Professional Documents
Culture Documents
Thermodynamics Properties
Best Practice Guide
August 2016
I. Table of Contents
I. Introduction .......................................................................................................................................... 1
II. List of Thermodynamics Property Packages in Aspen HYSYS ............................................................... 3
Equation of State (EOS) ............................................................................................................................. 3
Activity Models ......................................................................................................................................... 3
Chao Seader & Grayson Streed Models .................................................................................................... 4
Vapor Pressure Models ............................................................................................................................. 4
Acid Gas, Glycol, Electrolyte, Steam ......................................................................................................... 4
Aspen Properties ....................................................................................................................................... 5
III. Common Property Packages in Upstream & Midstream processes ................................................. 6
Peng-Robinson (PR)................................................................................................................................... 6
Sour PR ...................................................................................................................................................... 8
SRK ............................................................................................................................................................ 8
Sour SRK .................................................................................................................................................... 9
CPA ............................................................................................................................................................ 9
Acid Gas Property Packages .................................................................................................................... 10
a. Acid Gas Chemical Solvents ..................................................................................................... 10
b. Acid Gas Physical Solvents ....................................................................................................... 11
c. Acid Gas Liquid Treating ........................................................................................................... 11
Glycol Property Package ......................................................................................................................... 12
Sulsim (Sulfur Recovery) Property Package ............................................................................................ 13
GERG 2008 .............................................................................................................................................. 14
PC-SAFT ................................................................................................................................................... 16
PSRK ........................................................................................................................................................ 18
IV. Common Upstream & Midstream Processes .................................................................................. 19
Upstream Processes................................................................................................................................ 19
a. Hydrate Formation & Inhibition.................................................................................................. 19
b. Methanol Injection...................................................................................................................... 20
c. Mercury Partitioning & Removal ................................................................................................ 20
d. Gas-Oil Separation Process (GOSP) ............................................................................................. 21
Gas Plant Processes ................................................................................................................................ 22
a. Acid Gas Treating & Tail Gas Treating Process............................................................................ 22
b. Sulfur Recovery Process .............................................................................................................. 24
i
c. Dehydration ................................................................................................................................ 25
d. Sour Water Processes (including Sour Water Stripper) .............................................................. 26
e. Nitrogen & Helium Rejection ...................................................................................................... 26
f. NGL Fractionation ....................................................................................................................... 27
g. LNG Compression ........................................................................................................................ 28
h. LNG Regasification ...................................................................................................................... 28
V. Regressing Property Data.................................................................................................................... 28
Setting Up a Regression .......................................................................................................................... 29
Entering Supplemental Parameters ........................................................................................................ 29
Pure Component Data ............................................................................................................................ 29
Generating Binary VLE and LLE Data....................................................................................................... 30
Running the Regression .......................................................................................................................... 31
Example of Regressing Vapor Liquid Equilibrium Data for Ethanol and Ethyl-Acetate .......................... 31
ii
I. Introduction
Thermodynamic models are used to represent the phase equilibrium behavior and energy level of pure
compounds and mixture systems. The process simulation results are calculated based on the
thermodynamic models selected in the Aspen HYSYS Properties Environment. Often times, simulation
results do not reflect what is really happening in a plant or field operation because of
Aspen HYSYS provides over 30 different thermodynamic models in the Properties Environment, in
addition to the Aspen Properties models available through integration. Selection of the appropriate
property package is key to producing accurate simulations. In this paper, we will discuss the common
Upstream and Midstream processes and operation and provide guidance on selection of the appropriate
models.
A high-level view of the Upstream and Midstream processes and operations are shown in Figure 1.1,
although the order may vary depending on your plant configuration and objectives. The Table 1.1 shows
the recommended property packages and considerations when modeling these processes in Aspen
HYSYS. Further details of each of the process and fluid package will be discussed in the corresponding
section.
1
Table 1.1 Summary of property package recommendation
2
II. List of Thermodynamics Property Packages in Aspen HYSYS
Property packages are specialized collections of methods for calculating the properties of components
at temperature, pressure, and composition of interest within the simulation. The following groups of
models are available in Aspen HYSYS:
BWRS
CPA
GCEOS
Kabadi Danner
Lee-Kessler Plocker
Peng-Robinson
PRSV
Sour PR
Sour SRK
SRK
Zudkevitch Joffee
Activity Models
The Activity Models handle highly non-ideal systems and are much more empirical in nature when
compared to the property predictions in the hydrocarbon industry. Polar or non-ideal chemical systems
are traditionally handled using dual model approaches. In this type of approach, an equation of state is
used for predicting the vapour fugacity coefficients and an activity coefficient model is used for the
liquid phase. Since the experimental data for activity model parameters are fitted for a specific range,
these property methods cannot be used as reliably for generalized application.
Chien Null
Extended NRTL
General NRTL
3
Margules
NRTL
UNIQUAC
van Laar
Wilson
Chao Seader
Grayson Streed
Antoine
Braun K10
Esso Tabular
4
OLI_Electrolyte
Sulsim (Sulfur Recovery)
Aspen Properties
Aspen Properties is the physical property system used in Aspen Plus. Aspen HYSYS has ability to access
its thermodynamics calculation engine. Aspen Properties supports a very wide range of property
packages suitable for petrochemical, chemical, specialty chemical, electrolytic, and polymer application.
However, only the following property packages from Aspen Properties will be discussed further in this
Guide:
GERG 2008
PC-SAFT
PSRK
5
III. Common Property Packages in Upstream & Midstream processes
Peng-Robinson (PR)
The Peng-Robinