMC model for benzene in the liquid phase
57
2.50E-06
2.00E-06
mole fraction of Hg
1.50E-06
1.00E-06 exp
toluene (sep)
5.00E-07
toluene (sim)
0.00E+00
270 280 290 300 310
T[K]
UMR-PRMC model for toluene in the liquid phase
3.00E-06
2.50E-06
mole fraction of Hg
2.00E-06
1.50E-06
exp
1.00E-06
o-xylene (sep)
5.00E-07 o-xylene (sim)
0.00E+00
270 280 290 300 310
T[K]
UMR-PRMC model for o-xylene in the liquid phase
4.6.2 Main group CH2
The next Unifac parameters estimated are the ones for main groups of Hg and CH2.
Group CH2 contains all the normal and branched alkanes [11, 13] of the solubility
database in the liquid phase as it can be seen ta table 30. There are also other
components in the database which are also normal alkanes like methane [11] or
ethane [11] for instance, but these are considered to be individual groups, therefore
will be presented at a following subchapter separately.
58
Table 30: Unifacs group structure of each normal and branched alkane in the
solubility database
Unifac group CH2

Unifac subgroup CH3 CH2 CH C
Components
n.C3 2 1 - -
n.C5 2 3 - -
n.C6 2 4 - -
n.C7 2 5 - -
n.C8 2 6 - -
n.C10 2 8 - -
2.2-dm-C4 4 1 - 1
2.2.4-tm-C5 5 1 1 1
For the cyclo-alkanes, a new group will also be introduced later on because it was
observed that by fitting them with main group CH2, they could not be predicted well
enough. The interaction parameters between main groups of Hg and CH2 can be
estimated given the available data from the solubility database. Table 31 presents
the estimated parameters and some information about the results and table 32 the
absolute average deviation of the mole fraction of Hg estimated by the UMR-PRMC
model for each HC respectively, while using the Aij and Bij parameters for the
corresponding groups. It is mentioned that only parameters ACH2-Hg and BCH2-Hg were
estimated. The parameters AHg-CH2, BHg-CH2 had no effect on the results and were thus
set equal to 2000 each Aji and 0 each Bji respectively.
Table 31: Estimated unifac group interaction parameters between CH2 and Hg for the
UMR-PRMC model
i j Aij [K] Bij [-] Cij [K-1] Aji [K] Bji [-] Cji [K-1]
CH2 Hg Confidential data
Number Of Data Points 79

Abs.dev (%) 11.54
Table 32: Absolute average deviation for the mole fraction of Hg in each HC, as
estimated by the UMR-PRMC model in the liquid phase
Hydrocarbons Absolute deviation%

n.C3 11.35
n.C5 5.99
n.C6 5.57
n.C7 7.93
59
n.C8 12.28
n.C10 19.96
2.2-dm-C4 13.55
2.2.4-tm-C5 35.24
As table 32 suggests the results of the UMR-PRMC model with the fitted parameters
are not as good as the ones with the PR-MC and the SRK-Twu models. Tables F3-10
and figures F3-10 at appendix F present the results of the UMR-PRMC model
compared to the experimental data for each HC individually.
Furthermore as table 32 shows, only n.C5, n.C6 and n.C7 have an overall absolute
deviation of less than 10%. However all overall deviations of all binary mixtures
except for 2.2.4-tm-C5 and n.C10 are less than 14%, thus making the results very
accurate. These deviations none the less can be completely justified if one takes into
account the fact that the UMR-PRMC is a model based on the unifac groups as
mentioned, thus has no exclusive parameters for each binary mixture as it occurs for
the SRK-Twu and the PR-MC models, which present generally slightly better overall
deviations as table 22 shows.
As far as propane is concerned the UMR-PRMC model predicts its experimental data
accurately enough as figure 19 illustrates. As far as the other normal alkanes are
concerned the UMR-PRMC model does not predict the experimental data with the
same accuracy. It can be seen that except for n.C10, for all other alkanes the model
can predict the mole fraction of Hg at low temperatures, however as the
temperature rises, the accuracy of the model declines and under-predicts the
experimental data.
For the branched alkanes the UMR-PRMC model seems to constantly over-predict
the mole fraction of Hg in them. The same thing happened even if the corresponding
group was fitted exclusively to these two alkanes. This means that the UMR-PRMC
model appears to fail to accurately describe this type of HCs. That fact could pose a
problem for the model, however due to the overall deviations of these particular
HCs, where one is quite acceptable and one is not, it is not safe to assume anything
further than that one has to be cautious when using it with this type of HCs.
Finally figures 20-23 show the solubility of Hg for the HC, which are grouped
according to their carbon number. One thing that can be said about these figures is
that the calculated by the UMR-PRMC model values of the mole fraction of Hg in n.C8
and 2.2.4-tm-C5 and respectively in n.C6 and 2.2-dm-C4 are very close to each other.
60
3.50E-07
3.00E-07
2.00E-07
1.50E-07
UMR-PRMC
1.00E-07 exp
5.00E-08
0.00E+00
270 275 280 285 290 295
T[K]
Figure 19: Mole fractions of Hg in propane in the liquid phase estimated with the
UMR-PRMC model
7.00E-06
6.00E-06
mole fraction of Hg
5.00E-06
4.00E-06 nC6(exp)
3.00E-06 nC6(calc)
2.00E-06 22-dm-C4(exp)
1.00E-06 22-dm-C4(calc)
0.00E+00
270 280 290 300 310 320 330 340
T[K]
Figure 20: Mole fractions of Hg in HCs with CN=6 in the liquid phase, estimated with
the UMR-PRMC model
2.50E-06
2.00E-06
mole fraction of Hg
1.50E-06
1.00E-06 nC7(exp)
nC7(calc)
5.00E-07
0.00E+00
270 280 290 300 310 320
T[K]
Figure 21: Mole fractions of Hg in n.C7 in the liquid phase, estimated with the UMR-
PRMC model
61
2.50E-06
2.00E-06
mole fraction of Hg
1.50E-06
nC8(exp)
nC8(calc)
1.00E-06
224-tm-C5(exp)
5.00E-07 224-tm-C5(calc)
0.00E+00
270 280 290 300 310 320
T[K]
Figure 22: Mole fractions of Hg in HCs with CN=8 in the liquid phase, estimated with
the UMR-PRMC model
3.50E-06
3.00E-06
mole fraction of Hg
2.50E-06
2.00E-06
1.50E-06
1.00E-06 UMR-PRMC
exp
5.00E-07
0.00E+00
270 280 290 300 310 320
T[K]
Figure 23: Mole fractions of Hg in HCs with n.C10 in the liquid phase, estimated with
the UMR-PRMC model
One last thing that has to be examined is the data available for propane regarding its
vapor phase. These data were predicted with B.P.P calculations, the same way as it
was done for the PR-MC and SRK-Twu model. Table 33 and figure 24 present the
predictions from UMR-PRMC. Figure 24 also displays the results of the other two
models, since it is interesting to see how UMR-PRMC copes against them, given the
fact that it didnt have as accurate predictions in the liquid phase as the other two
models.
62
Table 33: Experimental and estimated mole fractions of Hg with C3 in the vapor
phase from the UMR-PRMC model
T[K] P[bar] yexp ycalc Abs. Deviation%

273.15 14.58
278.15 16.02
283.15 14.28
288.15 13.10
Overall deviation % 14.27
3.00E-07
2.50E-07
mole fraction of Hg
2.00E-07
1.50E-07 PR-MC
exp
1.00E-07
SRK-Twu
5.00E-08 UMR-PRMC
0.00E+00
270 275 280 285 290 295
T[K]
Figure 24: Experimental and estimated mole fractions of Hg in propane in the vapor
phase from all models
As figure 24 illustrates all three models provide almost the same predictions at low
temperatures, and only as the temperature rises, one can see a slight distinction
among them. This is of course very encouraging for UMR-PRMC since C3 is a very
important component in natural gas mixtures.
4.6.3 Main group cy-CH2
As already mentioned, for the better prediction of the cyclo-alkanes [13], the new
main group cy-CH2 is going to be introduced. Furthermore the Neqsim-tool [6] was
modified in order to include this new main group and do the fitting of the
parameters. For the interaction parameters between the unifac groups of Hg and
CH2 the values presented at table 31 were used. Once again, as expected, the
63
parameters AHg-cyCH2, BHg-cyCH2 had no effect on the results and were thus set equal to
2000 and 0 respectively. Finally the unifac group interaction parameters between
the groups of CH2 and cy-CH2 are considered to be equal to zero.
Table 34 presents the unifac group structure of all components comprising of the cy-
CH2 group. Tables 35 and 36 show the estimated parameters for this group and the
overall absolute deviation, as well as the absolute average deviation for the mole
fraction of Hg in each HC as estimated by the UMR-PRMC model, respectively.
Table 34: Unifacs group structure of each cyclo-alkane in the solubility database
Unifac group CH2 cy-CH2

Unifac subgroup CH3 cy-CH2 cy-CH
Components
cy-C6 - 6 -
m-cy-C6 1 5 1
cis-1.2-dm-cy-C6 2 4 2
cis-1.4-dm-cy-C6 2 4 2
trans-1.2-dm-cy-C6 2 4 2
Table 35: Estimated unifac group interaction parameters between cy-CH2 and Hg for
the UMR-PRMC model
cy-CH2 Hg Confidential data

Abs.ov.dev (%) 4.97
Table 36: Absolute average deviation for the mole fraction of Hg in each HC, as
estimated by the UMR-PRMC model in the liquid phase
Hydrocarbons Absolute deviation%

cy-C6 5.90
m-cy-C6 5.69
cis-1.2-dm-cy-C6 5.42
cis-1.4-dm-cy-C6 2.98
trans-1.2-dm-cy-C6 3.57
The analytical results of table 36 can be found at appendix F, at tables F11-16.

Figures 25 and 26 illustrate the results of the UMR-PRMC model compared to the
64
experimental data for cy-C6 and m-cy-C6 respectively as indicative ones, since these
are more usual components met in a natural gas mixture. The rest graphical
illustrations of the results are also to be found at appendix F, at figures F11-16.
As a comment one can say that the fitting and in extension the predictions of the
UMR-PRMC model for the components of this group are very satisfactory since they
are in addition to that very close to the predictions of the SRK-Twu and PR-MC
models as well, as table 22 presents. To confirmation of that, the highest overall
deviation met for the UMR-PRMC model is that of Hg with cy-C6, which ammounts
5.90%.
3.00E-06
2.50E-06
mole fraction of Hg
2.00E-06
1.50E-06
1.00E-06 exp
UMR-PRMC
5.00E-07
0.00E+00
285 290 295 300 305 310 315
T[K]
UMR-PRMC model for binary mixture with cy-C6 in the liquid phase
2.50E-06
2.00E-06
mole fraction of Hg
1.50E-06
1.00E-06
exp
5.00E-07 UMR-PRMC
0.00E+00
270 280 290 300 310
T[K]
UMR-PRMC model for binary mixture with m-cy-C6 in the liquid phase
65
4.6.4 Main group CH4
For the binary mixture of Hg in methane [11] the Neqsim-tool [6] was used for the
fitting of the unifac-group parameters of the UMR-PRMC model, as well as for the
prediction of the mole fractions of Hg in the vapor phase, in which the experimental
data are available. Table 37 presents the unifac parameters estimated for the main
groups of Hg and CH4.
Table 37: Estimated unifac group interaction parameters between CH4 and Hg for the
UMR-PRMC model
CH4 Hg Confidential data

Abs.ov.dev (%) 2.26
As table 37 presents the parameters have been fitted very satisfyingly to the
experimental data since the overall absolute deviation of the model is just 2.26%.
Figure 27 also confirms the good performance of the UMR-PRMC model given the
fact that its predictions match the experimental data almost precisely. The analytical
results can be found at Appendix F, at table F1.
9.00E-08
8.00E-08
7.00E-08
mole fraction of Hg
6.00E-08
5.00E-08
4.00E-08 exp
3.00E-08 UMR-PRMC
2.00E-08
1.00E-08
0.00E+00
240 250 260 270 280 290 300
T[K]
Figure 27: Experimental and predicted mole fractions of Hg in CH4 with the UMR-
PRMC model in the vapor phase
66
Figures 28 and 29 present predictions of the PR-MC and the SRK-Twu models
regarding the mole fractions of Hg in CH4. Table 38 re-presents the absolute overall
deviations of these models for this mixture. As table 38 shows, the deviations are
almost the same as the one of the UMR-PRMC model.
Table 38: Overall absolute deviations of SRK-Twu and PR-MC models for Hg and CH4
in the vapor phase
Model kij parameter Abs. Deviation (%)

SRK-Twu 2.35
PR-MC Confidential data 2.22
Like the UMR-PRMCs model, these models predictions match almost perfectly the
experimental mole fractions of Hg in the vapor phase as shown at table 23 and
figures 28 and 29. This is very important to know because the primary component of
every natural gas mixture is methane. Therefore it is utterly important that the
models are able to accurately predict the composition of Hg in the vapor phase in
this binary mixture, thus making the prediction of the final composition of Hg in the
sales gas of a natural gas processing plant more reliable. The analytical results of the
SRK-Twu and PR-MC models are presented at Appendix E and tables E1 and E22.
9.00E-08
8.00E-08
7.00E-08
mole fraction of Hg
6.00E-08
5.00E-08
4.00E-08 exp
3.00E-08 SRK-Twu
2.00E-08
1.00E-08
0.00E+00
240 250 260 270 280 290 300
T[K]
Figure 28: Experimental and predicted mole fractions of Hg in CH4 with the SRK-Twu
model
67
9.00E-08
8.00E-08
7.00E-08
mole fraction of Hg
6.00E-08
5.00E-08
4.00E-08 exp
3.00E-08 PR-MC
2.00E-08
1.00E-08
0.00E+00
240 250 260 270 280 290 300
T[K]
Figure 29: Experimental and predicted mole fractions of Hg in CH4 with the PR-MC
model
4.6.5 Main group C2H6
For the binary mixture of Hg in ethane [11] there are, as mentioned, experimental
data available in the liquid phase, as well as in the vapor phase. For the fitting of the
parameters required for the UMR-PRMC model, the data of the liquid phase have
been used. Afterwards these parameters were used with in order to predict the
experimental data in the vapor phase. For the parameter fitting and the prediction of
the mole fractions of Hg in the vapor phase, the Neqsim-tool [6] was employed once
again, doing B.P.P in the same fashion as described for previous binary mixtures as
well.
If assumed that i represents the C2H6 group and j the Hg group, the unifac
parameters estimated from the liquid phase data are the ones presented at table 39.
Table 39: Estimated unifac group interaction parameters between C2H6 and Hg for
the UMR-PRMC model
C2H6 Hg Confidential data

Abs.ov.dev (%) 0.27
68
The analytical results from the fitting can be located at appendices F, at table F2. It is
worth mentioning that the UMR-PRMC model clearly has an advantage when it
comes to the description of the liquid phase of Hg in ethane compared to the other
two models because their overall absolute deviation was about 5.7% as table 40
illustrates.
Table 40: Overall absolute deviations of SRK-Twu and PR-MC models for Hg and C2H6
in the liquid phase

SRK-Twu 5.82
Table 41 and figure 30 present the predictions of UMR regarding the vapor solubility
of Hg in ethane. It is interesting to also compare the predictions of the other two
models with the ones from UMR-PRMC as they appear at figure 30, because for
propane even though there were deviations regarding the liquid solubility of Hg, in
the vapor phase all three models provided similar results.
Table 41: B.P.P calculations with the UMR-PRMC model for Hg with ethane in the
vapor phase

273.15 8.26
278.15 8.79
283.15 6.19
288.15 2.95
As figure 30 shows the UMR-PRMC model predicts more accurately the mole fraction
of Hg in the vapor phase compared to the other two models. The PR-MC and SRK-
Twu models provide almost identical results with an overall average deviation
between 15% and 16%. Still the predictions of all three models are acceptable since
once again they are very close the non-ideal experimental error margin and within
the ideal experimental error margin as already explained.
69
1.00E-07
9.00E-08
7.00E-08
6.00E-08
exp
5.00E-08
PR-MC
4.00E-08
SRK-Twu
3.00E-08
2.00E-08 UMR-PRMC
1.00E-08
0.00E+00
270 275 280 285 290 295
T[K]
Figure 30: Experimental and estimated mole fractions of Hg in C2H6 in the vapor
phase from all three models
4.6.6 Main group CO2
For the binary mixture of Hg in CO2 [11] there are also experimental data available in
the liquid phase, as well as in the vapor phase. For the fitting of the Unifac
parameters, the data of the liquid phase have been used. Afterwards these
parameters were used with in order to predict the experimental data in the vapor
phase by conducting B.P.P calculations with the same methodology as before with
Neqsim-tool [6]. At Appendices F and G, the results of the fitting and the B.P.P
calculations with the UMR-PRMC model are presented at tables F19 and G7
respectively. Figures F25 and G3 present them graphically as well.
Table 42: Overall absolute deviations of SRK-Twu and PR-MC models for Hg and CO2
in the liquid phase

SRK-Twu 5.24
If assumed that i represents the CO2 group and j the Hg group, the unifac parameters
estimated from the liquid phase data are the ones presented at table 43.
70
Table 43: Estimated unifac group interaction parameters between CO2 and Hg for
the UMR-PRMC model in the liquid phase
CO2 Hg Confidential data

Abs.ov.dev (%) 0.55
Table 44 and figure 31 present the predictions of UMR-PRMC regarding the vapor
solubility of Hg in CO2. It is also interesting to compare the predictions of the other
two models with the ones from UMR-PRMC as they appear at figure 31, to see if
they have the same behavior as they did for the binary mixture of Hg with ethane or
the one with propane.
Table 44: B.P.P calculations with the UMR-PRMC model for Hg in CO2 in the vapor
phase

273.15 7.57
278.15 5.78
283.15 5.19
288.15 3.77
6.00E-08
5.00E-08
mole fraction of Hg
4.00E-08
PR-MC
3.00E-08
SRK-Twu
2.00E-08 exp
UMR-PRMC
1.00E-08
0.00E+00
270 275 280 285 290 295
T(K)
Figure 31: Experimental and estimated mole fractions of Hg in CO2 in the vapor
phase from the all three model
71
From figure 31 it can be concluded that this mixture follows a similar behavior to the
one of Hg with ethane regarding the results of Hg in vapor phase. As figure 31 shows
the UMR-PRMC model provides the most accurate estimations for the solubility of
Hg in the vapor phase in comparison with the other two models. The PR-MC model
provides the least accurate ones. However all models predictions are within both
the ideal and non-ideal experimental error margin. It is reminded that the
average deviation of the PR-MC and the SRK-Twu models is 12.36% and 2.28%
respectively for the vapor phase. This is the reason why at low temperatures the
SRK-Twu model has similar results to the ones from the UMR-PRMC model.
4.6.7 Main group N2
For this binary mixture experimental data are available only in the vapor phase [11],
like it happens for methane. The Neqsim-tool [6] was used for the fitting of the
unifac-group parameters as well as for the prediction of the mole fractions of Hg in
the vapor phase.
Like the previous groups, if i represents the N2 group and j the Hg group, the
Unifac parameters estimated are the following ones:
Table 45: Estimated unifac group interaction parameters between N2 and Hg for the
UMR-PRMC model in the vapor phase
N2 Hg Confidential data

Abs.ov.dev (%) 0.89
In this particular binary mixture all the experimental data have a constant
temperature of 273.15 K. Thus the B parameters, which pose the temperature
dependency of the unifac parameters cannot be estimated.
As figure 32 and table 38 present, the models predictions of the mole fractions of Hg
agree very well with the experimental data. The analytical results and the
experimental data are presented at Appendix F at table F23. Also figure F17 presents
a visual view of the results.
Like the other two models, UMR-PRMCs predictions match almost perfectly the
experimental mole fractions of Hg in the vapor phase as shown at figure 32, exactly
like it happened for the binary mixture of Hg with CH4. It is reminded that the
72
absolute deviations of PR-MC and SRK-Twu are 0.9% and 1.34% as presented at table
46, meaning almost identical to the 0.89% of the UMR-PRMC model.
4.50E-08
4.00E-08
3.50E-08
mole fraction of Hg
3.00E-08 exp
2.50E-08 UMR-PRMC
2.00E-08
1.50E-08
1.00E-08
5.00E-09
0.00E+00
5 25 45 65
P[bar]
Figure 32: Experimental and predicted mole fractions of Hg in N2 with the UMR-
PRMC model in the vapor phase
Table 46: Overall absolute deviations of SRK-Twu and PR-MC models for Hg and N2

SRK-Twu 1.34
Thus it is concluded that all models predictions are very accurate and acceptable
since they also can be located within both the ideal and the non-ideal
experimental error margin regarding this particular binary mixture.
4.6.8 Main group H2O
For the binary mixture of Hg with water as already established the only source of
experimental data is IUPAC [13]. The Neqsim-tool [6] was used for the fitting of the
unifac-group parameters as well as for the prediction of the mole fractions of Hg.
Like the previous groups, if i represents the H2O group and j the Hg group, the
unifac parameters estimated are the following ones:
73
Table 47: Estimated unifac group interaction parameters between H2O and Hg for
the UMR-PRMC model
H2O Hg Confidential data

Abs.ov.dev (%) 0.09
1.00E-07
9.00E-08
8.00E-08
mole fraction of Hg
7.00E-08
6.00E-08
5.00E-08
4.00E-08 exp
3.00E-08
UMR-PRMC
2.00E-08
1.00E-08
0.00E+00
250 300 350 400
T[K]
Figure 33: Experimental and predicted mole fractions of Hg in H2O with the UMR-
PRMC model
As figure 33 and table 39 present, the models predictions of the mole fractions of Hg
agree very well with the experimental data. This is very important and encouraging
because it means that the UMR-PRMC model is at an advantageous position since it
is implied by the fitting that it can be used in process simulations containing water as
well, thus closer to a realistic mixture that can be met in the natural gas industry.
The analytical results and the experimental data are presented at appendix F at table
F24.
In contrast to the other two models, UMR-PRMCs predictions match almost

perfectly the experimental mole fractions of Hg, as shown at figure 33 and table 47.
This is proof that the universal mixing rules are not bounded by the limitations of the
classical mixing rules regarding the hydrogen bond. It is reminded that the absolute
deviations of PR-MC and SRK-Twu are 13.06% and 12.86% as presented at table 20
with a maximum deviation around 27.5% shown by both models at the lowest and
highest temperatures respectively as shown at appendix E at tables E21 and E42
explicitly.
74
4.7 Generalized correlations for the binary interaction parameters
based on the Tb and the CN of the HC
As already mentioned there are not many available data in the literature concerning
binary mixtures consisting of Hg with HCs. There are many HCs not included in the
database like iso-hexane, p-xylene, and of course other branched HCs and HCs with a
CN higher than 10. These data are of great importance for the models under
development given the fact that the feed stream of a natural gas process plant
consists from all those components as well. Therefore it is important that the models
-namely SRK-Twu and PR-MC- will have kij parameters for all components so that
they will be able more accurately predict the amount of Hg that ends up in the
product stream. It is noted that no correlations are needed for the UMR-PRMC
model, since its a predictive model by itself.
One way to deal with this lack of data concerning these and other components as
well, is to use the existing data to create generalized correlations based on a
property of the HCs, which will provide an estimation of the value of the kij
parameters. The two properties that have been chosen are the normal boiling point
(Tb) of the HC and the carbon number (CN).
This choice was based on the fact that in contrast with other properties like the
molecular weight (Mr) or the acentric factor () each HC has a unique Tb, therefore
the correlations will be able to provide more accurate results for each HC. The Tb
values of each HC were found at Hysys software[14]. The choice of the CN property
was also made, as this is the easiest and fastest way for an engineer to produce
these parameters.
From scientific point of view looking at this subject however, it is clearly stated that
the generalized correlations based on the Tb property are the correct ones. This can
be easily justified through an example. If one has to estimate kij parameters for n.C4
and iso-butane for instance, it is obvious that both HCs cannot share the same
parameter. Thus their Tb property can guarantee that their parameters will be
different, something that does not happen with their CNs property.
In order to create these correlations, all mixtures are divided into three categories.
The first one includes all mixtures that contain Hg with paraffinic HC, the second one
includes all mixtures that contain Hg with naphthenic HC and the third one includes
all mixtures that contain Hg with aromatic HC. Tables 48-50 present all three
categories of HC along with their respective Tb values.
Table 48: First category
Binary mixtures of paraffinic HC with Hg

Name Tb [oC]
75
n.C5 36.05
n.C6 68.75
n.C7 98.45
n.C8 125.65
n.C10 174.15
Table 49: Second category
Binary mixtures of naphthenic HC with Hg

Name Tb [oC]
m-cy-C6 100.9
cy-C6 80.75
Table 50: Third category
Binary mixtures of aromatic HC with Hg

Name Tb [oC]
benzene 80.15
toluene 110.65
o-xylene 139.15
Table 51: Generalized correlations of the kij parameters based on the CN for each
category and model along with the R2 of each correlation
Models Paraffinic HC R2 Naphthenic HC R2 Aromatic HC R2

SRK-Twu(all) 0.993 0.979 0.910
PR-MC Confidential data 0.994 Confidential data 0.999 Confidential data 0.910
Table 52: Generalized correlations of the kij parameters based on the Tb [oC] for each
category and model along with the R2 of each correlation
Models Paraffinic HC R2 Naphthenic HC R2 Aromatic HC R2

SRK-Twu(all) 1 0.988 0.921
PR-MC Confidential data 1 Confidential data 0.997 Confidential data 0.921
As it can be seen at tables 51 and 52 all generalized correlations for the HC are linear
ones. This is happening because the second order equations curve, due to their
order, present a minimum at n.C10 and afterwards rise again monotonically. This
means that after a certain Tb the kij parameters provided from these correlations will
76
start rising instead of further decreasing as expected. Moreover the linear
correlations have a very decent R2, therefore their results will be reliable for usage.
At appendix H and figures H1-12 one can see the graphs of the generalized
correlations based both on the CN and the Tb along with the fitted kij parameters for
all models and HC categories. It shall finally be recommended that as far as the
correlation for the paraffinic HC goes, it can be used safely for HC from n.C4 up to
n.C10. For HC with a CN less than 4 the fitted parameters are recommended to be
used.
Table 53 presents the fitted kij parameter values as well as the kij values that are
estimated by both types of generalized correlations. In total the generalized
correlations, as expected, are predicting satisfyingly enough the kij parameters. It can
be observed that as the carbon number of the HCs is declining the value of the kij
parameter is rising for all models when it comes to the paraffinic and iso-paraffinic
HC. The same applies for the aromatic and the naphthenic HC as well.
The greatest deviations between the kij values are as expected those that concern
the binary mixtures of Hg with trans-1.2-dm-cy-C6, trans-1.4-dm-cy-C6 and cis-1.2-
dm-cy-C6 [13]. It is also important to emphasize the fact that these correlations
predict particularly well the kij values of all paraffinic HC for all models because these
are the main HC that compose a natural gas mixture once it is extracted.
Table 53: Values of fitted and estimated from generalized correlations kij parameters
for all models
Binary mixtures of SRK-Twu PR-MC

Hg with HC
Fitted kij Generalized Generalized Fitted kij Generalized Generalized kij
kij (CN) kij (Tb) kij (CN) (Tb)
i-C4
n.C4
n.C5
n.C6
n.C7
n.C8
n.C10 Confidential data Confidential data
2.2-dm-C4
2.2.4-tm-C5
cy-C6
m-cy-C6
cis-1.2-dm-cy-C6
cis-1.4-dm-cy-C6
trans-1.4-dm-cy-C6
trans-1.2-dm-cy-C6
77
benzene
toluene Confidential data Confidential data
o-xylene
4.7.1 Evaluation of the generalized correlations for the binary

interaction parameters based on the Tb and the CN of the HC
After the development of the generalized correlations it is essential that they are put
to the test in order to check their credibility. For this purpose two mixtures, one
ternary and one multicomponent, from the literature have been selected [10, 11], in
order to check the Hg solubility in them. Tables 54 and 55 present the composition
of these mixtures. It is noted that for the first mixture, the fitted parameters of C 3
were used for the testing.
Table 54: Composition of the first mixture
Component
C3
i-C4
Table 55: Composition of the second mixture
Component
n.C4
n.C5
n.C6
Figures 34 and 35 present the Hg solubility in these mixtures predicted from all three
models. In parenthesis next to the SRK-Twu and the PR-MC model, the type of the
generalized correlation used for the estimation of the kij parameters is stated.
78
1.00E-07
mole fraction of Hg
exp
SRK-Twu(CN)
PR-MC (CN)
PR-MC (Tb)
SRK-Twu(Tb)
UMR
1.00E-08
250 255 260 265 270 275 280
T [K]
Figure 34: Hg mole fraction in liquid iso-butane + C3 mixture at 69 bar
1.00E-06
mole fraction of Hg
exp
SRK-Twu(CN)
1.00E-07 PR-MC(CN)
PR-MC(Tb)
SRK-Twu(Tb)
UMR
1.00E-08
250 260 270 280 290 300
T [K]
Figure 35: Hg mole fraction in liquid n.C4 + n.C5 + n.C6 mixture at approximately 21
bar
At appendix J one can find the analytical results concerning figures 34 and 35 at
tables J1-J10. As far as the first mixture is concerned as figure 34 shows, all models
provide satisfyingly enough results. The UMR-PRMC model is the only one over-
predicting the experimental data. All the other models are under-predicting them. It
also appears that the models using the kij parameters based on the CN provide better
results than the ones using the Tb based correlations. This however is to be expected
because as it will be seen in the next chapter, the models using the CN correlations
are able to predict much more accurately the mole fractions of Hg in iso-butane. The
average deviations of the UMR-PRMC, the PR-MC(Tb), the PR-MC (CN), the SRK-
79
Twu(Tb) and the SRK-Twu (CN) models are 8.14%, 11.17%, 5.26%, 12.18% and 6.63%
respectively.
As far as the second multicomponent mixture is concerned, once again as figure 35

displays all models seem to over-predict the experimental data. Still the results are
very satisfactory. The UMR-PRMC model is the one that shows the biggest deviations
compared to the other ones. Furthermore in this case the models using the T b based
generalized correlations appear to have a slight advantage over the other models in
contrast to the previous mixture. The average deviations of the UMR-PRMC, the PR-
MC(Tb), the PR-MC (CN), the SRK-Twu(Tb) and the SRK-Twu (CN) models are 18.55%,
8.33%, 11.72%, 8.18% and 9.87% respectively.
Conclusively one can say that the generalized correlations seem to produce very
good results for the mixtures tested. It is up to the user to decide which type of
generalized correlations will be used depending on the work at hand and the field of
interest.
References:
[1] Peng, D. Y., and Robinson, D. B. (1976). "A New Two-Constant Equation of State".
Industrial and Engineering Chemistry: Fundamentals 15, p. 5964
[2] Soave (1972), G. Equilibrium Constants from a Modified RedlichKwong Equation

of State, Chem. Eng. Sci., 27, p. 1197-12 3
[3] Voutsas et al.(2005), Thermodynamic property calculations with the universal

mixing rule for EOS/GE models: Results with the Peng-Robinson EOS and a UNIFAC
model
[4] E.H.Benmekki et al.(1987), Van der Waals Mixing Rules for Cubic Equations of
State, ACS Symposium Series, Vol. 329, American Chemical Society
[5] Gross J., Sadowski G (2 2), Application of the perturbed chain SAFT equation of
state to associating systems, Industrial and Engineering Chemistry Research, vol. 41,
issue 22, p. 551 -5515
[6] Solbraa Even (2002), ''Measurement and Modelling of Absorption of Carbon

Dioxide into Methyldiethanolamine Solutions at High Pressures'', Norwegian
University of Science and Technology Department of Refrigeration and Air
Conditioning
[7] Mathias, P.M., Copeman, T.W. (1983), Extension of the Peng-Robinson equation
of state to complex mixtures, Fluid Phase Equilibria 13, p. 91-108
80
[8] Twu C. et al (1995),'' A new generalized alpha function for a cubic equation of
state Part 2. Redlich-Kwong equation'', Fluid Phase Equilibria 105, p. 61-69
[9] Prausnitz (1986), John M. et al., MOLECULAR THERMODYNAMICS OF FLUID-

PHASE EQUILIBRIA, Prentice Hall Inc.
[12] V. Louli et al (2012), ''Measurement and prediction of dew point curves of

natural gas mixtures '', Fluid Phase Equilibria 334, p. 2-4
[13] IUPAC, Solubility Data Series, Mercury in liquids, compressed gases molten
salts and other elements, Volume 29, p. 1 1-160
[14] Hysys simulation programs database, v.7.2
81
Chapter 5: Testing of PR-MC, SRK-Twu and UMR-PRMC
model in binary and multicomponent mixtures
5.1 Iso-butane
For research purposes, once a database is complete, it is useful to separate it into
two sections. One that will be used for the fitting of the required parameters and
one that will be used for predictions with the models developed. Thus one can have
a first evaluation of the models developed. In this master thesis the second part is
consisted of two multicomponent mixtures of Hg and experimental data on the
solubility of Hg in iso-butane [1]. Only iso-butane was chosen as binary mixture with
Hg for evaluation of the models, due to the lack of many experimental data on the
matter as explained at the 3rd chapter.
For the binary mixture of Hg in iso-butane [1] there are experimental data available
in the liquid phase, as well as in the vapor phase. However this time for the SRK-Twu
and the PR-MC models the parameters are not fitted to the data as explained. This
time the parameters are estimated from the generalized correlations of these
models, both with the ones based on the Tb of the components and the ones based
on the CN of the components.
These parameters are used with each model respectively in order to predict the
experimental data in the liquid phase at first with flash calculations using the
Neqsim-tool. Afterwards by using the liquid mole fractions of Hg estimated, the
Neqsim-tool performs B.P.P calculations in order to predict the vapor mole fraction
of Hg. The same procedure is followed for the UMR-PRMC model as well. However
since it is a predictive model, no new parameters are required for it.
At appendix I and tables I1-I5 one can see the analytical results of these calculations.
Table 56 shows the parameters estimated from the generalized correlations for each
model as well as their deviations from the experimental data. In parenthesis the type
of the correlation is stated. The UMR-PRMC model has no kij parameters, therefore
the corresponding Unifac parameters estimated in chapter 4 are used. Figures 36
and 37 present them graphically.
Table 56: Parameters and deviations all models for Hg and i-C4
Model kij parameter Abs. Deviation in liquid phase (%) Abs. Deviation in vapor phase (%)
SRK-Twu (Tb) 26.68 15.18
PR-MC (Tb) 25.11 12.73
SRK-Twu (CN) Confidential data 16.57 15.12
PR-MC (CN) 14.18 12.61
UMR-PRMC - 9.66 13.23
82
1.80E-07 exp
1.60E-07 UMR
PR-MC(Tb)
1.40E-07
SRK-Twu(Tb)
1.20E-07 PR-MC(CN)
mole fraction of Hg
1.00E-07 SRK-Twu(CN)
8.00E-08
6.00E-08
4.00E-08
2.00E-08
0.00E+00
260 265 270 275 280
T[K]
Figure 36: Experimental and estimated mole fractions of Hg in i-C4 in the liquid phase
from all models and both generalized correlations
3.00E-07 exp
UMR
2.50E-07 PR-MC(Tb)
SRK-Twu(Tb)
2.00E-07
PR-MC(CN)
mole fraction of Hg
SRK-Twu(CN)
1.50E-07
1.00E-07
5.00E-08
0.00E+00
260 265 270 275 280
T[K]
Figure 37: Experimental and estimated mole fractions of Hg in i-C4 in the vapor phase
from all models and both generalized correlations
As far as the mole fractions of Hg in the liquid phase are concerned figure 36 clearly
shows that the UMR-PRMC model can predict them better than the other two
models regardless of the generalized correlation used for the estimation of the k ij
parameters. It is also obvious that the SRK-Twu and PR-MC models using the
generalized correlations based on the CN can estimate better the mole fraction of Hg
in the liquid phase than the ones using the correlations based on the T b.
As far as the mole fractions of Hg in the vapor phase are concerned figure 37 shows
that all models provide almost the same results, which is expected as table 56
indicates at their absolute overall deviations from the experimental data. Thats the
83
reason why UMR-PRMC, SRK-Twu(Tb) and PR-MC(Tb) cover one another at figure
37. Generally the PR-MC model seems to have a slight advantage over the other
ones as tables I1-I5 at appendix I show and also as figure 37 implies. It should also be
mentioned the SRK-Twu and PR-MC models make almost identical predictions of the
mole fraction of Hg regardless of the generalized correlation used for the k ij
parameter.
5.2 K variable
After the completion of the estimation of the kij parameters for the SRK-Twu and the
PR-MC models and the unifac parameters for the UMR-PRMC model and the
comparison of the models results with the experimental data, it is also important to
investigate one more aspect for them. This is none other than the K variable. The K
variable is practically the division of the mole fraction of Hg in the vapor phase, with
the mole fraction of Hg in the liquid phase [3].
(eq. 37)
This variable can provide yet another indication of the accuracy of the models.
Obviously for K to be estimated, data in both the vapor and the liquid phase for Hg
are required. It is reminded that the components for which such data are available
are ethane, propane, iso-butane and CO2 [1, 2].
5.2.1 K variable for ethane
The first components K variable presented is the one for ethane. Table 57 presents
the estimated K for ethane of all models as well as their deviations from the
experimental ones for each temperature and pressure and figure 38 illustrates them
graphically.
Table 57: Experimental and estimated K variables from each model for ethane and
their respective deviations
Model PR-MC SRK-Twu UMR-PRMC

T [K] P [bar] Kexp Kcalc % Kcalc % Kcalc %
273.15 22.81 24.13 9.60
278.15 17.54 18.63 9.18
283.15 13.40 14.30 7.15
288.15 Confidential 10.64 Confidential 11.43 Confidential 3.79
293.15 Confidential data data 6.52 data 6.90 data 2.91
84
6.00E-01
5.00E-01
4.00E-01
'K' variable
3.00E-01 Kexp=yexp/xexp
PR-MC
2.00E-01
UMR-PRMC
1.00E-01 SRK-Twu
0.00E+00
270 275 280 285 290 295
T[K]
Figure 38: Experimental and estimated K variables from each model for Hg in
ethane
As figure 38 shows the UMR-PRMC model describes the K variable better than the
other two models. These show almost the same behavior, meaning that they begin
with a deviation around 23%, which as the temperature rises, declines for both
models to around 6.7%. Furthermore these models estimate almost the same K
variable with the PR-MC model to be slightly more accurate. It is also interesting to
note the fact that the K variable is rising along with the temperature.
The results of all models can be justified by the deviations they present -as shown at
tables 22, 23, 39, 41 and figure 30- in the estimation of the mole fraction of Hg in the
vapor phase, as well as in the liquid phase, although the kij parameters were fitted to
the experimental data of the liquid phase. In addition to that, these deviations
regarding the K variable are quite acceptable given the deviations the models
presented in the prediction of the mole fractions of Hg in both the liquid and the
vapor phase.
5.2.2 K variable for propane
The next components K variable presented is the one for propane. Table 58
presents its estimated K for all models as well as their deviations from the
experimental data for each temperature and pressure. Figure 39 illustrates the
results graphically.
85
Table 58: Experimental and estimated K variables from each model for propane and

273.15 13.09 14.53 5.85
278.15 13.50 14.90 5.55
283.15 12.19 13.57 3.43
288.15 Confidential Confidential 10.93 Confidential 12.27 Confidential 1.46
293.15 data data 10.11 data 11.40 data 0.06
7.60E-01
7.40E-01
7.20E-01
7.00E-01
'K' variable
6.80E-01
SRK-Twu
6.40E-01 PR-MC
6.20E-01 UMR-PRMC
6.00E-01
270 275 280 285 290 295
T[K]
Figure 39: Experimental and estimated K variables from each model for Hg in
propane
As figure 39 shows the UMR-PRMC model once again is able to describe the K
variable better than the other two models. For propane the PR-MC and SRK-Twu
models present once more almost the same behavior, meaning that they begin with
a deviation around 13% and 14% respectively, which as the temperature rises,
declines - slower compared to the corresponding rate presented for Hg in ethane -
for both models to around 10% and 11% respectively. It is also noted that for this
binary mixture the K variable rises along with the temperature for the first 10 K but
after that remains relatively the same.
The results of these models can be explained by the deviations they present -as
shown at tables 23, 33 and figure 24- in the estimation of the mole fraction of Hg in
the vapor phase of the binary mixture, since the kij parameters were fitted very well
to the experimental data of the liquid phase, whereas the unifac parameters had a
deviation of 11.35%.
86
5.2.3 K variable for CO2
he penultimate component providing the K variable is CO2. Table 59 presents its

estimated K for all models as well as their deviations from the experimental data
for each temperature and pressure. Figure 40 illustrates the results graphically.
Table 59: Experimental and estimated K variables from each model for CO2 and

34.92 19.93 15.00 7.16
39.74 16.13 11.60 6.79
45.12 10.75 6.30 4.69
Confidential Confidential Confidential Confidential
50.98 5.96 1.73 3.37
data data data data
57.40 2.30 1.38 3.03
8.00E-01
7.50E-01
7.00E-01
'K'variable
SRK-Twu
6.00E-01
PR-MC
5.50E-01 UMR-PRMC
5.00E-01
270 275 280 285 290 295
T[K]
Figure 40: Experimental and estimated K variables from each model for Hg in CO2
As figure 40 shows the model that least accurately describes the K variable is PR-
MC. It presents a maximum deviation of 19.93% at 273.15 K, however as the
temperature rises this deviation is declining and at 293.15 K it is just 2.30%.
Therefore given the deviations this model has in the prediction of the mole fraction
of Hg in the vapor and the liquid phase -as shown at tables 22, 23 and figure 31- , the
results of the estimation of the K variable are considered to be expected and
acceptable.
The SRK-Twu model is the one estimating the variable better than the other two
models for temperatures equal or higher than around 285 K. It is also the first time
that a model appears to over-predict this variable at the temperature of 293.15 K by
1.38%. Both the SRK-Twu and the PR-MC model however appear to have a
87
constantly rising prediction for the K variable in contrast with the UMR-PRMC
model. Of course the results of this model are also expected and acceptable for the
same reasons mentioned for the PR-MC model.
The UMR-PRMC model appears to be the most reliable one compared to the other
two models. This is said not only because it is able to more accurately predict the
mole fraction of Hg in both the vapor and the liquid phase, but also because it
systematically under-predicts the K variable by about 3% - 7%. That means that its
estimations are consistent and there appears to be no danger to over-estimate the
variable, as the temperature rises, by a big deviation. The trend-lines of the other
models imply that when the temperature rises higher than the data available, they
will over-predict the variable by a significant deviation, given the fact that from
figure 40, K, constantly appears to be around 0.745 and 0.770.
5.2.4 K variable for iso-butane
The last component examined for the K variable is iso-butane. This component is
special compared to the other ones presented so far. This is because its data were
immediately predicted from the models and were not included in the database used
for the fitting of the respective parameter for each model.
Tables 60 and 61 present its estimated K for all models as well as their deviations
from the experimental data for each temperature and pressure. Figure 41 illustrates
the results graphically. In parenthesis is stated the type of the generalized
correlation used for the estimation of the kij parameters for the SRK-Twu and the PR-
MC models.
Table 60: Experimental and estimated K variables from UMR-PRMC for iso-butane
and their respective deviations
Model UMR-PRMC
T [K] P [bar] Kexp Kcalc %
263.15 24.86
268.15 21.78
273.15 18.57
278.15 Confidential Confidential 17.96
263.15 data data 24.86
88
Table 61: Experimental and estimated K variables from each EOS for iso-butane and
Model PR-MC (CN) PR-MC (Tb) SRK-Twu (CN) SRK-Twu (Tb)

T [K] P [bar] Kexp Kcalc % Kcalc % Kcalc % Kcalc %
263.15 0.44 14.71 0.35 14.00
268.15 1.55 16.47 1.51 15.61
273.15 3.68 18.38 3.51 17.39
278.15 Confidential Conf. 2.54 Conf. 16.57 Conf. 2.25 Conf. 15.74
263.15 data data 0.44 data 14.71 data 0.35 data 14.00
2.40
Kexp=yexp/xexp
2.20 UMR
2.00 PR-MC(Tb)
SRK-Twu(Tb)
'K' variable
1.80
PR-MC(CN)
1.60 SRK-Twu(CN)
1.40
1.20
1.00
260 265 270 275 280
T[K]
Figure 41: Experimental and estimated K variables from each model for Hg in iso-
butane
First of all the UMR-PRMC model is the only model under-predicting the K variable
in contrast with the other models. The deviations are between 18% and 25%. These
deviations are to be expected in a way, given the fact that the unifac groups
describing this component have been fitted to a large amount of experimental data
and therefore naturally present deviations at the estimation of the mole fraction of
Hg in the vapor, as well as in the liquid phase. Worth-noting is also the fact that the
K estimated from this model is almost constant and around 1.4 and 1.41.
The PR-MC and SRK-Twu models using the CN based generalized correlations provide
K variables almost in complete agreement to the experimental ones. This is
expected of course since as figure 41 shows, these correlations describe with almost
the same deviations the experimental mole fractions of Hg in both liquid and vapor
phase.
89
The PR-MC and SRK-Twu models using the Tb based generalized correlations in
contrast to the previous ones over-predict the experimental Ks. This is expected
because as figure 36 implies as well, these models under-predict the liquid mole
fraction of Hg. However since their deviations are between 18.4% and 14%, their
respective results are also considered to be acceptable.
In conclusion the K variables of four different binary mixtures have been estimated
and compared to their respective experimental data. The results of all models
examined can be justified by the analysis of chapter 4 for all binary mixtures. Thus all
models -and generalized correlations- appear to be in position to provide reliable
results concerning the distribution of Hg in a natural gas process. Finally based on
the analysis done, the UMR-PRMC model appears to be the most consistent one in
terms of accuracy -over or under-estimation of the experimental K variables- and
deviations compared to the other models.
5.3 Multi-component mixtures

After the completion of the estimation of all necessary parameters for the models
depending on binary mixtures and vapor pressure data for Hg, it is important to see
how they cope against experimental data regarding more complex than binary
symstems containing Hg. That way one can be more certain that the results
produced by them regarding the Hg distribution within a process will be more
reliable. This is the final and most important test for the models, since there are no
experimental data from any process that can confirm the results for it.
For this testing two different mixtures will be examined. Tables 62 to 63 present
their components and their respective composition. It is mentioned that the
experimental data of the first mixture are available at two different pressures,
namely at 27.58 bar and at 69 bar. The second mixture has data available at
approximately 21 bar. Also at appendix D at table D3 one can find the unifac group
parameters between all other groups, except for Hg, used for the prediction of the
two mixtures with the UMR-PRMC model.
Table 62: Composition of the first mixture
Component
CH4
C2H6
C3
N2
CO2
Sum
90
Table 63: Composition of the second mixture
Component
n.C4
n.C5
n.C6
Sum
Tables 64 and 65 as well as figures 42 and 43 illustrate the results concerning the
first multicomponent mixture. It can be seen that all three models are able to predict
very accurately the experimental data. Also as table 64 indicates, at low
temperatures the models under-predict the experimental data at 69 bar and at 20 oC
they over-predict them. However at 27.58 bar they constantly under-predict the
data. In total all models present very close results with one another. This is also the
reason why at figures 42 and 43 the line presenting the results from the PR-MC
practically covers the one from the SRK-Twu model. This can be made even clearer
by comparing the results displayed at tables 64 and 65 for each model against the
other two.
Finally as figures 42 and 43 illustrate all models fail to accurately predict the
experimental data at 10 oC at both pressures. Still the figures imply that the data
follow a linear trend-line. Thus one may assume that at this particular temperature
there has been some sort of experimental error.
Table 64: Experimental data, conditions and results of the first mixture with all three
models
Conditions Model SRK-Twu PR-MC UMR-PRMC

P [bar] T [oC] Exp. Mole fraction of Hg Calc. Mole fraction of Hg
Confidential data Confidential data Confidential data
91
1.00E-07
mole fraction of Hg
exp
1.00E-08
SRK-Twu
PR-MC
UMR
1.00E-09
-15 -10 -5 0 5 10 15 20 25
T [oC]
Figure 42: Experimental and predicted mole fractions of Hg in the first

multicomponent mixture with the all three models at 69 bar
Table 65: Experimental data, conditions and results of the first mixture with all three
models
Conditions Model SRK-Twu PR-MC UMR-PRMC

P [bar] T [oC] Exp. Mole fraction of Hg Calc. Mole fraction of Hg
1.00E-07
mole fraction of Hg
1.00E-08 exp
SRK-Twu
PR-MC
UMR
1.00E-09
-15 -10 -5 0 5 10 15 20 25
T [oC]
Figure 43: Experimental and predicted mole fractions of Hg in the first

multicomponent mixture with the all three models at 27.58 bar
92
Tables 66 and 67 and figure 44 present the experimental data and the results of the
second multicomponent mixture. It is mentioned that for this mixture the k ij
parameters for n.C4 have been estimated by both types of generalized correlations,
which are shown in parenthesis next to the models. The rest kij parameters taken
into employment from the models are the fitted ones.
Table 66: Experimental data, conditions and results of the second mixture with the
UMR-PRMC model
Conditions Model UMR-PRMC

P [bar] T [oC] Exp. data Calc. Mole fraction of Hg
Table 67: Experimental data, conditions and results of the second mixture with the
PR-MC and SRK-Twu models using the generalized correlations for the kij of n.C4
Conditions Model SRK-Twu (CN) SRK-Twu (Tb) PR-MC (CN) PR-MC (Tb)
P [bar] T [oC] Exp. data Calc. Mole fraction of Hg
Confidential data Confidential Confidential data

data
1.00E-06
mole fraction of Hg
exp
1.00E-07
UMR
SRK-Twu(CN)
PR-MC(CN)
PR-MC(Tb)
SRK-Twu(Tb)
1.00E-08
-20 -15 -10 -5 0 5 10 15 20 25
T [oC]
Figure 44: Experimental and predicted mole fractions of Hg in the second

multicomponent mixture with the all models at approximately 21 bar
93
It can be seen that all three models are able to predict very well the experimental
data. Also table 67 indicates the differences between the models using the CN based
and the Tb based generalized correlations are practically negligible. This can be
attributed to the fact that for the paraffinic HCs, both had a R2 around 0.99. In
addition to that as table 53 illustrates, both correlations estimate very similar
parameters for n.C4.
Conclusively it appears that all three models can be characterized as trustworthy for
the prediction of the Hg distribution in multicomponent mixtures as well as in binary
mixtures. It is also underlined that this happens despite the fact that all models were
not able to accurately estimate the K variable of propane and the PR-MC and SRK-
Twu the one for ethane. In addition it is also important to specially emphasize the
fact that the multicomponent mixture of table 62 basically resembles a typical feed
stream of any natural gas industry. Therefore the fact that the models predictions
match very well the experimental data is also an indication that the models can be
used later on for process simulations. Furthermore the results of this particular
mixture are within the range of the study presented by Carnell P. et al, at table 2,
fact which adds credibility to the study, the experimental data and the models as
well.
References:
[3] Tassios et al (2001), ''Applied Thermodynamics of Chemical Engineering'', NTUA,

p. 521
[4] Michael M.Abbott, Hendrick C. Van Ness (1983), '''', Translated by

Konstantinos . Pattas, Nikolaos A. Kiriakis, Schaum outline series, Mc Graw-Hill,
New York, p. 238-239, 262-263
[5] Carnell P. et al (2004),'' MERCURY DISTRIBUTION ON GAS PROCESSING PLANTS '',

83rd annual GPA convention proceedings, New Orleans.
94
Chapter 6: Conclusions
In this master thesis three thermodynamic models have been developed, using the
PR and the SRK EOSs as their basis, in order to predict the elemental Hg distribution
throughout a natural gas processing plant.
The first step was the evaluation of two different sets of critical properties about Hg
that are suggested in literature. One of them was proposed by a report from NIST
and the other one was proposed by DIPPR. Since the set of NIST has no real
advantage over the other set and in addition the set from DIPPR is used by official
softwares like Hysys and PROII it was decided that this set will be used for the rest of
the thesis.
The second step was the compilation of the data regarding the P s of Hg. These data
were used afterwards in order to estimate the Mathias-Copeman parameters for the
PR EOS and the Twu-Coon parameters for the SRK EOS for a temperature range from
the triple point of Hg, up to its critical point. These parameters were incorporated
via the alpha parameter in the EOSs and with them the ability of the models to
accurately predict the Ps of elemental Hg improved significantly.
The third step was the creation of a database concerning the solubility of Hg in HCs
and other components. This database contains LLE data for all binary mixtures
except for those concerning Hg in propane CO2, ethane and iso-butane, which are
VLE data. Furthermore there are data regarding the solubility of Hg only in the vapor
phase, namely those of Hg in methane and N2. These data have been evaluated and
then used in order to estimate the binary interaction parameters between Hg and
each HC as constants for each model respectively. These parameters assist the
models to provide a better estimation for the phase equilibriums of systems
containing elemental Hg with HCs and other components.
The fourth step was the creation of temperature dependent binary interaction
parameters for Hg and water. This was found to be necessary because with the
respective constant parameters all models failed to accurately predict the phase
equilibriums of this particular binary system. This can be attributed to the hydrogen
bond of water that the classical EOSs like PR-MC and SRK-Twu fail to take into
consideration successfully.
The fifth step was the estimation of the Unifac group contribution parameters for
the UMR-PRMC model for all groups of which the experimental data consist of, by
using the data of the assembled database. Given the fact that this model is based on
the PR-MC model, no further action was needed for its completion.
95
Due to the lack of experimental data regarding the solubility of Hg in many HCs, the
sixth step was the creation of generalized correlations for each model respectively.
These correlations are able to provide estimation about the binary interaction
parameters that cannot be calculated otherwise. The properties of the HCs chosen
to create these correlations were their normal boiling point and their respective
carbon number. In addition to that, the generalized correlations have been tested
against experimental data regarding a ternary and a multicomponent mixture.
The final part of this thesis was the evaluation of the three models developed as well
as their corresponding generalized correlations, against experimental data
concerning multicomponent mixtures and also against the experimental K variable,
which is the division of the Hg mole fraction in the vapor phase with the Hg mole
fraction in the liquid phase, where possible.
These evaluations had three objectives. The first one was to test how the developed
models cope against more complex systems than just binary mixtures. That way it
would be safer to conclude whether or not the models can be used in process
simulations at a later stage while providing reliable results, given the fact that there
are no experimental data regarding processes. The second one was to explore the
capability of the models to predict the K variable, which is very important when it
comes down to evaluating models in a process. The last one was to find the optimum
type of binary interaction parameters that should be used for process simulations in
the future.
The conclusions can be summarized as follows:
1. All models, namely the PR-MC, the SRK-Twu EOS and the UMR-PRMC are able
to accurately predict the mole fraction of elemental Hg both in binary and
multicomponent mixtures.
2. All models are capable of predicting the K variable of the corresponding

mixtures very accurately.
3. As far as the optimum type of the generalized correlation is concerned, both

types appear to provide very similar results in all cases, except for the
mixtures containing iso-butane, where the correlations based on the CN
seem to have the advantage. Therefore it is concluded that it is up to the user
to decide which type is appropriate according to the occasion, since both
types provide reliable results.
4. All models and generalized correlations can be used in process simulations at

a future stage with some caution regarding the branched alkanes and the
96
aromatic HCs due to their tentative experimental data, providing reliable
results.
97
Chapter 7: Future work
The models created are able to predict the distribution of Hg in binary, ternary and
multicomponent mixtures. However there are several steps necessary to be made, in
order to improve their predictive abilities and also assist in a further evaluation for
them. Some options include:
1. The usage of the models at a real process of a natural gas plant where the
distribution of elemental Hg throughout the plant will be evaluated.
2. The veryfication of the results of the models via experimental data from the
process.
3. The extension of the database concerning the solubility of Hg in HC for binary

mixtures of Hg with components heavier than n.C10, so that binary
interaction parameters can be fitted for these components as well.
4. The extension of the same database for the solubility of Hg in other

components like H2S and He, which are of interest to the natural gas industry.
5. The veryfication of tha data included in the database of this master thesis
that are characterized as tentative ones.
6. It has been mentioned that although Hg is mainly found in its elemental form
in the natural gas, this is not its only form that is contained there. Therefore
the chemistry of Hg, meaning its chemical reactions and various forms can
be included in the models as well.
7. There can also be a further extension of the models tested for the prediction
of the distribution of Hg. Except for PR and SRK EOS there are other mode
advanced models. One of them is the PC-SAFT model, which can also make
predictions about the phase equilibrium of water. This is important
enviromentaly speaking because that way the industrial world will be able to
know the composition of their waste in Hg and thus proceed to take any
measures if necessary.
98
Appendix A
Table A1: Experimental data and comparison with results from DIPPRs equation for
Ps
Exp. Data Dippr's eq. Deviation %

Researcher (uncert. level %) T [K] P [Kpa] P [Kpa] ((Pexp-Pcalc)/Pexp)%
417.10 0.293 0.293 -0.07
426.20 0.424 0.426 -0.42
432.28 0.538 0.541 -0.61
439.29 0.706 0.708 -0.27
441.72 0.774 0.775 -0.16
447.68 0.964 0.965 -0.10
451.38 1.101 1.102 -0.10
454.12 1.213 1.214 -0.11
456.32 1.309 1.312 -0.19
462.63 1.627 1.629 -0.13
469.18 2.024 2.027 -0.15
474.57 2.414 2.415 -0.04
479.04 2.784 2.785 -0.02
485.15 3.369 3.368 0.04
Ambrose (<0.03) 491.86 4.128 4.126 0.04
497.53 4.882 4.879 0.07
549.47 19.193 19.175 0.09
554.72 21.742 21.700 0.19
562.76 26.162 26.106 0.21
579.98 38.203 38.112 0.24
589.082 46.244 46.130 0.25
597.32 54.686 54.547 0.25
605.65 64.484 64.316 0.26
611.99 72.866 72.686 0.25
621.15 86.564 86.340 0.26
627.81 97.795 97.543 0.26
628.88 99.711 99.458 0.25
629.95 101.643 101.388 0.25
638.37 118.032 117.728 0.26
639.86 121.141 120.845 0.24
645.49 133.558 133.189 0.28
654.71 155.987 155.590 0.25
663.19 179.295 178.846 0.25
671.78 205.659 205.159 0.24
681.17 237.992 237.429 0.24
685.43 253.933 253.351 0.23
702.72 327.808 327.089 0.22
711.62 371.975 371.197 0.21
99
Continuation of table A1
726.55 456.609 455.737 0.19
739.69 543.039 542.119 0.17
749.79 617.883 616.913 0.16
771.24 802.526 802.659 -0.02
481.65 3.023 3.022 0.03
488.13 3.689 3.688 0.02
494.93 4.522 4.520 0.05
500.62 5.339 5.336 0.06
506.66 6.342 6.336 0.10
513.69 7.708 7.699 0.11
Ambrose (<0.03) 520.26 9.205 9.191 0.15
526.17 10.753 10.739 0.13
533.78 13.074 13.052 0.17
541.59 15.879 15.850 0.18
546.93 18.080 18.045 0.19
555.22 21.998 21.953 0.20
572.032 32.173 32.099 0.23
587.99 45.215 45.103 0.25
596.47 53.760 53.625 0.25
605.05 63.675 63.568 0.17
612.93 74.144 73.999 0.20
621.86 87.728 87.491 0.27
630.16 102.037 101.766 0.27
624.85 92.667 92.429 0.26
627.96 98.061 97.812 0.25
632.34 106.099 105.827 0.26
623.25 89.986 89.755 0.26
624.85 92.662 92.429 0.25
626.41 95.342 95.098 0.26
633.75 108.792 108.519 0.25
636.49 114.202 113.914 0.25
629.47 100.764 100.510 0.25
633.75 108.794 108.519 0.25
635.13 111.489 111.207 0.25
630.92 103.440 103.177 -1.83
629.76 101.325 101.049 0.27
622.96 89.523 89.280 0.27
624.56 92.188 91.936 0.27
Beattie (0.03) 626.055 94.739 94.484 0.27
627.69 97.591 97.327 0.27
629.077 100.075 99.806 0.27
630.58 102.819 102.545 0.27
632.0069 105.472 105.199 0.26
100
633.43 108.168 107.903 0.25
634.71 110.655 110.388 0.24
635.65 112.492 112.230 0.23
634.15 109.565 109.294 0.25
632.80 106.976 106.702 0.26
631.40 104.344 104.071 0.26
629.96 101.688 101.414 0.27
628.42 98.897 98.634 0.27
Beattie (0.03)
626.85 96.126 95.867 0.27
625.39 93.605 93.352 0.27
623.29 90.0733 89.827 0.27
624.99 92.914 92.664 0.27
626.76 95.967 95.710 0.27
631.19 103.940 103.669 0.26
632.63 106.646 106.372 0.26
633.94 109.160 108.891 0.25
634.92 111.047 110.785 0.24
632.26 105.943 105.672 0.26
627.62 97.470 97.204 0.27
285.22 8.45E-05 8.46E-05 -0.13
288.15 0.000111 0.000110 0.97
291.10 0.000145 0.000143 1.25
294.11 0.000185 0.000185 -0.15
297.22 0.000241 0.000241 -0.05
300.25 0.000312 0.000310 0.45
303.18 0.000393 0.000394 -0.03
306.17 0.000501 0.000499 0.25
309.29 0.000636 0.000637 -0.15
Ernsberger (1)
312.11 0.000793 0.000790 0.36
315.15 0.000995 0.000992 0.25
318.17 0.00124 0.00124 0.22
321.15 0.00154 0.00154 0.16
324.12 0.00190 0.00190 0.20
326.63 0.00226 0.002258 0.08
293.24 0.000174 0.000172 0.78
296.22 0.000223 0.000222 0.32
299.20 0.000287 0.000285 0.69
423.10 0.374 0.376 -0.56
464.58 1.736 1.740 -0.21
526.92 10.943 10.949 -0.06
Menzies (0.5) 533.29 12.911 12.890 0.16
537.10 14.201 14.185 0.11
540.92 15.604 15.592 0.08
101
544.75 17.141 17.119 0.13
548.61 18.805 18.785 0.11
556.53 22.641 22.632 0.04
565.97 28.103 28.065 0.14
571.23 31.592 31.539 0.17
577.63 36.429 36.245 0.50
583.62 41.294 41.167 0.31
587.20 44.652 44.367 0.64
599.51 57.186 56.988 0.35
602.72 60.942 60.728 0.35
607.01 66.396 66.041 0.53
611.05 71.667 71.390 0.39
615.66 78.228 77.926 0.39
619.97 84.660 84.475 0.22
Menzies (0.5) 620.57 85.622 85.421 0.23
624.95 93.023 92.598 0.46
625.06 93.022 92.785 0.25
627.66 97.609 97.282 0.33
628.79 99.700 99.292 0.41
630.13 102.050 101.718 0.32
630.54 102.560 102.471 0.09
633.86 108.920 108.731 0.17
635.24 111.510 111.424 0.08
638.43 118.090 117.858 0.20
641.34 124.340 123.989 0.28
645.54 133.660 133.294 0.27
651.46 147.810 147.370 0.30
656.74 161.400 160.923 0.30
659.45 168.730 168.262 0.28
662.35 176.800 176.414 0.22
666.82 190.120 189.604 0.27
677.05 223.350 222.807 0.24
679.94 232.550 232.990 -0.19
683.43 246.290 245.784 0.21
689.93 271.59 271.116 0.17
692.87 283.42 283.241 0.06
699.29 311.21 311.233 -0.01
706.78 346.53 346.641 -0.03
707.47 349.88 350.058 -0.05
1051.44 9006.415 9338.783 -3.69
1186.24 19992.89 20352.414 -1.80
Schonherr (3) 1322.14 38198.32 38520.392 -0.84
1424.66 57101.07 58131.280 -1.80
102
1510.37 78204.03 79191.073 -1.26
1581.99 99103.96 100534.610 -1.44
1632.52 115498.60 117887.026 -2.07
1665.97 127496.70 130509.868 -2.36
Schonherr (3) 1686.68 134500.80 138805.413 -3.20
1704.76 142996.60 146360.096 -2.35
1716.40 148999.00 151382.444 -1.60
1726.05 153498.70 155642.459 -1.40
1735.51 157499.80 159904.817 -1.53
554.11 21.52 21.392 0.59
560.61 25.05 24.860 0.76
567.11 28.93 28.789 0.49
578.91 37.40 37.254 0.39
600.61 58.54 58.247 0.50
604.11 62.61 62.409 0.32
621.11 86.35 86.280 0.08
622.21 88.11 88.053 0.06
623.71 90.38 90.518 -0.15
626.81 95.47 95.793 -0.34
629.41 100.20 100.408 -0.21
635.21 111.70 111.365 0.30
643.31 127.90 128.285 -0.30
619.71 84.19 84.067 0.15
628.11 98.15 98.078 0.07
Shpil'rain (0.6 - 0.8) 628.31 98.89 98.434 0.46
630.91 103.40 103.153 0.24
635.61 112.30 112.155 0.13
639.81 121.00 120.733 0.22
642.11 125.70 125.654 0.04
643.81 129.30 129.394 -0.07
664.30 182.54 182.074 0.26
707.29 349.21 349.164 0.01
742.78 561.47 564.202 -0.49
755.98 663.72 666.639 -0.44
774.67 839.05 836.017 0.36
648.01 138.62 139.027 -0.29
677.30 225.31 223.674 0.73
681.50 237.79 238.641 -0.36
694.70 290.70 291.006 -0.11
724.09 441.07 440.824 0.06
732.39 491.11 492.654 -0.31
752.68 638.37 639.761 -0.22
774.57 834.49 835.030 -0.06
103
786.77 962.52 962.405 0.01
796.86 1079.80 1078.732 0.10
814.46 1305.70 1307.372 -0.13
836.25 1635.51 1640.125 -0.28
803.66 1161.17 1163.067 -0.16
815.46 1319.73 1321.399 -0.13
Shpil'rain (0.6 - 0.8) 822.45 1423.02 1422.710 0.02
831.85 1569.96 1568.223 0.11
845.39 1795.51 1797.513 -0.11
847.24 1825.05 1830.722 -0.31
854.54 1967.28 1966.343 0.05
856.44 1999.68 2002.861 -0.16
866.64 2211.86 2207.780 0.18
880.43 2514.78 2509.577 0.21
882.13 2553.52 2548.829 0.18
883.23 2572.95 2574.472 -0.06
533.83 13.06 13.066 -0.04
549.81 19.337 19.330 0.04
558.95 23.954 23.932 0.09
564.72 27.351 27.289 0.23
565.74 27.964 27.922 0.15
573.61 33.293 33.226 0.20
Spedding (0.03) 586.013 43.39 43.284 0.25
594.74 51.918 51.785 0.26
597.25 54.588 54.474 0.21
604.29 62.792 62.627 0.26
613.87 75.568 75.355 0.28
620.25 85.144 84.922 0.26
630.24 102.22 101.927 0.29
234.30 2.921E-07 2.93276E-07 -0.40
Douglas et al (<1.5) 253.17 0.00000311 3.13695E-06 -0.74
273.15 0.00002661 2.68401E-05 -0.86
104
Table A2 : Analytical results from Neqsim tool using SRK and PR EoS and deviations
from the results of DIPPRs equation.
)
(eq. A1)
T [K] DIPPR Ps Neqsim- Ps % Neqsim- Ps % Neqsim- Ps % Neqsim- Ps %

[bar] SRK (set 1) SRK(set 2) PR(set 1) PR(set 2)
238.15 4.90E-09 6.13E-10 87.49 1.63E-09 66.65 3.30E-09 32.70 9.29E-09 -89.50
248.15 1.73E-08 2.44E-09 85.87 6.14E-09 64.49 1.19E-08 31.45 3.14E-08 -81.66
258.15 5.52E-08 8.72E-09 84.21 2.08E-08 62.39 3.85E-08 30.25 9.66E-08 -74.86
268.15 1.62E-07 2.83E-08 82.50 6.41E-08 60.33 1.15E-07 29.09 2.73E-07 -68.92
278.15 4.37E-07 8.41E-08 80.76 1.82E-07 58.32 3.15E-07 27.97 7.15E-07 -63.70
288.15 1.10E-06 2.31E-07 78.99 4.81E-07 56.36 8.05E-07 26.89 1.75E-06 -59.09
298.15 2.61E-06 5.94E-07 77.21 1.19E-06 54.45 1.93E-06 25.85 4.04E-06 -54.99
308.15 5.82E-06 1.43E-06 75.41 2.76E-06 52.60 4.38E-06 24.85 8.81E-06 -51.33
318.15 1.24E-05 3.26E-06 73.61 6.08E-06 50.80 9.41E-06 23.88 1.83E-05 -48.05
328.15 2.51E-05 7.06E-06 71.80 1.28E-05 49.05 1.93E-05 22.94 3.64E-05 -45.11
338.15 4.86E-05 1.46E-05 70.00 2.56E-05 47.34 3.79E-05 22.03 6.93E-05 -42.44
348.15 9.08E-05 2.89E-05 68.21 4.93E-05 45.69 7.16E-05 21.15 1.27E-04 -40.03
358.15 1.64E-04 5.49E-05 66.42 9.15E-05 44.09 1.30E-04 20.29 2.26E-04 -37.84
368.14 2.85E-04 1.01E-04 64.65 1.64E-04 42.53 2.30E-04 19.47 3.88E-04 -35.84
378.14 4.83E-04 1.79E-04 62.90 2.85E-04 41.02 3.93E-04 18.67 6.47E-04 -34.00
388.14 7.95E-04 3.09E-04 61.17 4.80E-04 39.56 6.53E-04 17.89 1.05E-04 -32.33
398.14 1.28E-03 5.17E-04 59.45 7.89E-04 38.14 0.00106 17.14 0.00167 -30.78
408.14 2.00E-03 8.44E-04 57.77 0.00126 36.76 0.00167 16.42 0.00258 -29.36
418.14 3.06E-03 1.34E-03 56.10 0.00198 35.43 0.00258 15.71 0.00392 -28.05
428.14 4.60E-03 2.09E-03 54.47 0.00303 34.14 0.00390 15.03 0.00583 -26.83
438.14 6.78E-03 3.20E-03 52.86 0.00455 32.89 0.00580 14.37 0.00852 -25.70
448.14 0.0098 4.78E-03 51.28 0.00670 31.68 0.00846 13.74 0.0122 -24.65
458.14 0.0140 7.02E-03 49.72 0.00971 30.51 0.0121 13.12 0.0173 -23.66
468.14 0.0196 1.01E-02 48.20 0.0138 29.37 0.0171 12.53 0.0240 -22.75
478.14 0.0271 1.44E-02 46.71 0.0194 28.27 0.0238 11.95 0.0330 -21.89
488.14 0.0369 2.02E-02 45.26 0.0269 27.21 0.0327 11.39 0.0447 -21.08
498.14 0.0497 2.79E-02 43.83 0.0367 26.18 0.0443 10.86 0.0598 -20.32
508.14 0.0660 3.80E-02 42.43 0.0494 25.19 0.0592 10.34 0.0790 -19.61
518.14 0.0869 5.12E-02 41.07 0.0658 24.23 0.0783 9.84 0.103 -18.93
528.14 0.1130 6.81E-02 39.74 0.0867 23.30 0.102 9.35 0.134 -18.29
538.14 0.1456 8.96E-02 38.44 0.113 22.41 0.133 8.89 0.171 -17.69
548.14 0.1857 1.17E-01 37.18 0.146 21.54 0.170 8.44 0.218 -17.11
558.14 0.2349 1.50E-01 35.94 0.186 20.71 0.216 8.01 0.274 -16.57
568.14 0.2946 1.92E-01 34.74 0.236 19.90 0.272 7.59 0.342 -16.04
105
578.14 0.3664 2.43E-01 33.57 0.296 19.12 0.340 7.19 0.423 -15.55
588.14 0.452 3.06E-01 32.43 0.369 18.37 0.422 6.81 0.521 -15.07
598.14 0.555 3.81E-01 31.32 0.457 17.65 0.519 6.44 0.636 -14.62
608.14 0.675 4.71E-01 30.23 0.561 16.95 0.634 6.08 0.771 -14.19
618.14 0.816 5.78E-01 29.18 0.684 16.27 0.770 5.74 0.929 -13.77
628.14 0.981 7.05E-01 28.16 0.828 15.62 0.928 5.41 1.113 -13.37
638.14 1.173 8.54E-01 27.17 0.997 15.00 1.113 5.10 1.325 -12.98
648.14 1.393 1.028 26.21 1.193 14.39 1.327 4.79 1.569 -12.61
658.14 1.647 1.231 25.27 1.419 13.81 1.573 4.50 1.849 -12.26
668.14 1.937 1.465 24.36 1.680 13.25 1.855 4.22 2.167 -11.91
678.14 2.266 1.734 23.47 1.978 12.71 2.177 3.95 2.528 -11.58
688.14 2.639 2.043 22.62 2.318 12.19 2.542 3.69 2.937 -11.26
698.14 3.0606 2.394 21.78 2.703 11.69 2.955 3.44 3.396 -10.95
708.14 3.534 2.793 20.98 3.138 11.21 3.421 3.21 3.911 -10.65
718.14 4.0641 3.243 20.19 3.627 10.74 3.943 2.98 4.485 -10.36
728.14 4.6555 3.751 19.43 4.176 10.29 4.527 2.76 5.125 -10.08
738.14 5.313 4.320 18.69 4.789 9.86 5.178 2.54 5.834 -9.81
748.14 6.0418 4.956 17.97 5.471 9.45 5.900 2.34 6.619 -9.55
758.14 6.847 5.664 17.28 6.228 9.04 6.700 2.15 7.484 -9.30
768.14 7.734 6.450 16.60 7.0645 8.66 7.583 1.96 8.434 -9.05
778.14 8.709 7.320 15.95 7.987 8.29 8.554 1.78 9.476 -8.81
788.14 9.776 8.279 15.31 9.001 7.93 9.619 1.60 10.615 -8.58
798.14 10.942 9.334 14.70 10.113 7.58 10.785 1.43 11.857 -8.36
808.14 12.214 10.491 14.10 11.384 7.25 12.058 1.27 13.208 -8.14
818.14 13.596 11.758 13.52 12.654 6.92 13.444 1.12 14.674 -7.93
828.14 15.0949 13.139 12.96 14.097 6.61 14.949 0.96 16.261 -7.72
838.14 16.718 14.643 12.41 15.662 6.32 16.581 0.82 17.976 -7.53
848.14 18.471 16.276 11.88 17.357 6.03 18.345 0.68 19.825 -7.33
858.14 20.360 18.046 11.36 19.189 5.75 20.249 0.54 21.815 -7.15
868.14 22.392 19.959 10.86 21.165 5.48 22.300 0.41 23.951 -6.96
878.14 24.574 22.0236 10.38 23.291 5.22 24.504 0.28 26.242 -6.79
888.14 26.913 24.247 9.91 25.576 4.97 26.870 0.16 28.694 -6.62
898.14 29.416 26.636 9.45 28.025 4.73 29.403 0.04 31.313 -6.45
908.14 32.0883 29.198 9.01 30.646 4.50 32.112 -0.07 34.106 -6.29
918.14 34.939 31.943 8.58 33.447 4.27 35.003 -0.18 37.081 -6.13
928.14 37.974 34.876 8.16 36.435 4.05 38.085 -0.29 40.244 -5.98
938.14 41.200 38.007 7.75 39.617 3.84 41.364 -0.40 43.602 -5.83
948.14 44.625 41.344 7.35 43.001 3.64 44.848 -0.50 47.163 -5.69
958.14 48.257 44.893 6.97 46.594 3.45 48.546 -0.60 50.933 -5.55
968.14 52.101 48.664 6.60 50.404 3.26 52.463 -0.70 54.920 -5.41
978.14 56.166 52.664 6.23 54.439 3.07 56.609 -0.79 59.130 -5.28
988.14 60.459 56.902 5.88 58.706 2.90 60.990 -0.88 63.571 -5.15
998.14 64.986 61.386 5.54 63.212 2.73 65.615 -0.97 68.250 -5.02
106
1008.14 69.756 66.123 5.21 67.966 2.57 70.492 -1.05 73.174 -4.90
1018.14 74.776 71.123 4.89 72.975 2.41 75.627 -1.14 78.350 -4.78
1028.14 80.0533 76.3923 4.57 78.246 2.26 81.0288 -1.22 83.786 -4.66
1038.14 85.595 81.942 4.27 83.788 2.11 86.705 -1.30 89.488 -4.55
1048.14 91.408 87.777 3.97 89.607 1.97 92.664 -1.37 95.464 -4.44
1058.14 97.501 93.908 3.69 95.712 1.83 98.913 -1.45 101.721 -4.33
1068.14 103.881 100.342 3.41 102.110 1.70 105.460 -1.52 108.266 -4.22
1078.14 110.555 107.087 3.14 108.808 1.58 112.313 -1.59 115.107 -4.12
1088.14 117.530 114.152 2.87 115.816 1.46 119.480 -1.66 122.249 -4.02
1098.14 124.815 121.545 2.62 123.138 1.34 126.967 -1.72 129.702 -3.91
1108.14 132.417 129.275 2.37 130.784 1.23 134.784 -1.79 137.470 -3.82
1118.14 140.342 137.348 2.13 138.760 1.13 142.938 -1.85 145.563 -3.72
1128.14 148.600 145.774 1.90 147.074 1.03 151.437 -1.91 153.985 -3.62
1138.14 157.196 154.560 1.68 155.734 0.93 160.288 -1.97 162.746 -3.53
1148.14 166.139 163.714 1.46 164.747 0.84 169.499 -2.02 171.852 -3.44
1158.14 175.437 173.245 1.25 174.119 0.75 179.0785 -2.08 181.309 -3.35
1168.14 185.0960 183.161 1.05 183.860 0.67 189.0334 -2.13 191.125 -3.26
1178.14 195.125 193.468 0.85 193.975 0.59 199.372 -2.18 201.306 -3.17
1188.14 205.531 204.177 0.66 204.471 0.52 210.101 -2.22 211.860 -3.08
1198.14 216.321 215.293 0.48 215.357 0.45 221.229 -2.27 222.794 -2.99
1208.14 227.505 226.825 0.30 226.639 0.38 232.764 -2.31 234.114 -2.91
1218.14 239.0879 238.782 0.13 238.324 0.32 244.712 -2.35 245.827 -2.82
1228.14 251.0791 251.169 -0.04 250.420 0.26 257.0821 -2.39 257.940 -2.73
1238.14 263.486 263.996 -0.19 262.932 0.21 269.881 -2.43 270.460 -2.65
1248.14 276.317 277.270 -0.35 275.869 0.16 283.117 -2.46 283.394 -2.56
1258.14 289.579 290.999 -0.49 289.237 0.12 296.798 -2.49 296.748 -2.48
1268.14 303.280 305.189 -0.63 303.042 0.08 310.930 -2.52 310.529 -2.39
1278.14 317.429 319.849 -0.76 317.292 0.04 325.522 -2.55 324.744 -2.30
1288.14 332.033 334.986 -0.89 331.993 0.01 340.580 -2.57 339.399 -2.22
1298.14 347.101 350.607 -1.01 347.152 -0.01 356.113 -2.60 354.502 -2.13
1308.14 362.641 366.720 -1.12 362.776 -0.04 372.128 -2.62 370.058 -2.05
1318.14 378.660 383.332 -1.23 378.871 -0.06 388.633 -2.63 386.074 -1.96
1328.14 395.168 400.450 -1.34 395.443 -0.07 405.634 -2.65 402.558 -1.87
1338.14 412.172 418.081 -1.43 412.499 -0.08 423.139 -2.66 419.515 -1.78
1348.14 429.681 436.234 -1.52 430.046 -0.08 441.156 -2.67 436.952 -1.69
1358.14 447.704 454.914 -1.61 448.089 -0.09 459.692 -2.68 454.876 -1.60
1368.14 466.250 474.129 -1.69 466.636 -0.08 478.754 -2.68 473.292 -1.51
1378.14 485.326 493.885 -1.76 485.691 -0.08 498.350 -2.68 492.209 -1.42
1388.14 504.943 514.191 -1.83 505.263 -0.06 518.487 -2.68 511.631 -1.32
1398.14 525.108 535.052 -1.89 525.356 -0.05 539.173 -2.68 531.566 -1.23
1408.14 545.831 556.475 -1.95 545.976 -0.03 560.414 -2.67 552.019 -1.13
1418.14 567.122 578.468 -2.00 567.131 0.00 582.219 -2.66 572.998 -1.04
1428.14 588.989 601.037 -2.05 588.825 0.03 604.594 -2.65 594.509 -0.94
107
1438.14 611.441 624.189 -2.08 611.065 0.06 627.548 -2.63 616.558 -0.84
1448.14 634.489 647.930 -2.12 633.857 0.10 651.086 -2.62 639.152 -0.73
1458.14 658.143 672.267 -2.15 657.206 0.14 675.217 -2.59 662.296 -0.63
1468.14 682.411 697.207 -2.17 681.119 0.19 699.948 -2.57 685.998 -0.53
1478.14 707.305 722.755 -2.18 705.601 0.24 725.286 -2.54 710.264 -0.42
1488.14 732.833 748.920 -2.20 730.657 0.30 751.239 -2.51 735.099 -0.31
1498.14 759.007 775.706 -2.20 756.294 0.36 777.813 -2.48 760.511 -0.20
1508.14 785.837 803.121 -2.20 782.516 0.42 805.017 -2.44 786.505 -0.09
1518.14 813.334 831.170 -2.19 809.331 0.49 832.858 -2.40 813.089 0.03
1528.14 841.508 859.861 -2.18 836.742 0.57 861.342 -2.36 840.267 0.15
1538.14 870.370 889.198 -2.16 864.756 0.64 890.477 -2.31 868.048 0.27
1548.14 899.932 919.189 -2.14 893.379 0.73 920.272 -2.26 896.436 0.39
1558.14 930.206 949.840 -2.11 922.615 0.82 950.732 -2.21 925.439 0.51
1568.14 961.201 981.157 -2.08 952.469 0.91 981.866 -2.15 955.063 0.64
1578.14 992.932 1013.145 -2.04 982.947 1.01 1013.681 -2.09 985.313 0.77
1588.14 1025.408 1045.811 -1.99 1014.055 1.11 1046.184 -2.03 1016.197 0.90
1598.14 1058.644 1079.161 -1.94 1045.798 1.21 1079.383 -1.96 1047.721 1.03
1608.14 1092.651 1113.201 -1.88 1078.180 1.32 1113.285 -1.89 1079.890 1.17
1618.14 1127.441 1147.937 -1.82 1111.207 1.44 1147.898 -1.81 1112.712 1.31
1628.14 1163.029 1183.374 -1.75 1144.885 1.56 1183.229 -1.74 1146.193 1.45
1638.14 1199.4262 1219.519 -1.68 1179.217 1.68 1219.287 -1.66 1180.340 1.59
1648.14 1236.647 1256.377 -1.60 1214.210 1.81 1256.077 -1.57 1215.157 1.74
1658.14 1274.701 1293.954 -1.51 1249.868 1.95 1293.609 -1.48 1250.653 1.89
1668.14 1313.615 1332.256 -1.42 1286.196 2.09 1331.890 -1.39 1286.833 2.04
1678.14 1353.390 1371.288 -1.32 1323.199 2.23 1370.926 -1.30 1323.705 2.19
1688.14 1394.0452 1411.057 -1.22 1360.883 2.38 1410.727 -1.20 1361.274 2.35
1698.14 1435.595 1451.568 -1.11 1399.251 2.53 1451.300 -1.09 1399.546 2.51
1708.14 1478.055 - - 1438.308 2.69 - - 1438.529 2.67
1718.14 1521.440 - - 1478.061 2.85 - - 1478.229 2.84
1728.14 1565.767 - - - - - - - -
1734.14 1592.822 - - - - - - -
108
Appendix B
Table B1: Vapor pressures calculated using PR-MC from Neqsim tool and their
deviations from the vapor pressures of DIPPRs equation
)
(eq. B1)
As experimental values are considered the Ps calculated from DIPPRs equation
DIPPRs eq. PR-MC

T [K] Ps [bar] s
P [bar] Ps %
238.15 4.90E-09 4.86E-09 0.86
248.15 1.73E-08 1.72E-08 0.72
258.15 5.52E-08 5.49E-08 0.62
268.15 1.62E-07 1.61E-07 0.53
278.15 4.37E-07 4.35E-07 0.47
288.15 1.10E-06 1.10E-06 0.41
298.15 2.61E-06 2.60E-06 0.37
308.15 5.82E-06 5.81E-06 0.33
318.15 1.24E-05 1.23E-05 0.30
328.15 2.51E-05 2.50E-05 0.28
338.15 4.86E-05 4.85E-05 0.25
348.15 9.08E-05 9.06E-05 0.23
358.15 1.64E-04 1.63E-04 0.21
368.15 2.85E-04 2.85E-04 0.19
378.15 4.83E-04 4.82E-04 0.17
388.15 7.95E-04 7.94E-04 0.15
398.15 1.28E-03 0.00127 0.14
408.15 2.00E-03 0.00199 0.12
418.15 3.06E-03 0.00306 0.10
428.15 4.60E-03 0.00459 0.09
438.15 6.78E-03 0.00677 0.07
448.15 0.0098 0.0098 0.06
458.15 0.014 0.014 0.04
468.15 0.019 0.019 0.03
478.15 0.027 0.027 0.02
488.15 0.036 0.036 0.01
498.15 0.0497 0.0497 0.00
508.15 0.066 0.066 -0.01
518.15 0.087 0.087 -0.02
528.15 0.113 0.113 -0.02
538.15 0.146 0.146 -0.03
548.15 0.186 0.186 -0.03
558.15 0.235 0.235 -0.03
568.15 0.295 0.295 -0.03
578.15 0.366 0.366 -0.03
109
Continuation of table B1
588.15 0.452 0.452 -0.03
598.15 0.555 0.555 -0.02
608.15 0.675 0.675 -0.02
618.15 0.816 0.816 -0.01
628.15 0.981 0.981 0.00
638.15 1.173 1.173 0.01
648.15 1.393 1.393 0.02
658.15 1.647 1.646 0.03
668.15 1.936 1.936 0.04
678.15 2.266 2.265 0.05
688.15 2.639 2.638 0.07
698.15 3.061 3.058 0.08
708.15 3.534 3.531 0.09
718.15 4.064 4.060 0.11
728.15 4.655 4.650 0.12
738.15 5.313 5.306 0.13
748.15 6.041 6.033 0.14
758.15 6.847 6.836 0.16
768.15 7.734 7.721 0.17
778.15 8.709 8.693 0.18
788.15 9.776 9.758 0.19
798.15 10.942 10.921 0.20
808.15 12.214 12.188 0.21
818.15 13.595 13.567 0.21
828.15 15.094 15.062 0.22
838.15 16.717 16.680 0.22
848.15 18.470 18.428 0.23
858.15 20.359 20.313 0.23
868.15 22.392 22.340 0.23
878.15 24.574 24.517 0.23
888.15 26.913 26.851 0.23
898.15 29.415 29.348 0.23
908.15 32.088 32.016 0.22
918.15 34.938 34.862 0.22
928.15 37.973 37.893 0.21
938.15 41.199 41.116 0.20
948.15 44.625 44.539 0.19
958.15 48.256 48.168 0.18
968.15 52.101 52.013 0.17
978.15 56.165 56.078 0.16
988.15 60.458 60.374 0.14
998.15 64.986 64.906 0.12
1008.15 69.756 69.683 0.11
1018.15 74.776 74.712 0.09
1028.15 80.053 80.001 0.07
110
1038.15 85.595 85.557 0.04
1048.15 91.408 91.389 0.02
1058.15 97.501 97.504 0.00
1068.15 103.880 103.910 -0.03
1078.15 110.554 110.614 -0.05
1088.15 117.530 117.625 -0.08
1098.15 124.815 124.950 -0.11
1108.15 132.416 132.598 -0.14
1118.15 140.342 140.575 -0.17
1128.15 148.599 148.891 -0.20
1138.15 157.195 157.552 -0.23
1148.15 166.139 166.567 -0.26
1158.15 175.436 175.944 -0.29
1168.15 185.095 185.690 -0.32
1178.15 195.124 195.814 -0.35
1188.15 205.530 206.324 -0.39
1198.15 216.321 217.228 -0.42
1208.15 227.504 228.533 -0.45
1218.15 239.087 240.247 -0.48
1228.15 251.079 252.379 -0.52
1238.15 263.486 264.938 -0.55
1248.15 276.316 277.929 -0.58
1258.15 289.578 291.363 -0.62
1268.15 303.280 305.247 -0.65
1278.15 317.429 319.588 -0.68
1288.15 332.033 334.396 -0.71
1298.15 347.101 349.679 -0.74
1308.15 362.640 365.444 -0.77
1318.15 378.659 381.699 -0.80
1328.15 395.167 398.454 -0.83
1338.15 412.171 415.716 -0.86
1348.15 429.681 433.493 -0.89
1358.15 447.704 451.795 -0.91
1368.15 466.249 470.628 -0.94
1378.15 485.326 490.002 -0.96
1388.15 504.942 509.925 -0.99
1398.15 525.107 530.405 -1.01
1408.15 545.831 551.451 -1.03
1418.15 567.121 573.071 -1.05
1428.15 588.988 595.274 -1.07
1438.15 611.441 618.068 -1.08
1448.15 634.489 641.463 -1.10
1458.15 658.143 665.466 -1.11
1468.15 682.411 690.087 -1.12
1478.15 707.305 715.334 -1.14
111
1488.15 732.833 741.216 -1.14
1498.15 759.007 767.742 -1.15
1508.15 785.837 794.921 -1.16
1518.15 813.334 822.761 -1.16
1528.15 841.508 851.281 -1.16
1538.15 870.370 880.472 -1.16
1548.15 899.932 910.352 -1.16
1558.15 930.206 940.931 -1.15
1568.15 961.201 972.218 -1.15
1578.15 992.932 1004.221 -1.14
1588.15 1025.408 1036.952 -1.13
1598.15 1058.644 1070.418 -1.11
1608.15 1092.651 1104.630 -1.10
1618.15 1127.441 1139.598 -1.08
1628.15 1163.029 1175.331 -1.06
1638.15 1199.426 1211.839 -1.03
1648.15 1236.647 1249.132 -1.01
1658.15 1274.706 1287.220 -0.98
1668.15 1313.615 1326.114 -0.95
1678.15 1353.390 1365.824 -0.92
1688.15 1394.045 1406.360 -0.88
1698.15 1435.595 1447.732 -0.85
1708.15 1478.055 - -
1718.15 1478.055 - -
Table B2: Vapor pressures calculated using SRK-Twu from Neqsim tool and their
deviations from the vapor pressures of DIPPRs equation
)
(eq. B2)
As experimental values are considered the Ps calculated from DIPPRs equation
DIPPRs eq. SRK-Twu

T [K] Ps [bar] s
P [bar] Ps %
238.15 4.90E-09 4.71E-09 3.82%
248.15 1.73E-08 1.67E-08 3.11%
258.15 5.52E-08 5.38E-08 2.53%
268.15 1.62E-07 1.58E-07 2.05%
278.15 4.37E-07 4.30E-07 1.65%
288.15 1.10E-06 1.09E-06 1.32%
298.15 2.61E-06 2.58E-06 1.05%
308.15 5.82E-06 5.78E-06 0.82%
318.15 1.24E-05 1.23E-05 0.63%
328.15 2.51E-05 2.49E-05 0.48%
112
338.15 4.86E-05 4.85E-05 0.35%
348.15 9.08E-05 9.06E-05 0.25%
358.15 1.64E-04 1.63E-04 0.16%
368.15 2.85E-04 2.85E-04 0.09%
378.15 4.83E-04 4.83E-04 0.03%
388.15 7.95E-04 7.95E-04 0.02%
398.15 1.28E-03 1.28E-03 0.06%
408.15 2.00E-03 2.00E-03 0.09%
418.15 3.06E-03 3.07E-03 0.11%
428.15 4.60E-03 4.61E-03 0.13%
438.15 6.78E-03 6.79E-03 0.15%
448.15 0.0098 9.83E-03 0.16%
458.15 0.014 1.40E-02 0.17%
468.15 0.019 1.96E-02 0.17%
478.15 0.027 2.71E-02 0.17%
488.15 0.036 3.70E-02 0.17%
498.15 0.0497 4.97E-02 0.17%
508.15 0.066 6.62E-02 0.16%
518.15 0.087 8.70E-02 0.16%
528.15 0.113 0.113 0.15%
538.15 0.146 0.146 0.14%
548.15 0.186 0.186 0.13%
558.15 0.235 0.235 0.11%
568.15 0.295 0.295 0.10%
578.15 0.366 0.367 0.08%
588.15 0.452 0.453 0.07%
598.15 0.555 0.555 0.05%
608.15 0.675 0.675 0.04%
618.15 0.816 0.817 0.02%
628.15 0.981 0.981 0.00%
638.15 1.173 1.172 0.02%
648.15 1.393 1.393 0.04%
658.15 1.647 1.646 0.05%
668.15 1.936 1.935 0.07%
678.15 2.266 2.264 0.09%
688.15 2.639 2.637 0.11%
698.15 3.061 3.057 0.13%
708.15 3.534 3.529 0.14%
718.15 4.064 4.058 0.16%
728.15 4.655 4.647 0.18%
738.15 5.313 5.303 0.19%
748.15 6.041 6.030 0.20%
758.15 6.847 6.832 0.22%
768.15 7.734 7.717 0.23%
778.15 8.709 8.688 0.24%
113
788.15 9.776 9.752 0.25%
798.15 10.942 10.914 0.26%
808.15 12.214 12.182 0.26%
818.15 13.595 13.559 0.27%
828.15 15.094 15.054 0.27%
838.15 16.717 16.672 0.27%
848.15 18.470 18.420 0.27%
858.15 20.359 20.305 0.27%
868.15 22.392 22.332 0.27%
878.15 24.574 24.510 0.26%
888.15 26.913 26.845 0.25%
898.15 29.415 29.343 0.25%
908.15 32.088 32.013 0.24%
918.15 34.938 34.860 0.22%
928.15 37.973 37.893 0.21%
938.15 41.199 41.119 0.20%
948.15 44.625 44.545 0.18%
958.15 48.256 48.179 0.16%
968.15 52.101 52.027 0.14%
978.15 56.165 56.098 0.12%
988.15 60.458 60.400 0.10%
998.15 64.986 64.939 0.07%
1008.15 69.756 69.723 0.05%
1018.15 74.776 74.761 0.02%
1028.15 80.053 80.059 0.01%
1038.15 85.595 85.627 0.04%
1048.15 91.408 91.470 0.07%
1058.15 97.501 97.598 0.10%
1068.15 103.880 104.018 0.13%
1078.15 110.554 110.738 0.17%
1088.15 117.530 117.766 0.20%
1098.15 124.815 125.110 0.24%
1108.15 132.416 132.777 0.27%
1118.15 140.342 140.776 0.31%
1128.15 148.599 149.114 0.35%
1138.15 157.195 157.800 0.38%
1148.15 166.139 166.841 0.42%
1158.15 175.436 176.246 0.46%
1168.15 185.095 186.022 0.50%
1178.15 195.124 196.177 0.54%
1188.15 205.530 206.720 0.58%
1198.15 216.321 217.658 0.62%
1208.15 227.504 229.000 0.66%
1218.15 239.087 240.753 0.70%
1228.15 251.079 252.925 0.74%
114
1238.15 263.486 265.525 0.77%
1248.15 276.316 278.561 0.81%
1258.15 289.578 292.040 0.85%
1268.15 303.280 305.970 0.89%
1278.15 317.429 320.361 0.92%
1288.15 332.033 335.219 0.96%
1298.15 347.101 350.553 0.99%
1308.15 362.640 366.370 1.03%
1318.15 378.659 382.680 1.06%
1328.15 395.167 399.490 1.09%
1338.15 412.171 416.808 1.12%
1348.15 429.681 434.642 1.15%
1358.15 447.704 453.001 1.18%
1368.15 466.249 471.893 1.21%
1378.15 485.326 491.325 1.24%
1388.15 504.942 511.306 1.26%
1398.15 525.107 531.844 1.28%
1408.15 545.831 552.948 1.30%
1418.15 567.121 574.625 1.32%
1428.15 588.988 596.884 1.34%
1438.15 611.441 619.732 1.36%
1448.15 634.489 643.180 1.37%
1458.15 658.143 667.233 1.38%
1468.15 682.411 691.902 1.39%
1478.15 707.305 717.193 1.40%
1488.15 732.833 743.117 1.40%
1498.15 759.007 769.680 1.41%
1508.15 785.837 796.891 1.41%
1518.15 813.334 824.767 1.41%
1528.15 841.508 853.301 1.40%
1538.15 870.370 882.508 1.39%
1548.15 899.932 912.397 1.39%
1558.15 930.206 942.976 1.37%
1568.15 961.201 974.255 1.36%
1578.15 992.932 1006.242 1.34%
1588.15 1025.408 1038.945 1.32%
1598.15 1058.644 1072.373 1.30%
1608.15 1092.651 1106.535 1.27%
1618.15 1127.441 1141.440 1.24%
1628.15 1163.029 1177.096 1.21%
1638.15 1199.426 1213.512 1.17%
1648.15 1236.647 1250.697 1.14%
1658.15 1274.706 1288.659 1.09%
1668.15 1313.615 1327.409 1.05%
1678.15 1353.390 1366.954 1.00%
115
1688.15 1394.045 1407.304 0.95%
1698.15 1435.595 1448.467 0.90%
1708.15 1478.055 - -
1718.15 1478.055 - -
116
Appendix C
Table C1: Available solubility data of Hg in HC from IUPACs book
Solvent T [K] Mole fraction of Hg (exp)

278.15 1.90E-07
283.15 2.70E-07
288.15 3.60E-07
n.C5 293.15 4.60E-07
298.15 6.80E-07
303.15 9.30E-07
308.15 1.20E-06
313.15 1.60E-06
273.15 1.80E-07
278.15 2.40E-07
283.15 3.40E-07
288.15 4.60E-07
293.15 5.90E-07
298.15 8.30E-07
n.C6 303.15 1.10E-06
308.15 1.42E-06
313.15 1.90E-06
318.15 2.38E-06
323.15 3.03E-06
328.15 3.84E-06
333.15 4.83E-06
338.15 6.03E-06
273.15 2.00E-07
278.15 2.90E-07
283.15 3.70E-07
288.15 5.40E-07
n.C7 293.15 7.00E-07
298.15 9.70E-07
303.15 1.26E-06
308.15 1.63E-06
313.15 2.20E-06
273.15 2.50E-07
278.15 3.30E-07
283.15 4.70E-07
n.C8 288.15 6.10E-07
293.15 7.90E-07
298.15 1.10E-06
303.15 1.40E-06
117
Continuation of table C1
n.C8 308.15 1.78E-06
313.15 2.30E-06
273.15 4.00E-07
278.15 4.90E-07
283.15 5.80E-07
288.15 8.70E-07
n.C10 293.15 9.60E-07
298.15 1.37E-06
303.15 1.69E-06
308.15 1.90E-06
313.15 2.52E-06
318.15 2.66E-06
273.15 1.70E-07
288.15 3.70E-07
2.2-dm-C4 293.15 4.90E-07
298.15 6.70E-07
303.15 8.10E-07
308.15 9.90E-07
273.15 1.60E-07
286.65 3.40E-07
288.15 3.60E-07
2.2.4-tm-C5 293.15 5.40E-07
298.15 7.00E-07
303.15 8.90E-07
308.15 1.11E-06
288.15 8.20E-07
293.15 1.02E-06
298.15 1.32E-06
cy-C6 303.15 1.62E-06
308.15 1.92E-06
313.15 2.47E-06
273.15 2.70E-07
288.15 6.60E-07
m-cy-C6 293.15 8.80E-07
298.15 1.17E-06
303.15 1.46E-06
308.15 1.88E-06
118
Continuation of table C1
289.15 8.10E-07
293.15 1.06E-06
cis-1.2-dm-cy-C6 298.15 1.36E-06
303.15 1.71E-06
308.15 2.06E-06
288.15 6.90E-07
293.15 9.10E-07
trans-1.2-dm-cy-C6 298.15 1.23E-06
303.15 1.55E-06
308.15 1.85E-06
288.15 7.40E-07
293.15 1.02E-06
cis-1.4-dm-cy-C6 298.15 1.33E-06
303.15 1.64E-06
308.15 2.11E-06
288.15 6.90E-07
293.15 9.10E-07
trans-1.4-dm-cy-C6 298.15 1.22E-06
303.15 1.55E-06
308.15 1.86E-06
288.15 5.80E-07
293.15 7.90E-07
295.65 9.27E-07
benzene 298.15 1.06E-06
303.15 1.37E-06
308.15 1.91E-06
273.15 3.20E-07
288.15 6.90E-07
293.15 1.03E-06
toluene 298.15 1.28E-06
303.15 1.73E-06
308.15 2.14E-06
273.15 3.10E-07
293.15 1.12E-06
o-xylene 298.15 1.45E-06
303.15 1.89E-06
308.15 2.62E-06
119
Table C2: Available solubility data of Hg in n.propane and n.butane from IUPACs
book that were not used in the master thesis
T[K] P[atm] Mole fraction of Hg (exp)

457.15 9.8 0.00140
457.15 18.9 0.000747
491.15 10.6 0.00384
C3 491.15 20.6 0.00204
491.15 29.4 0.00143
529.15 11.6 0.0100
529.15 22.5 0.00527
529.15 32.5 0.00369
457.15 9.6 0.00143

457.15 17.8 0.000778
457.15 24.1 0.000571
491.15 10.5 0.00391
n.C4 491.15 19.6 0.00211
491.15 27.2 0.00153
529.15 11.5 0.0103
529.15 21.9 0.00549
529.15 30.6 0.00390
Table C3: Raw data for n.C8, cy-C6, benzene and toluene
HC T[K] mole/L (raw data) Mole fraction of Hg (exp)

n.C6 298.15 6.4 8.41E-07
313.15 13.5 1.81E-06

336.15 50.8 7.06E-06
Robert R.
Kuntz and cy-C6 298.15 11 1.20E-06
Gilbert J.
Mains toluene 298.15 12.5 1.33E-06
benzene 298.15 12 1.074E-06
Table C4: Raw data for n.C8 and toluene
HC T[K] ppm (1000g/kg) (raw data) Mole fraction of Hg (exp)

M. M. 383.15 54.2 3.09E-05
MIEDANER, A. A.
n.C8 423.15 184 1.10E-04
MIGDISOV, and
A. E. WILLIAMS- 473.15 821 4.70E-04
JONES
120
M. M. Continuation of table C4
MIEDANER, A. A.
393.15 14 6.43E-06
MIGDISOV, and
A. E. WILLIAMS- toluene 423.15 75 3.45E-05
JONES 443.15 111 5.10E-05
473.15 280 1.30E-04
Table C5: Available solubility data of Hg in various HCs, CO2 and N2
Confidential data
Table C6: Available solubility data of Hg in iso-butane
Confidential data
Table C7: Available solubility data of Hg in H2O from IUPACs book
T [K] P [bar] Mole fraction of Hg (exp)

273.15 1 3.73E-09
278.15 1 3.91E-09
283.15 1 4.15E-09
288.15 1 4.46E-09
293.15 1 4.83E-09
298.15 1 5.28E-09
303.15 1 5.83E-09
308.15 1 6.48E-09
313.15 1 7.27E-09
318.15 1 8.20E-09
323.15 1 9.31E-09
328.15 1 1.06E-08
333.15 1 1.22E-08
338.15 1 1.41E-08
343.15 1 1.64E-08
348.15 1 1.91E-08
353.15 1 2.24E-08
358.15 1 2.63E-08
363.15 1 3.10E-08
373.15 1 4.36E-08
383.15 1 6.21E-08
393.15 1 8.91E-08
121
7.00E-06
6.00E-06
5.00E-06
mole fraction of Hg
4.00E-06
3.00E-06
n.C6
2.00E-06 cy-c6
1.00E-06
0.00E+00
270 280 290 300 310 320 330 340
T[K]
Figure C1: Solubility of Hg in cyclo and n.alkanes with carbon number 6
2.50E-06
2.00E-06
mole fraction of Hg
1.50E-06
1.00E-06 n.C7
m-cy-c6
5.00E-07
0.00E+00
270 275 280 285 290 295 300 305 310 315
T[K]
122
2.50E-06 n.C8
cis-1,2-dm-cy-c6
2.00E-06 trans-1,2-dm-cy-c6
cis-1,4-dm-cy-c6
mole fraction of Hg
1.50E-06 trans-1,4-dm-cy-c6
1.00E-06
5.00E-07
0.00E+00
270 275 280 285 290 295 300 305 310 315
T[K]
2.50E-06
2.00E-06
mole fraction of Hg
1.50E-06
1.00E-06
n.C6
benzene
5.00E-07
0.00E+00
270 275 280 285 290 295 300 305 310 315
T[K]
Figure C4: Solubility of Hg in aromatic HC and n.alkanes with carbon number 6
123
2.50E-06
2.00E-06
mole fraction of Hg
1.50E-06
1.00E-06
n.C7
5.00E-07 toluene
0.00E+00
270 275 280 285 290 295 300 305 310 315
T[K]
3.00E-06
2.50E-06
mole fraction of Hg
2.00E-06
1.50E-06
n.C8
1.00E-06
o-xylene
5.00E-07
0.00E+00
270 275 280 285 290 295 300 305 310 315
T[K]
124
7.00E-06
6.00E-06
5.00E-06
mole fraction of Hg
4.00E-06
3.00E-06 n.C6
2,2-dm-C4
2.00E-06
1.00E-06
0.00E+00
270 280 290 300 310 320 330 340
T[K]
Figure C7: Solubility of Hg in branched and n.alkanes with carbon number 6
2.50E-06
2.00E-06
mole fraction of Hg
1.50E-06
1.00E-06
n.C8
2,2,4-tm-c5
5.00E-07
0.00E+00
270 275 280 285 290 295 300 305 310 315
T[K]
Figure C8: Solubility of Hg in branched and n.alkanes with carbon number 8
125
2,2-dm-c4
2.5E-06
2,2,4-tm-c5
cy-c6
2.0E-06
cis-1,2-dm-cy-c6
trans-1,2-dm-cy-c6
mole fraction of Hg
1.5E-06 cis-1,4-dm-cy-c6
trans-1,4-dm-cy-c6
1.0E-06
5.0E-07
0.0E+00
270 275 280 285 290 295 300 305 310 315
T[K]
Figure C9: Solubility of Hg in branched and cyclo-alkanes
n.C5
n.C6
2.5E-06 n.C7
n.C8
n.C10
2.0E-06 2,2-dm-c4
2,2,4-tm-c5
cy-c6
mole fraction of Hg
m-cy-c6
1.5E-06 cis-1,2-dm-cy-c6
trans-1,2-dm-cy-c6
cis-1,4-dm-cy-c6
trans-1,4-dm-cy-c6
1.0E-06
benzene
toluene
o-xylene
5.0E-07
0.0E+00
270 275 280 285 290 295 300 305 310 315
T[K]
Figure C10: Solubility of Hg in all HC
126
Appendix D
Table D1: Twu-Coon parameters for all components
NAME L M N
Hg 0.09245 0.9784 2.244
CH4 0.5144 0.9903 1
C2H6 0.2424 0.88 1.9845
C3 0.2349 0.8662 2.2076
i-C4 0.2179 0.8489 2.2840
n.C5 0.5316 0.87 1.481
n.C6 0.1581 0.8728 3.8442
n.C7 0.1648 0.8697 4.0451
n.C8 0.23 0.8433 3.243
n.C10 0.2028 0.8418 3.8472
benzene 0.103 0.8886 4.2837
toluene 0.3319 0.8432 2.0677
o-xylene 0.2561 0.8532 2.762
cy-C6 0.0796 0.9161 6.0039
m-cy-C6 0.1416 0.8755 3.6415
cis-1.2-dm-cy-C6 0.0853 0.8909 4.9261
cis-1.4-dm-cy-C6 0.0691 0.9052 5.9034
trans-1.2-dm-cy-C6 0.106 0.8821 4.3185
trans-1.4-dm-cy-C6 0.0977 0.8868 4.583
2.2-dim-C4 0.3067 0.8562 2.1437
2.2.4-tm-C5 0.2828 0.8453 2.4878
CO2 1.4136 -0.6060 1.1018
N2 0.1523 0.8945 2.3404
water 0.3533 0.8741 2.4996
Table D2: Mathias-Copeman parameters for all components
NAME MC1 MC2 MC3

Hg 0.1491 -0.1652 0.1447
CH4 0.3923 0 0
C2H6 0.5201 0.0043 0.1029
C3 0.6184 -0.1607 0.4879
i-C4 0.6562 -0.1387 0.5039
n.C5 0.7506 -0.1073 0.3871
n.C6 0.8347 -0.3180 1.0209
n.C7 0.8648 0.0716 0.0669
n.C8 0.9633 -0.2563 0.8864
n.C10 1.0682 -0.0724 0.4364
benzene 0.6831 -0.1052 0.6671
127
Continuation of table D2
toluene 0.7646 -0.0974 0.4257
o-xylene 0.8445 -0.2051 0.6115
cy-C6 0.6805 -0.0690 0.5765
m-cy-C6 0.7162 -0.0204 0.3396
cis-1.2-dm-cy-C6 0.6288 0.7011 -0.7844
cis-1.4-dm-cy-C6 0 0 0
2.2-dim-C4 0.6989 0.1119 -0.0635
2.2.4-tm-C5 0 0 0
CO2 0.7137 -0.4476 2.4375
N2 0.4363 0 0
water 0.9237 -0.3794 0.4424
Table D3: Unifac group parameters of all components included in the mixtures used
for the predictions of the mole fraction of Hg
CH4 C2H6 96.49 0.5073 0 -71.15 -0.4012 0
CH4 CO2 -83.73 -2.2570 0 301.98 1.6951 0
N2 CH4 -119.61 -0.7538 0 194.04 0.8318 0
CO2 C2H6 90.86 -0.4893 0 117.63 -0.3263 0
N2 C2H6 -154.52 -0.9021 0 303.92 0.9467 0
N2 CO2 -137.90 -1.6936 0 382.68 1.5736 0
CH4 CH2 328.80 1.8031 -0.011159 -214.41 -1.2202 0.004627
C2H6 CH2 154.49 -0.2815 0.011039 -140.19 -0.1858 -0.005389
CO2 CH2 70.06 -0.7276 0.002452 84.68 -0.6989 0.002966
N2 CH2 112.52 -1.4903 0.000607 -4.26 0.6264 0.002043
128
Appendix E
9.00E-08
8.00E-08
7.00E-08
mole fraction of Hg
6.00E-08
5.00E-08
4.00E-08 exp
3.00E-08 SRK-Twu
2.00E-08
1.00E-08
0.00E+00
240 250 260 270 280 290 300
T[K]
Figure E1: Mole fraction of Hg in methane calculated with SRK-Twu EOS with kij
parameters in the vapor phase
2.00E-07
1.80E-07 exp
1.60E-07
SRK-Twu
mole fraction of Hg
1.40E-07
1.20E-07
1.00E-07
8.00E-08
6.00E-08
4.00E-08
2.00E-08
0.00E+00
270 275 280 285 290 295
T[K]
Figure E2: Mole fraction of Hg in ethane calculated with SRK-Twu EOS with kij
parameters in the liquid phase
129
3.50E-07
exp
3.00E-07
mole fraction of Hg SRK-Twu
2.50E-07
2.00E-07
1.50E-07
1.00E-07
5.00E-08
0.00E+00
270 275 280 285 290 295
T [K]
Figure E3: Mole fraction of Hg in propane calculated with SRK-Twu EOS with kij
1.80E-06
1.60E-06
exp
mole fraction of Hg
1.40E-06
1.20E-06 SRK-Twu
1.00E-06
8.00E-07
6.00E-07
4.00E-07
2.00E-07
0.00E+00
275 280 285 290 295 300 305 310 315
T[K]
Figure E4: Mole fraction of Hg in n.C5 calculated with SRK-Twu EOS with kij
7.00E-06
6.00E-06 exp
mole fraction of Hg
5.00E-06 SRK-Twu
4.00E-06
3.00E-06
2.00E-06
1.00E-06
0.00E+00
270 280 290 300 310 320 330 340
T[K]
130
2.50E-06
mole fraction of Hg 2.00E-06 exp
SRK-Twu
1.50E-06
1.00E-06
5.00E-07
0.00E+00
270 280 290 300 310
T[K]
3.00E-06
2.50E-06
exp
mole fraction of Hg
2.00E-06
SRK-Twu
1.50E-06
1.00E-06
5.00E-07
0.00E+00
270 280 290 300 310 320
T[K]
4.00E-06
3.50E-06 exp
mole fraction of Hg
3.00E-06
SRK-Twu
2.50E-06
2.00E-06
1.50E-06
1.00E-06
5.00E-07
0.00E+00
270 280 290 300 310 320 330
T[K]
131
1.40E-06
1.20E-06 exp
mole fraction of Hg
1.00E-06 SRK-Twu
8.00E-07
6.00E-07
4.00E-07
2.00E-07
0.00E+00
270 275 280 285 290 295 300 305 310
T[K]
Figure E9: Mole fraction of Hg in 2.2-dm-C4 calculated with SRK-Twu EOS with kij
1.40E-06
1.20E-06 exp
mole fraction of Hg
1.00E-06
SRK-Twu
8.00E-07
6.00E-07
4.00E-07
2.00E-07
0.00E+00
270 275 280 285 290 295 300 305 310
T[K]
Figure E10: Mole fraction of Hg in 2.2.4-tm-C5 calculated with SRK-Twu EOS with kij
3.00E-06
exp
2.50E-06
mole fraction of Hg
2.00E-06 SRK-Twu
1.50E-06
1.00E-06
5.00E-07
0.00E+00
285 290 295 300 305 310 315
T[K]
Figure E11: Mole fraction of Hg in cy-C6 calculated with SRK-Twu EOS with kij
132
2.50E-06
exp
1.50E-06 SRK-Twu
1.00E-06
5.00E-07
0.00E+00
270 275 280 285 290 295 300 305 310
T[K]
Figure E12: Mole fraction of Hg in m-cy-C6 calculated with SRK-Twu EOS with kij
2.50E-06
exp
2.00E-06
mole fraction of Hg
SRK-Twu
1.50E-06
1.00E-06
5.00E-07
0.00E+00
285 290 295 300 305 310
T[K]
Figure E13: Mole fraction of Hg in trans-1.2-dm-cy-C6 calculated with SRK-Twu EOS

with kij parameters in the liquid phase
2.50E-06
exp
2.00E-06
mole fraction of Hg
SRK-Twu
1.50E-06
1.00E-06
5.00E-07
0.00E+00
285 290 295 300 305 310
T[K]
Figure E14: Mole fraction of Hg in trans-1.4-dm-cy-C6 calculated with SRK-Twu EOS

133
2.50E-06
exp
SRK-Twu
1.50E-06
1.00E-06
5.00E-07
0.00E+00
285 290 295 300 305 310
T[K]
Figure E15: Mole fraction of Hg in cis-1.2-dm-cy-C6 calculated with SRK-Twu EOS with
kij parameters in the liquid phase
2.50E-06
exp
mole fraction of Hg
2.00E-06
SRK-Twu
1.50E-06
1.00E-06
5.00E-07
0.00E+00
285 290 295 300 305 310
T[K]
Figure E16: Mole fraction of Hg in cis-1.4-dm-cy-C6 calculated with SRK-Twu EOS with
2.50E-06
exp
mole fraction of Hg
2.00E-06
1.50E-06 SRK-Twu
1.00E-06
5.00E-07
0.00E+00
285 290 295 300 305 310
T[K]
Figure E17: Mole fraction of Hg in benzene calculated with SRK-Twu EOS with kij
134
2.50E-06
exp
SRK-Twu
1.50E-06
1.00E-06
5.00E-07
0.00E+00
270 275 280 285 290 295 300 305 310
T[K]
Figure E18: Mole fraction of Hg in toluene calculated with SRK-Twu EOS with kij
3.00E-06
2.50E-06
mole fraction of Hg
exp
2.00E-06 SRK-Twu
1.50E-06
1.00E-06
5.00E-07
0.00E+00
270 275 280 285 290 295 300 305 310
T[K]
Figure E19: Mole fraction of Hg in o-xylene calculated with SRK-Twu EOS with kij
4.50E-08
4.00E-08 exp
3.50E-08 SRK-Twu
mole fraction of Hg
3.00E-08
2.50E-08
2.00E-08
1.50E-08
1.00E-08
5.00E-09
0.00E+00
5 15 25 35 45 55 65 75
P[bar]
Figure E20: Mole fraction of Hg in N2 calculated with SRK-Twu EOS with kij
135
1.00E-07
9.00E-08
exp
8.00E-08
SRK-Twu
mole fraction of Hg
7.00E-08
6.00E-08
5.00E-08
4.00E-08
3.00E-08
2.00E-08
1.00E-08
0.00E+00
250 270 290 310 330 350 370 390
T[K]
Figure E22: Mole fraction of Hg in water calculated with SRK-Twu EOS with kij
parameters
9.00E-08
8.00E-08
7.00E-08
mole fraction of Hg
6.00E-08
5.00E-08
4.00E-08 exp
3.00E-08 PR-MC
2.00E-08
1.00E-08
0.00E+00
240 250 260 270 280 290 300
T[K]
Figure E22: Mole fraction of Hg in methane calculated with PR-MC EOS with kij
136
2.00E-07
1.80E-07 exp
mole fraction of Hg 1.60E-07 PR-MC
1.40E-07
1.20E-07
1.00E-07
8.00E-08
6.00E-08
4.00E-08
2.00E-08
0.00E+00
270 275 280 285 290 295
T[K]
Figure E23: Mole fraction of Hg in ethane calculated with PR-MC EOS with kij
3.50E-07
exp
3.00E-07
PR-MC
mole fraction of Hg
2.50E-07
2.00E-07
1.50E-07
1.00E-07
5.00E-08
0.00E+00
270 275 280 285 290 295
T [K]
Figure E24: Mole fraction of Hg in propane calculated with PR-MC EOS with kij
137
1.80E-06
1.60E-06
exp
1.40E-06
mole fraction of Hg
PR-MC
1.20E-06
1.00E-06
8.00E-07
6.00E-07
4.00E-07
2.00E-07
0.00E+00
275 280 285 290 295 300 305 310 315
T[K]
Figure E25: Mole fraction of Hg in n.C5 calculated with PR-MC EOS with kij
7.00E-06
6.00E-06 exp
mole fraction of Hg
5.00E-06 PR-MC
4.00E-06
3.00E-06
2.00E-06
1.00E-06
0.00E+00
270 280 290 300 310 320 330 340
T[K]
2.50E-06
exp
2.00E-06
mole fraction of Hg
PR-MC
1.50E-06
1.00E-06
5.00E-07
0.00E+00
270 280 290 300 310
T[K]
138
3.00E-06
exp
PR-MC
2.00E-06
1.50E-06
1.00E-06
5.00E-07
0.00E+00
270 280 290 300 310 320
T[K]
4.00E-06
3.50E-06
exp
mole fraction of Hg
3.00E-06
PR-MC
2.50E-06
2.00E-06
1.50E-06
1.00E-06
5.00E-07
0.00E+00
270 280 290 300 310 320 330
T[K]
139
1.20E-06
1.00E-06 exp
mole fraction of Hg
8.00E-07 PR-MC
6.00E-07
4.00E-07
2.00E-07
0.00E+00
270 275 280 285 290 295 300 305 310
T[K]
Figure E30: Mole fraction of Hg in 2.2-dm-C4 calculated with PR-MC EOS with kij
1.40E-06
1.20E-06 exp
mole fraction of Hg
1.00E-06 PR-MC
8.00E-07
6.00E-07
4.00E-07
2.00E-07
0.00E+00
270 275 280 285 290 295 300 305 310
T[K]
Figure E31: Mole fraction of Hg in 2.2.4-tm-C5 calculated with PR-MC EOS with kij
3.00E-06
exp
2.50E-06
mole fraction of Hg
PR-MC
2.00E-06
1.50E-06
1.00E-06
5.00E-07
0.00E+00
285 290 295 300 305 310 315
T[K]
Figure E32: Mole fraction of Hg in cy-C6 calculated with PR-MC EOS with kij
140
2.50E-06
exp
PR-MC
1.50E-06
1.00E-06
5.00E-07
0.00E+00
270 275 280 285 290 295 300 305 310
T[K]
Figure E33: Mole fraction of Hg in m-cy-C6 calculated with PR-MC EOS with kij
2.50E-06
exp
2.00E-06
mole fraction of Hg
PR-MC
1.50E-06
1.00E-06
5.00E-07
0.00E+00
285 290 295 300 305 310
T[K]
Figure E34: Mole fraction of Hg in cis-1.2-dm-cy-C6 calculated with PR-MC EOS with
2.50E-06
exp
2.00E-06
mole fraction of Hg
PR-MC
1.50E-06
1.00E-06
5.00E-07
0.00E+00
285 290 295 300 305 310
T[K]
Figure E35: Mole fraction of Hg in trans-1.2-dm-cy-C6 calculated with PR-MC EOS

141
2.50E-06
exp
PR-MC
1.50E-06
1.00E-06
5.00E-07
0.00E+00
285 290 295 300 305 310
T[K]
Figure E36: Mole fraction of Hg in trans-1.4-dm-cy-C6 calculated with PR-MC EOS

2.50E-06
exp
mole fraction of Hg
2.00E-06
PR-MC
1.50E-06
1.00E-06
5.00E-07
0.00E+00
285 290 295 300 305 310
T[K]
Figure E37: Mole fraction of Hg in cis-1.4-dm-cy-C6 calculated with PR-MC EOS with
2.50E-06
exp
2.00E-06
mole fraction of Hg
PR-MC
1.50E-06
1.00E-06
5.00E-07
0.00E+00
285 290 295 300 305 310
T[K]
Figure E38: Mole fraction of Hg in benzene calculated with PR-MC EOS with kij
142
2.50E-06
exp
PR-MC
1.50E-06
1.00E-06
5.00E-07
0.00E+00
270 275 280 285 290 295 300 305 310
T[K]
Figure E39: Mole fraction of Hg in toluene calculated with PR-MC EOS with kij
3.00E-06
2.50E-06 exp
mole fraction of Hg
2.00E-06 PR-MC
1.50E-06
1.00E-06
5.00E-07
0.00E+00
270 275 280 285 290 295 300 305 310
T[K]
Figure E40: Mole fraction of Hg in o-xylene calculated with PR-MC EOS with kij
4.50E-08
4.00E-08
3.50E-08
mole fraction of Hg
3.00E-08
exp
2.50E-08
PR-MC
2.00E-08
1.50E-08
1.00E-08
5.00E-09
0.00E+00
5 15 25 35 45 55 65 75
P[bar]
Figure E41: Mole fraction of Hg in N2 calculated with PR-MC EOS with kij parameters
in the vapor phase
143
1.00E-07
9.00E-08 exp
8.00E-08 PR-MC
mole fraction of Hg
7.00E-08
6.00E-08
5.00E-08
4.00E-08
3.00E-08
2.00E-08
1.00E-08
0.00E+00
250 270 290 310 330 350 370 390
T[K]
Figure E42: Mole fraction of Hg in water calculated with PR-MC EOS with kij
parameters
Table E1: Analytical results and absolute deviations of the mole fraction of Hg in
methane for the SRK-Twu EOS in the vapor phase
Confidential data
ethane for the SRK-Twu EOS in the liquid phase
Confidential data
propane for the SRK-Twu EOS in the liquid phase
Confidential data
Table E4: Analytical results and absolute deviations of the mole fraction of Hg in n.C 5
for the SRK-Twu EOS in the liquid phase
kij=const
T[K] xexp xcalc Dx%
278.15 1.90E-07 1.94E-07 2.00
283.15 2.70E-07 2.68E-07 0.67
288.15 3.60E-07 3.67E-07 1.98
293.15 4.60E-07 4.97E-07 8.12
298.15 6.80E-07 6.67E-07 1.89
303.15 9.30E-07 8.87E-07 4.67
308.15 1.20E-06 1.17E-06 2.68
313.15 1.60E-06 1.53E-06 4.66
144
kij=const
T[K] xexp xcalc Dx%
273.15 1.80E-07 1.72E-07 4.62
278.15 2.40E-07 2.40E-07 0.16
283.15 3.40E-07 3.31E-07 2.74
288.15 4.60E-07 4.52E-07 1.84
293.15 5.90E-07 6.10E-07 3.45
298.15 8.30E-07 8.17E-07 1.56
298.15 8.41E-07 8.17E-07 2.81
303.15 1.10E-06 1.08E-06 1.46
308.15 1.42E-06 1.43E-06 0.38
313.15 1.81E-06 1.86E-06 2.63
313.15 1.90E-06 1.86E-06 2.16
318.15 2.38E-06 2.41E-06 1.06
323.15 3.03E-06 3.09E-06 1.94
328.15 3.84E-06 3.94E-06 2.56
333.15 4.83E-06 4.99E-06 3.25
338.15 6.03E-06 6.27E-06 4.05
Table E6: Analytical results and absolute deviations of the mole fraction of Hg in n.C7
kij=const
T[K] xexp xcalc Dx%
273.15 2.00E-07 2.02E-07 1.10
278.15 2.90E-07 2.81E-07 2.95
283.15 3.70E-07 3.87E-07 4.70
288.15 5.40E-07 5.28E-07 2.28
293.15 7.00E-07 7.12E-07 1.66
298.15 9.70E-07 9.51E-07 1.99
303.15 1.26E-06 1.26E-06 0.11
308.15 1.63E-06 1.65E-06 1.34
313.15 2.20E-06 2.15E-06 2.25
kij=const
T[K] xexp xcalc Dx%
273.15 2.50E-07 2.33E-07 6.76
278.15 3.30E-07 3.24E-07 1.90
145
Continuation of table E7
283.15 4.70E-07 4.45E-07 5.39
288.15 6.10E-07 6.04E-07 0.90
293.15 7.90E-07 8.14E-07 2.99
298.15 1.10E-06 1.08E-06 1.38
303.15 1.40E-06 1.43E-06 2.39
308.15 1.78E-06 1.88E-06 5.51
313.15 2.30E-06 2.44E-06 6.12
n.C10 for the SRK-Twu EOS in the liquid phase
kij=const
T[K] xexp xcalc Dx%
273.15 4.00E-07 2.90E-07 27.38
278.15 4.90E-07 4.01E-07 18.10
283.15 5.80E-07 5.48E-07 5.43
288.15 8.70E-07 7.42E-07 14.70
293.15 9.60E-07 9.95E-07 3.60
298.15 1.37E-06 1.32E-06 3.60
303.15 1.69E-06 1.74E-06 2.88
308.15 2.04E-06 2.27E-06 11.27
313.15 2.52E-06 2.94E-06 16.68
318.15 3.09E-06 3.78E-06 22.33
Table E9: Analytical results and absolute deviations of the mole fraction of Hg in 2.2-
dm-C4 for the SRK-Twu EOS in the liquid phase
kij=const
T[K] xexp xcalc Dx%
273.15 1.70E-07 1.31E-07 23.15
288.15 3.70E-07 3.53E-07 4.70
293.15 4.90E-07 4.80E-07 1.98
298.15 6.30E-07 6.48E-07 2.81
303.15 8.10E-07 8.65E-07 6.81
308.15 9.90E-07 1.15E-06 15.68
2.2.4-tm-C5 for the SRK-Twu EOS in the liquid phase
kij=const
T[K] xexp xcalc Dx%
273.15 1.60E-07 1.38E-07 13.69
146
286.65 3.40E-07 3.40E-07 0.12
288.15 3.60E-07 3.85E-07 7.04
293.15 5.40E-07 5.09E-07 5.74
298.15 7.00E-07 6.87E-07 1.88
303.15 8.90E-07 9.18E-07 3.15
308.15 1.11E-06 1.22E-06 9.56
318.15 3.09E-06 3.78E-06 22.33
Table E11: Analytical results and absolute deviations of the mole fraction of Hg in cy-
C6 for the SRK-Twu EOS in the liquid phase
kij=const
T[K] xexp xcalc Dx%
288.15 8.20E-07 7.07E-07 13.79
293.15 1.02E-06 9.42E-07 7.67
298.15 1.32E-06 1.24E-06 5.81
298.15 1.20E-06 1.24E-06 3.83
303.15 1.62E-06 1.63E-06 0.45
308.15 1.92E-06 2.11E-06 10.02
313.15 2.47E-06 2.72E-06 10.15
Table E12: Analytical results and absolute deviations of the mole fraction of Hg in m-
cy-C6 for the SRK-Twu EOS in the liquid phase
kij=const
T[K] xexp xcalc Dx%
273.15 2.70E-07 2.54E-07 6.10
288.15 6.60E-07 6.50E-07 1.55
293.15 8.80E-07 8.71E-07 1.00
298.15 1.17E-06 1.16E-06 1.09
303.15 1.46E-06 1.52E-06 4.34
308.15 1.88E-06 1.99E-06 5.77
Table E13: Analytical results and absolute deviations of the mole fraction of Hg in cis-
1.2-dm-cy-C6 for the SRK-Twu EOS in the liquid phase
kij=const
T[K] xexp xcalc Dx%
289.15 8.10E-07 7.87E-07 2.79
293.15 1.06E-06 9.94E-07 6.26
298.15 1.36E-06 1.32E-06 2.78
303.15 1.71E-06 1.73E-06 1.38
147
308.15 2.06E-06 2.26E-06 9.76
1.4-dm-cy-C6 for the SRK-Twu EOS in the liquid phase
kij=const
T[K] xexp xcalc Dx%
288.15 7.40E-07 7.28E-07 1.62
293.15 1.02E-06 9.74E-07 4.48
298.15 1.33E-06 1.29E-06 2.56
303.15 1.64E-06 1.70E-06 3.57
308.15 2.11E-06 2.21E-06 4.96
trans-1.2-dm-cy-C6 for the SRK-Twu EOS in the liquid phase
kij=const
T[K] xexp xcalc Dx%
288.15 6.90E-07 6.60E-07 4.32
293.15 9.10E-07 8.87E-07 2.53
298.15 1.23E-06 1.18E-06 4.34
303.15 1.55E-06 1.56E-06 0.46
308.15 1.85E-06 2.04E-06 10.09
trans-1.4- dm-cy-C6 for the SRK-Twu EOS in the liquid phase
kij=const
T[K] xexp xcalc Dx%
288.15 6.90E-07 6.60E-07 4.34
293.15 9.10E-07 8.86E-07 2.58
298.15 1.22E-06 1.18E-06 3.24
303.15 1.55E-06 1.56E-06 0.36
308.15 1.86E-06 2.03E-06 9.36
148
benzene for the SRK-Twu EOS in the liquid phase
kij=const
T[K] xexp xcalc Dx%
288.15 5.80E-07 6.01E-07 3.68
293.15 7.90E-07 8.02E-07 1.50
295.65 9.27E-07 9.23E-07 0.46
298.15 1.06E-06 1.06E-06 0.05
298.15 1.07E-06 1.06E-06 1.33
303.15 1.37E-06 1.39E-06 1.30
308.15 1.91E-06 1.80E-06 5.62
toluene for the SRK-Twu EOS in the liquid phase
kij=const
T[K] xexp xcalc Dx%
273.15 3.20E-07 2.95E-07 7.70
288.15 6.90E-07 7.43E-07 7.70
293.15 1.03E-06 9.90E-07 3.84
298.15 1.28E-06 1.31E-06 2.17
298.15 1.33E-06 1.31E-06 1.82
303.15 1.73E-06 1.71E-06 1.07
308.15 2.14E-06 2.22E-06 3.79
Table E19: Analytical results and absolute deviations of the mole fraction of Hg in o-
xylene for the SRK-Twu EOS in the liquid phase
kij=const
T[K] xexp xcalc Dx%
273.15 3.10E-07 3.29E-07 6.08
293.15 1.12E-06 1.10E-06 1.95
298.15 1.45E-06 1.45E-06 0.08
303.15 1.89E-06 1.89E-06 0.26
308.15 2.62E-06 2.46E-06 6.20
Table E20: Analytical results and absolute deviations of the mole fraction of Hg in N 2
for the SRK-Twu EOS in the vapor phase
Confidential data
149
water for the SRK-Twu EOS
kij T [K] P [bar] Mole fraction of Hg (exp) Mole fraction of Hg (calc) abs dev %
0.5804 273.15 1 3.73E-09 2.71E-09 27.36
0.5931 278.15 1 3.91E-09 3.14E-09 19.72
0.6058 283.15 1 4.15E-09 3.63E-09 12.53
0.6185 288.15 1 4.46E-09 4.19E-09 6.04
0.6312 293.15 1 4.83E-09 4.83E-09 0.00
0.6439 298.15 1 5.28E-09 5.56E-09 5.27
0.6566 303.15 1 5.83E-09 6.39E-09 9.56
0.6693 308.15 1 6.48E-09 7.33E-09 13.11
0.6820 313.15 1 7.27E-09 8.40E-09 15.55
0.6947 318.15 1 8.20E-09 9.62E-09 17.26
0.7074 323.15 1 9.31E-09 1.10E-08 18.08
0.7202 328.15 1 1.06E-08 1.25E-08 18.36
0.7329 333.15 1 1.22E-08 1.43E-08 17.21
0.7456 338.15 1 1.41E-08 1.63E-08 15.60
0.7583 343.15 1 1.64E-08 1.86E-08 13.18
0.7710 348.15 1 1.91E-08 2.11E-08 10.57
0.7837 353.15 1 2.24E-08 2.40E-08 7.20
0.7964 358.15 1 2.63E-08 2.73E-08 3.73
0.8091 363.15 1 3.10E-08 3.10E-08 0.00
0.8345 373.15 1 4.36E-08 3.99E-08 8.59
0.8599 383.15 1 6.21E-08 5.12E-08 17.59
0.8853 393.15 1 8.91E-08 6.56E-08 26.36
methane for the PR-MC EOS in the vapor phase
Confidential data
ethane for the PR-MC EOS in the liquid phase
Confidential data
propane for the PR-MC EOS in the liquid phase
Confidential data
150
n.C5 for the PR-MC EOS in the liquid phase
kij=const
T[K] xexp xcalc Dx%
278.15 1.90E-07 1.95E-07 2.76
283.15 2.70E-07 2.69E-07 0.20
288.15 3.60E-07 3.68E-07 2.20
293.15 4.60E-07 4.97E-07 8.11
298.15 6.80E-07 6.66E-07 2.10
303.15 9.30E-07 8.83E-07 5.05
308.15 1.20E-06 1.16E-06 3.23
313.15 1.60E-06 1.51E-06 5.35
kij=const
T[K] xexp xcalc Dx%
273.15 1.80E-07 1.75E-07 2.68
278.15 2.40E-07 2.44E-07 1.46
283.15 3.40E-07 3.35E-07 1.53
288.15 4.60E-07 4.55E-07 0.98
293.15 5.90E-07 6.14E-07 4.00
298.15 8.30E-07 8.19E-07 1.34
298.15 8.41E-07 8.19E-07 2.59
303.15 1.10E-06 1.08E-06 1.53
308.15 1.42E-06 1.42E-06 0.03
313.15 1.81E-06 1.85E-06 2.02
313.15 1.90E-06 1.85E-06 2.74
318.15 2.38E-06 2.39E-06 0.22
323.15 3.03E-06 3.06E-06 0.87
328.15 3.84E-06 3.89E-06 1.28
333.15 4.83E-06 4.92E-06 1.77
338.15 6.03E-06 6.17E-06 2.37
kij=const
T[K] xexp xcalc Dx%
273.15 2.00E-07 2.04E-07 1.99
278.15 2.90E-07 2.83E-07 2.32
283.15 3.70E-07 3.89E-07 5.14
151
288.15 5.40E-07 5.29E-07 2.10
293.15 7.00E-07 7.11E-07 1.61
298.15 9.70E-07 9.48E-07 2.28
303.15 1.26E-06 1.25E-06 0.63
308.15 1.63E-06 1.64E-06 0.59
313.15 2.20E-06 2.13E-06 3.20
kij=const
T[K] xexp xcalc Dx%
273.15 2.50E-07 2.36E-07 5.46
278.15 3.30E-07 3.27E-07 0.91
283.15 4.70E-07 4.48E-07 4.78
288.15 6.10E-07 6.06E-07 0.61
293.15 7.90E-07 8.13E-07 2.96
298.15 1.10E-06 1.08E-06 1.72
303.15 1.40E-06 1.42E-06 1.74
308.15 1.78E-06 1.86E-06 4.55
313.15 2.30E-06 2.41E-06 4.88
kij=const
T[K] xexp xcalc Dx%
273.15 4.00E-07 2.93E-07 26.86
278.15 4.90E-07 4.04E-07 17.60
283.15 5.80E-07 5.51E-07 4.96
288.15 8.70E-07 7.45E-07 14.38
293.15 9.60E-07 9.97E-07 3.83
298.15 1.37E-06 1.32E-06 3.53
303.15 1.69E-06 1.74E-06 2.79
308.15 2.04E-06 2.26E-06 10.99
313.15 2.52E-06 2.93E-06 16.19
318.15 3.09E-06 3.76E-06 21.60
152
2.2-dm-C4 for the PR-MC EOS in the liquid phase
kij=const
T[K] xexp xcalc Dx%
273.15 1.70E-07 1.33E-07 22.03
288.15 3.70E-07 3.54E-07 4.26
293.15 4.90E-07 4.81E-07 1.81
298.15 6.30E-07 6.47E-07 2.71
303.15 8.10E-07 8.62E-07 6.44
308.15 9.90E-07 1.14E-06 15.01
2.2.4-tm-C5 for the PR-MC EOS in the liquid phase
kij=const
T[K] xexp xcalc Dx%
273.15 1.60E-07 1.38E-07 13.59
286.65 3.40E-07 3.40E-07 0.10
288.15 3.60E-07 3.85E-07 7.05
293.15 5.40E-07 5.09E-07 5.76
298.15 7.00E-07 6.87E-07 1.91
303.15 8.90E-07 9.18E-07 3.10
308.15 1.11E-06 1.22E-06 9.48
Table E32: Analytical results and absolute deviations of the mole fraction of Hg in cy-
C6 for the PR-MC EOS in the liquid phase
kij=const
T[K] xexp xcalc Dx%
288.15 8.20E-07 7.11E-07 13.31
293.15 1.02E-06 9.45E-07 7.35
298.15 1.32E-06 1.25E-06 5.68
298.15 1.20E-06 1.25E-06 3.98
303.15 1.62E-06 1.63E-06 0.39
308.15 1.92E-06 2.11E-06 9.73
313.15 2.47E-06 2.71E-06 9.64
153
Table E33: Analytical results and absolute deviations of the mole fraction of Hg in m-
cy-C6 for the PR-MC EOS in the liquid phase
kij=const
T[K] xexp xcalc Dx%
273.15 2.70E-07 2.56E-07 5.12
288.15 6.60E-07 6.52E-07 1.27
293.15 8.80E-07 8.72E-07 0.95
298.15 1.17E-06 1.16E-06 1.28
303.15 1.46E-06 1.52E-06 3.91
308.15 1.88E-06 1.98E-06 5.10
1.2-dm-cy-C6 for the PR-MC EOS in the liquid phase
kij=const
T[K] xexp xcalc Dx%
289.15 8.10E-07 7.89E-07 2.62
293.15 1.06E-06 9.95E-07 6.14
298.15 1.36E-06 1.32E-06 2.75
303.15 1.71E-06 1.73E-06 1.31
308.15 2.06E-06 2.26E-06 9.54
1.4-dm-cy-C6 for the PR-MC EOS in the liquid phase
kij=const
T[K] xexp xcalc Dx%
288.15 7.40E-07 7.24E-07 2.13
293.15 1.02E-06 9.72E-07 4.70
298.15 1.33E-06 1.29E-06 2.54
303.15 1.64E-06 1.70E-06 3.80
308.15 2.11E-06 2.22E-06 5.37
trans-1.2-dm-cy-C6 for the PR-MC EOS in the liquid phase
kij=const
T[K] xexp xcalc Dx%
288.15 6.90E-07 6.63E-07 3.84
293.15 9.10E-07 8.89E-07 2.27
298.15 1.23E-06 1.18E-06 4.30
303.15 1.55E-06 1.55E-06 0.26
154
308.15 1.85E-06 2.03E-06 9.59
trans-1.4-dm-cy-C6 for the PR-MC EOS in the liquid phase
kij=const
T[K] xexp xcalc Dx%
288.15 6.90E-07 6.59E-07 4.57
293.15 9.10E-07 8.86E-07 2.68
298.15 1.22E-06 1.18E-06 3.23
303.15 1.55E-06 1.56E-06 0.44
308.15 1.86E-06 2.04E-06 9.52
benzene for the PR-MC EOS in the liquid phase
kij=const
T[K] xexp xcalc Dx%
288.15 5.80E-07 6.04E-07 4.12
293.15 7.90E-07 8.03E-07 1.69
295.65 9.27E-07 9.23E-07 0.39
298.15 1.06E-06 1.06E-06 0.09
298.15 1.07E-06 1.06E-06 1.38
303.15 1.37E-06 1.38E-06 1.02
308.15 1.91E-06 1.79E-06 6.08
toluene for the PR-MC EOS in the liquid phase
kij=const
T[K] xexp xcalc Dx%
273.15 3.20E-07 3.00E-07 6.38
288.15 6.90E-07 7.46E-07 8.08
293.15 1.03E-06 9.91E-07 3.80
298.15 1.28E-06 1.30E-06 1.91
298.15 1.33E-06 1.30E-06 2.06
303.15 1.73E-06 1.70E-06 1.58
308.15 2.14E-06 2.20E-06 2.99
155
Table E40: Analytical results and absolute deviations of the mole fraction of Hg in o-
xylene for the PR-MC EOS in the liquid phase
kij=const
T[K] xexp xcalc Dx%
273.15 3.10E-07 3.33E-07 7.56
293.15 1.12E-06 1.10E-06 2.02
298.15 1.45E-06 1.44E-06 0.47
303.15 1.89E-06 1.88E-06 0.43
308.15 2.62E-06 2.43E-06 7.11
Table E41: Analytical results and absolute deviations of the mole fraction of Hg in N 2
for the PR-MC EOS in the vapor phase
Confidential data
water for the PR-MC EOS
kij T [K] P [bar] Mole fraction Mole fraction abs dev %

of Hg (exp) of Hg (calc)
0.5736 273.15 1 3.73E-09 2.70E-09 27.68
0.5859 278.15 1 3.91E-09 3.13E-09 20.00
0.5981 283.15 1 4.15E-09 3.63E-09 12.59
0.6104 288.15 1 4.46E-09 4.19E-09 6.06
0.6227 293.15 1 4.83E-09 4.83E-09 0.00
0.6350 298.15 1 5.28E-09 5.56E-09 5.30
0.6473 303.15 1 5.83E-09 6.39E-09 9.60
0.6595 308.15 1 6.48E-09 7.35E-09 13.36
0.6718 313.15 1 7.27E-09 8.42E-09 15.78
0.6841 318.15 1 8.20E-09 9.63E-09 17.47
0.6964 323.15 1 9.31E-09 1.10E-08 18.25
0.7086 328.15 1 1.06E-08 1.26E-08 18.76
0.7209 333.15 1 1.22E-08 1.44E-08 17.66
0.7332 338.15 1 1.41E-08 1.64E-08 15.96
0.7455 343.15 1 1.64E-08 1.86E-08 13.45
0.7578 348.15 1 1.91E-08 2.12E-08 10.75
0.7700 353.15 1 2.24E-08 2.41E-08 7.44
0.7823 358.15 1 2.63E-08 2.73E-08 3.86
0.7946 363.15 1 3.10E-08 3.10E-08 0.00
0.8192 373.15 1 4.36E-08 3.98E-08 8.77
0.8437 383.15 1 6.21E-08 5.10E-08 17.83
0.8683 393.15 1 8.91E-08 6.52E-08 26.78
156
Appendix F
9.00E-08
8.00E-08
7.00E-08
mole fraction of Hg
6.00E-08
5.00E-08
4.00E-08 exp
3.00E-08 UMR-PRMC
2.00E-08
1.00E-08
0.00E+00
240 250 260 270 280 290 300
T[K]
Figure F1: Mole fraction of Hg in methane calculated with the UMR-PRMC model in
the vapor phase
2.00E-07
1.80E-07 exp
1.60E-07 UMR-PRMC
mole fraction of Hg
1.40E-07
1.20E-07
1.00E-07
8.00E-08
6.00E-08
4.00E-08
2.00E-08
0.00E+00
270 275 280 285 290 295
T[K]
Figure F2: Mole fraction of Hg in ethane calculated with the UMR-PRMC model in the
liquid phase
157
3.50E-07
exp
3.00E-07
mole fraction of Hg UMR-PRMC
2.50E-07
2.00E-07
1.50E-07
1.00E-07
5.00E-08
0.00E+00
270 275 280 285 290 295
T[K]
Figure F3: Mole fraction of Hg in propane calculated with the UMR-PRMC model in
the liquid phase
1.80E-06
1.60E-06 exp
1.40E-06 UMR-PRMC
mole fraction of Hg
1.20E-06
1.00E-06
8.00E-07
6.00E-07
4.00E-07
2.00E-07
0.00E+00
270 280 290 300 310 320
T[K]
Figure F4: Mole fraction of Hg in n.C5 calculated with the UMR-PRMC model in the
liquid phase
7.00E-06
exp
6.00E-06
UMR-PRMC
mole fraction of Hg
5.00E-06
4.00E-06
3.00E-06
2.00E-06
1.00E-06
0.00E+00
270 280 290 300 310 320 330 340
T[K]
liquid phase
158
2.50E-06
exp
mole fraction of Hg 2.00E-06 UMR-PRMC
1.50E-06
1.00E-06
5.00E-07
0.00E+00
270 280 290 300 310 320
T[K]
liquid phase
2.50E-06
exp
2.00E-06
mole fraction of Hg
UMR-PRMC
1.50E-06
1.00E-06
5.00E-07
0.00E+00
270 280 290 300 310 320
T[K]
liquid phase
3.50E-06
3.00E-06 UMR-PRMC
mole fraction of Hg
exp
2.50E-06
2.00E-06
1.50E-06
1.00E-06
5.00E-07
0.00E+00
270 280 290 300 310 320
T[K]
liquid phase
159
1.40E-06
UMR-PRMC
1.20E-06
mole fraction of Hg exp
1.00E-06
8.00E-07
6.00E-07
4.00E-07
2.00E-07
0.00E+00
270 275 280 285 290 295 300 305 310
T[K]
Figure F9: Mole fraction of Hg in 2.2-dm-C4 calculated with the UMR-PRMC model in
the liquid phase
1.80E-06
1.60E-06 UMR-PRMC
1.40E-06 exp
mole fraction of Hg
1.20E-06
1.00E-06
8.00E-07
6.00E-07
4.00E-07
2.00E-07
0.00E+00
270 275 280 285 290 295 300 305 310
T[K]
Figure F10: Mole fraction of Hg in 2.2.4-tm-C5 calculated with the UMR-PRMC model
in the liquid phase
3.00E-06
exp
mole fraction of Hg
2.50E-06
UMR-PRMC
2.00E-06
1.50E-06
1.00E-06
5.00E-07
0.00E+00
285 290 295 300 305 310 315
T[K]
Figure F11: Mole fraction of Hg in cy-C6 calculated the UMR-PRMC model in the
liquid phase
160
2.50E-06
exp
UMR-PRMC
1.50E-06
1.00E-06
5.00E-07
0.00E+00
270 275 280 285 290 295 300 305 310
T[K]
Figure F12: Mole fraction of Hg in m-cy-C6 calculated with the UMR-PRMC model in
the liquid phase
2.50E-06
exp
2.00E-06 UMR-PRMC
mole fraction of Hg
1.50E-06
1.00E-06
5.00E-07
0.00E+00
285 290 295 300 305 310
T[K]
Figure F13: Mole fraction of Hg in cis-1.2-dm-cy-C6 calculated with the UMR-PRMC

model in the liquid phase
2.50E-06
exp
mole fraction of Hg
2.00E-06 UMR-PRMC
1.50E-06
1.00E-06
5.00E-07
0.00E+00
285 290 295 300 305 310
T[K]
Figure F14: Mole fraction of Hg in trans-1.2-dm-cy-C6 calculated with the UMR-PRMC

161
2.50E-06
exp
mole fraction of Hg
2.00E-06
UMR-PRMC
1.50E-06
1.00E-06
5.00E-07
0.00E+00
285 290 295 300 305 310
T[K]
Figure F15: Mole fraction of Hg in trans-1.4-dm-cy-C6 calculated with the UMR-PRMC

2.50E-06
exp
mole fraction of Hg
2.00E-06
UMR-PRMC
1.50E-06
1.00E-06
5.00E-07
0.00E+00
285 290 295 300 305 310
T[K]
Figure F16: Mole fraction of Hg in cis-1.4-dm-cy-C6 calculated the UMR-PRMC model

in the liquid phase
4.50E-08
4.00E-08
mole fraction of Hg
3.50E-08 exp
3.00E-08 UMR-PRMC
2.50E-08
2.00E-08
1.50E-08
1.00E-08
5.00E-09
0.00E+00
5 15 25 35 45 55 65 75
P[bar]
Figure F17: Mole fraction of Hg in N2 calculated with the UMR-PRMC model in the
vapor phase
162
1.00E-07
9.00E-08 exp
8.00E-08 UMR-PRMC
mole fraction of Hg
7.00E-08
6.00E-08
5.00E-08
4.00E-08
3.00E-08
2.00E-08
1.00E-08
0.00E+00
270 290 310 330 350 370 390
T [K]
Figure F18: Mole fraction of Hg in water calculated with the UMR-PRMC model
7.00E-08
exp
6.00E-08
UMR-PRMC
5.00E-08
mole fraction of Hg
4.00E-08
3.00E-08
2.00E-08
1.00E-08
0.00E+00
270 275 280 285 290 295
T[K]
Figure F19: Mole fraction of Hg in CO2 calculated with the UMR-PRMC model
Table F1: Analytical results and absolute deviations of the mole fraction of Hg in
methane for the UMR-PRMC model in the vapor phase
Confidential data
ethane for the UMR-PRMC model in the liquid phase
Confidential data
163
propane for the UMR-PRMC model in the liquid phase
Confidential data
Table F4: Analytical results and absolute deviations of the mole fraction of Hg in n.C 5
for the UMR-PRMC model in the liquid phase
T[K] P[bar] xexp xcalc Dx%

278.15 1.00 1.90E-07 2.08E-07 9.52
283.15 1.00 2.70E-07 2.83E-07 4.99
288.15 1.00 3.60E-07 3.82E-07 6.18
293.15 1.00 4.60E-07 5.10E-07 10.97
298.15 1.00 6.80E-07 6.76E-07 0.66
303.15 1.00 9.30E-07 8.86E-07 4.71
308.15 1.00 1.20E-06 1.15E-06 3.91
313.15 1.00 1.60E-06 1.49E-06 6.97
Table F5: Analytical results and absolute deviations of the mole fraction of Hg in n.C6

273.15 1.00 1.80E-07 1.74E-07 3.11
278.15 1.00 2.40E-07 2.40E-07 0.10
283.15 1.00 3.40E-07 3.27E-07 3.72
288.15 1.00 4.60E-07 4.42E-07 4.01
293.15 1.00 5.90E-07 5.90E-07 0.03
298.15 1.00 8.30E-07 7.81E-07 5.94
298.15 1.00 8.41E-07 7.81E-07 7.14
303.15 1.00 1.10E-06 1.02E-06 6.88
308.15 1.00 1.42E-06 1.33E-06 6.12
313.15 1.00 1.81E-06 1.72E-06 4.97
313.15 1.00 1.90E-06 1.72E-06 9.41
318.15 1.00 2.38E-06 2.21E-06 7.32
323.15 1.00 3.03E-06 2.81E-06 7.36
328.15 1.00 3.84E-06 3.55E-06 7.61
333.15 1.00 4.83E-06 4.46E-06 7.76
338.15 1.00 6.03E-06 5.57E-06 7.69
164

273.15 1.00 2.00E-07 1.94E-07 3.19
278.15 1.00 2.90E-07 2.67E-07 8.03
283.15 1.00 3.70E-07 3.63E-07 1.78
288.15 1.00 5.40E-07 4.90E-07 9.22
293.15 1.00 7.00E-07 6.55E-07 6.45
298.15 1.00 9.70E-07 8.67E-07 10.65
303.15 1.00 1.26E-06 1.14E-06 9.74
308.15 1.00 1.63E-06 1.48E-06 9.20
313.15 1.00 2.20E-06 1.91E-06 13.14

273.15 1.00 2.50E-07 2.11E-07 15.76
278.15 1.00 3.30E-07 2.90E-07 12.06
283.15 1.00 4.70E-07 3.96E-07 15.84
288.15 1.00 6.10E-07 5.34E-07 12.51
293.15 1.00 7.90E-07 7.13E-07 9.74
298.15 1.00 1.10E-06 9.44E-07 14.18
303.15 1.00 1.40E-06 1.24E-06 11.51
308.15 1.00 1.78E-06 1.61E-06 9.41
313.15 1.00 2.30E-06 2.08E-06 9.47
n.C10 for the UMR-PRMC model in the liquid phase

273.15 1.00 4.00E-07 2.39E-07 40.29
278.15 1.00 4.90E-07 3.29E-07 32.84
283.15 1.00 5.80E-07 4.49E-07 22.67
288.15 1.00 8.70E-07 6.05E-07 30.45
293.15 1.00 9.60E-07 8.08E-07 15.78
298.15 1.00 1.37E-06 1.07E-06 21.88
303.15 1.00 1.69E-06 1.40E-06 16.89
308.15 1.00 2.04E-06 1.83E-06 10.39
313.15 1.00 2.52E-06 2.36E-06 6.33
318.15 1.00 3.09E-06 3.03E-06 2.10
165
Table F9: Analytical results and absolute deviations of the mole fraction of Hg in 2.2-
dm-C4 for the UMR-PRMC model in the liquid phase

273.15 1.00 1.70E-07 1.63E-07 4.01
288.15 1.00 3.70E-07 4.13E-07 11.61
293.15 1.00 4.90E-07 5.52E-07 12.57
298.15 1.00 6.30E-07 7.30E-07 8.96
303.15 1.00 8.10E-07 9.58E-07 18.25
308.15 1.00 9.90E-07 1.25E-06 25.91
2.2.4-tm-C5 for the UMR-PRMC model in the liquid phase

273.15 1.00 1.60E-07 2.05E-07 28.12
286.65 1.00 3.40E-07 4.74E-07 39.56
288.15 1.00 3.60E-07 5.18E-07 44.02
293.15 1.00 5.40E-07 6.92E-07 28.22
298.15 1.00 7.00E-07 9.16E-07 30.88
303.15 1.00 8.90E-07 1.20E-06 35.03
308.15 1.00 1.11E-06 1.56E-06 40.86
318.15 1.00 3.09E-06 2.05E-07 28.12
Table F11: Analytical results and absolute deviations of the mole fraction of Hg in cy-
C6 for the UMR-PRMC model in the liquid phase

288.15 1.00 8.20E-07 7.08E-07 13.68
293.15 1.00 1.02E-06 9.22E-07 9.62
298.15 1.00 1.32E-06 1.19E-06 9.79
298.15 1.00 1.20E-06 1.19E-06 0.56
303.15 1.00 1.62E-06 1.53E-06 5.81
308.15 1.00 1.92E-06 1.94E-06 1.08
313.15 1.00 2.47E-06 2.45E-06 0.77
Table F12: Analytical results and absolute deviations of the mole fraction of Hg in m-
cy-C6 for the UMR-PRMC model in the liquid phase

273.15 1.00 2.70E-07 2.97E-07 9.95
166
Continuation of table F12
288.15 1.00 6.60E-07 7.04E-07 6.72
293.15 1.00 8.80E-07 9.22E-07 4.79
298.15 1.00 1.17E-06 1.20E-06 2.31
303.15 1.00 1.46E-06 1.54E-06 5.58
308.15 1.00 1.88E-06 1.97E-06 4.78
Table F13: Analytical results and absolute deviations of the mole fraction of Hg in cis-
1.2-dm-cy-C6 for the UMR-PRMC model in the liquid phase

289.15 1.00 8.10E-07 7.74E-07 4.38
293.15 1.00 1.06E-06 9.63E-07 9.14
298.15 1.00 1.36E-06 1.26E-06 7.44
303.15 1.00 1.71E-06 1.62E-06 5.14
308.15 1.00 2.06E-06 2.08E-06 0.99
Table F14: Analytical results and absolute deviations of the mole fraction of Hg in cis-
1.4-dm-cy-C6 for the UMR-PRMC model in the liquid phase

288.15 1.00 7.40E-07 7.32E-07 1.02
293.15 1.00 1.02E-06 9.62E-07 5.65
298.15 1.00 1.33E-06 1.25E-06 5.45
303.15 1.00 1.64E-06 1.62E-06 1.23
308.15 1.00 2.11E-06 2.08E-06 1.58
trans-1.2-dm-cy-C6 for the UMR-PRMC model in the liquid phase

288.15 1.00 6.90E-07 7.11E-07 3.08
293.15 1.00 9.10E-07 9.35E-07 2.73
298.15 1.00 1.23E-06 1.22E-06 1.28
303.15 1.00 1.55E-06 1.57E-06 1.59
308.15 1.00 1.85E-06 2.02E-06 9.16
trans-1.4- dm-cy-C6 for the UMR-PRMC model in the liquid phase

288.15 1.00 6.90E-07 7.29E-07 5.72
167
Continuation of table F16
293.15 1.00 9.10E-07 9.59E-07 5.34
298.15 1.00 1.22E-06 1.25E-06 2.45
303.15 1.00 1.55E-06 1.61E-06 4.11
308.15 1.00 1.86E-06 2.07E-06 11.23
benzene for the UMR-PRMC model in the liquid phase (separate fit)

288.15 1.00 5.80E-07 5.79E-07 0.11
293.15 1.00 7.90E-07 7.88E-07 0.24
295.65 1.00 9.27E-07 8.89E-07 4.12
298.15 1.00 1.06E-06 1.06E-06 0.12
298.15 1.00 1.07E-06 1.06E-06 1.17
303.15 1.00 1.37E-06 1.42E-06 3.32
308.15 1.00 1.91E-06 1.87E-06 2.06
toluene for the UMR-PRMC model in the liquid phase (separate fit)

273.15 1.00 3.20E-07 2.89E-07 9.65
288.15 1.00 6.90E-07 7.35E-07 6.56
293.15 1.00 1.03E-06 9.83E-07 4.54
298.15 1.00 1.28E-06 1.30E-06 1.76
298.15 1.00 1.33E-06 1.30E-06 2.21
303.15 1.00 1.73E-06 1.71E-06 1.15
308.15 1.00 2.14E-06 2.23E-06 4.04
Table F19: Analytical results and absolute deviations of the mole fraction of Hg in o-
xylene for the UMR-PRMC model in the liquid phase (separate fit)

273.15 1.00 3.10E-07 3.22E-07 3.92
293.15 1.00 1.12E-06 1.10E-06 1.40
298.15 1.00 1.45E-06 1.47E-06 1.09
303.15 1.00 1.89E-06 1.93E-06 2.01
308.15 1.00 2.62E-06 2.51E-06 4.02
168
benzene for the UMR-PRMC model in the liquid phase (simultaneous fit)

288.15 1.00 5.80E-07 6.03E-07 3.90
293.15 1.00 7.90E-07 8.04E-07 1.74
295.65 1.00 9.27E-07 9.00E-07 2.96
298.15 1.00 1.06E-06 1.06E-06 0.20
298.15 1.00 1.07E-06 1.06E-06 1.09
303.15 1.00 1.37E-06 1.39E-06 1.54
308.15 1.00 1.91E-06 1.81E-06 5.40
toluene for the UMR-PRMC model in the liquid phase (simultaneous fit)

273.15 1.00 3.20E-07 2.80E-07 12.54
288.15 1.00 6.90E-07 7.26E-07 5.27
293.15 1.00 1.03E-06 9.77E-07 5.10
298.15 1.00 1.28E-06 1.30E-06 1.77
298.15 1.00 1.33E-06 1.30E-06 2.20
303.15 1.00 1.73E-06 1.72E-06 0.56
308.15 1.00 2.14E-06 2.25E-06 5.25
Table F22: Analytical results and absolute deviations of the mole fraction of Hg in o-
xylene for the UMR-PRMC model in the liquid phase (simultaneous fit)

273.15 1.00 3.10E-07 3.27E-07 5.57
293.15 1.00 1.12E-06 1.11E-06 1.11
298.15 1.00 1.45E-06 1.47E-06 1.09
303.15 1.00 1.89E-06 1.92E-06 1.75
308.15 1.00 2.62E-06 2.50E-06 4.51
Table F23: Analytical results and absolute deviations of the mole fraction of Hg in N 2
for the UMR-PRMC model in the vapor phase
Confidential data
169
water for the UMR-PRMC model
T [K] P [bar] Mole fraction of Hg Mole fraction of Hg abs dev %

(exp) (calc)
273.15 1 3.73E-09 3.72E-09 0.17
278.15 1 3.91E-09 3.91E-09 0.04
283.15 1 4.15E-09 4.15E-09 0.11
288.15 1 4.46E-09 4.46E-09 0.02
293.15 1 4.83E-09 4.83E-09 0.05
298.15 1 5.28E-09 5.29E-09 0.10
303.15 1 5.83E-09 5.83E-09 0.00
308.15 1 6.48E-09 6.48E-09 0.04
313.15 1 7.27E-09 7.26E-09 0.10
318.15 1 8.20E-09 8.19E-09 0.08
323.15 1 9.31E-09 9.30E-09 0.07
328.15 1 1.06E-08 1.06E-08 0.27
333.15 1 1.22E-08 1.22E-08 0.10
338.15 1 1.41E-08 1.41E-08 0.04
343.15 1 1.64E-08 1.64E-08 0.16
348.15 1 1.91E-08 1.91E-08 0.04
353.15 1 2.24E-08 2.24E-08 0.18
358.15 1 2.63E-08 2.63E-08 0.04
363.15 1 3.10E-08 3.10E-08 0.07
373.15 1 4.36E-08 4.36E-08 0.06
383.15 1 6.21E-08 6.21E-08 0.05
393.15 1 8.91E-08 8.92E-08 0.09
CO2 for the UMR-PRMC model in the liquid phase
Confidential data
170
Appendix G
Table G1: B.P.P calculation with the UMR-PRMC model for the estimation of the
mole fraction of Hg in ethane in the vapor phase
Confidential data
Table G2: B.P.P calculation with the PR-MC model for the estimation of the mole
fraction of Hg in ethane in the vapor phase
Confidential data
Table G3: B.P.P calculation with the SRK-Twu model for the estimation of the mole
fraction of Hg in ethane in the vapor phase
Confidential data
mole fraction of Hg in propane in the vapor phase
Confidential data
fraction of Hg in propane in the vapor phase
Confidential data
fraction of Hg in propane in the vapor phase
Confidential data
mole fraction of Hg in CO2 in the vapor phase
Confidential data
fraction of Hg in CO2 in the vapor phase
Confidential data
fraction of Hg in CO2 in the vapor phase
Confidential data
171
1.00E-07
9.00E-08
8.00E-08
mole fraction of Hg
7.00E-08
6.00E-08
5.00E-08 exp
4.00E-08 PR-MC
3.00E-08
SRK-Twu
2.00E-08
UMR-PRMC
1.00E-08
0.00E+00
270 275 280 285 290 295
T[K]
Figure G1: Mole fractions of Hg with ethane in the vapor phase estimated by all
models by B.P.P calculation
3.00E-07
2.50E-07
mole fraction of Hg
2.00E-07
PR-MC
1.50E-07
exp
1.00E-07 SRK-Twu
UMR-PRMC
5.00E-08
0.00E+00
270 275 280 285 290 295
T[K]
Figure G2: Mole fractions of Hg with propane in the vapor phase estimated by all
models by B.P.P calculation
172
6.00E-08
4.00E-08
PR-MC
3.00E-08
SRK-Twu
2.00E-08 exp
UMR-PRMC
1.00E-08
0.00E+00
270 275 280 285 290 295
T[K]
Figure G3: Mole fractions of Hg with CO2 in the vapor phase estimated by all models
by B.P.P calculation
173
Appendix H
6.00E-02
4.00E-02 y = -0.0008x + 0.0686
2.00E-02 R = .9997
0.00E+00
fitted kij
kij
-2.00E-02 0 50 100 150 200

Linear (fitted kij)
-4.00E-02
-6.00E-02
-8.00E-02
T[C]
Figure H1: Generalized correlations for the PR-MCs kij parameters of paraffinic HCs
based on their Tb property
6.00E-02
4.00E-02 y = -0.0008x + 0.0788

R = .9995
2.00E-02
0.00E+00
fitted kij
kij
0 50 100 150 200

-2.00E-02
Linear (fitted kij)
-4.00E-02
-6.00E-02
-8.00E-02
T[C]
Figure H2: Generalized correlations for the SRK-Twus kij parameters of paraffinic HCs
6.00E-02
4.00E-02 y = -0.0224x + 0.1478
2.00E-02 R = .9942
0.00E+00
-2.00E-02 0 2 4 6 8 10 12 fitted kij
kij
-4.00E-02 Linear (fitted kij)

-6.00E-02
-8.00E-02
-1.00E-01
CN
Figure H3: Generalized correlations for the PR-MCs kij parameters of paraffinic HCs
based on their CN
174
6.00E-02
4.00E-02
y = -0.0211x + 0.1534
2.00E-02 R = .9929
0.00E+00
fitted kij
kij
0 2 4 6 8 10 12
-2.00E-02
Linear (fitted kij)
-4.00E-02
-6.00E-02
-8.00E-02
CN
Figure H4: Generalized correlations for the SRK-Twus kij parameters of paraffinic HCs
based on their CN
4.00E-02
y = -0.0009x + 0.11
3.00E-02
R = .9968
2.00E-02
fitted kij
kij
1.00E-02 Linear (fitted kij)
0.00E+00
50 70 90 110 130
-1.00E-02
T[C]
Figure H5: Generalized correlations for the PR-MCs kij parameters of naphthenic HCs
6.00E-02
5.00E-02
y = -0.0012x + 0.1476
4.00E-02
R = .9875
3.00E-02
fitted kij
kij
2.00E-02
Linear (fitted kij)
1.00E-02
0.00E+00
-1.00E-02 50 70 90 110 130
T[C]
Figure H6: Generalized correlations for the SRK-Twus kij parameters of naphthenic
HCs based on their Tb property
175
4.00E-02
3.50E-02
3.00E-02 y = -0.0195x + 0.1538
2.50E-02 R = .9994
2.00E-02
fitted kij
kij
1.50E-02
Linear (fitted kij)
1.00E-02
5.00E-03
0.00E+00
-5.00E-03 4 5 6 7 8 9
CN
based on their CN
6.00E-02
5.00E-02
4.00E-02 y = -0.0257x + 0.2044

R = .9794
3.00E-02
fitted kij
kij
2.00E-02
Linear (fitted kij)
1.00E-02
0.00E+00
4 5 6 7 8 9
-1.00E-02
CN
based on their CN
1.20E-01 y = -0.0011x + 0.1889

R = .92 5
1.00E-01
8.00E-02
fitted kij
kij
6.00E-02
Linear (fitted kij)
4.00E-02
2.00E-02
0.00E+00
50 70 90 110 130 150
T[C]
Figure H9: Generalized correlations for the PR-MCs kij parameters of aromatic HCs
176
1.40E-01
1.20E-01
1.00E-01 y = -0.0011x + 0.1975
R = .921
8.00E-02
kij
6.00E-02 fitted kij

4.00E-02 Linear (fitted kij)
2.00E-02
0.00E+00
50 70 90 110 130 150
T[C]
Figure H10: Generalized correlations for the SRK-Twus kij parameters of aromatic
HCs based on their Tb property
1.20E-01
y = -0.0317x + 0.2918
1.00E-01 R = .9 96
8.00E-02
kij
6.00E-02
fitted kij
4.00E-02
Linear (fitted kij)
2.00E-02
0.00E+00
5 6 7 8 9
CN
Figure H11: Generalized correlations for the PR-MCs kij parameters of aromatic HCs
based on their CN
1.40E-01
1.20E-01 y = -0.0309x + 0.2979
1.00E-01 R = .91 1
8.00E-02
kij
6.00E-02 fitted kij

4.00E-02
Linear (fitted kij)
2.00E-02
0.00E+00
5 6 7 8 9
CN
Figure H12: Generalized correlations for the SRK-Twus kij parameters of aromatic
HCs based on their CN
177
Appendix I
Table I1: Analytical results for Hg in iso-butane with the UMR-PRMC model for both
liquid and vapor phase and their respective absolute deviations
Confidential data
Table I2: Analytical results for Hg in iso-butane with the PR-MC model using the Tb
based generalized correlations for both liquid and vapor phase and their respective
absolute deviations
Confidential data
Table I3: Analytical results for Hg in iso-butane with the PR-MC model using the CN
absolute deviations
Confidential data
Table I4: Analytical results for Hg in iso-butane with the SRK-Twu model using the Tb
absolute deviations
Confidential data
Table I5: Analytical results for Hg in iso-butane with the SRK-Twu model using the CN
absolute deviations
Confidential data
178
Appendix J
Table J1: Analytical results for Hg in iso-butane and propane with the UMR-PRMC
model in the liquid phase and their respective absolute deviations
Confidential data
Table J2: Analytical results for Hg in iso-butane and propane with the PR-MC model
using the Tb based generalized correlations in the liquid phase and their respective
absolute deviations
Confidential data
Table J3: Analytical results for Hg in iso-butane and propane with the PR-MC model
using the CN based generalized correlations in the liquid phase and their respective
absolute deviations
Confidential data
Table J4: Analytical results for Hg in iso-butane and propane with the SRK-Twu model
using the Tb based generalized correlations in the liquid phase and their respective
absolute deviations
Confidential data
Table J5: Analytical results for Hg in iso-butane and propane with the SRK-Twu model
absolute deviations
Confidential data
Table J6: Analytical results for Hg in n.C4, n.C5 and n.C6 with the UMR-PRMC model in
the liquid phase and their respective absolute deviations
Confidential data
Table J7: Analytical results for Hg in n.C4, n.C5 and n.C6 with the PR-MC model using
the Tb based generalized correlations in the liquid phase and their respective
absolute deviations
Confidential data
Table J8: Analytical results for Hg in n.C4, n.C5 and n.C6 with the PR-MC model using
the CN based generalized correlations in the liquid phase and their respective
absolute deviations
Confidential data
179
Table J9: Analytical results for Hg in n.C4, n.C5 and n.C6 with the SRK-Twu model using
the Tb based generalized correlations in the liquid phase and their respective
absolute deviations
Confidential data
Table J10: Analytical results for Hg in n.C4, n.C5 and n.C6 with the SRK-Twu model
absolute deviations
Confidential data
180

Diploma Thesis: Modelling of Mercury (HG) Distribution in Natural Gas Mixtures

Uploaded by

Document Informationclick to expand document information

Copyright:

Available Formats

Diploma Thesis: Modelling of Mercury (HG) Distribution in Natural Gas Mixtures

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Diploma Thesis: Modelling of Mercury (HG) Distribution in Natural Gas Mixtures

Uploaded by

Copyright:

Available Formats

Diploma Thesis

Modelling of Mercury (Hg) Distribution in Natural

Supervisor: Dr. Epaminondas C. Voutsas

Summary (English) ......................... 8

Chapter 1: Introduction .............. 11

Chapter 2: Theory .................... 15

2.1 Forms of mercury in natural gas ............ 15

2.2 Mercury distribution on gas processing plants ......... 16

Chapter 3: Evaluation of experimental data (critical properties, vapor pressure and

3.1 Available Tc, Pc, for Hg in literature ............... 20

3.2 Creation of Ps database and Solubility database ................. 20

3.2.1 Evaluation of Ps experimental data ................. 20

3.2.2 Evaluation of DIPPRs equation for the prediction of Ps .. 22

3.2.3 Development of database for binary Vapor-Liquid-Equilibrium (VLE)

3.2.4 Experimental errors of the data assembled ...... 32

Chapter 4: Thermodynamic modeling ........................... 37

4.1 SRK and PR EOS ................................ 37

4.3 Calculation of the Mathias-Copeman parameters for Hg for the PR EOS .. 42

4.4 Calculation of Twu parameters for Hg for the SRK EOS .. 44

4.5.1 Calculation of the kij parameters .................. 46

4.6 The UMR-PRMC model ........................... 53

4.6.1 Main groups ACH and ACCH3 ..................... 55

4.6.2 Main group CH2 ...................... 58

4.6.3 Main group cy-CH2 ....................... 63

4.6.4 Main group CH4 ...................... 66

4.6.5 Main group C2H6 ....................... 68

4.6.6 Main group CO2 .............................. 70

4.6.7 Main group N2 ................................ 72

4.6.8 Main group H2O ............................. 73

4.7.1 Evaluation of the generalized correlations for the binary interaction

Chapter 5: Testing of PR-MC, SRK-Twu and UMR-PRMC model in binary and

5.1 Iso-butane .................................. 82

5.2 K variable ....................................... 84

5.2.1 K variable for ethane ................................. 84

5.2.2 K variable for propane ............................... 85

5.2.3 K variable for CO2 ...................................... 87

5.3 Multi-component mixtures .................................... 90

Chapter 6: Conclusions .................................... 95

Chapter 7: Future work ......................................... 98

Appendix B .......................................................... 109

Appendix C .......................................................... 117

Appendix D .......................................................... 128

Appendix E ........................................................... 130

Appendix F ........................................................... 160

Appendix G .......................................................... 175

Appendix H .......................................................... 179

Appendix I ............................................................ 183

Appendix J ........................................................... 185

, Peng-Robinson (PR) Soave-Redlich-Kwong (SRK)

Mercury (Hg) is a toxic chemical element, commonly known as quicksilver or

Furthermore two predictive correlations have been developed, in order to estimate

Finally the three models developed as well as their corresponding generalized

I would like to express my sincere thankfulness to my supervisors and all my

I am indebted to my father for living, but to my teachers for living well.

Finally I would like to thank my family, my closest friends Mitrou Michalis,

This is the combined action of mercury and moisture producing a corrosion

Hg + Al Hg(Al) amalgam (1)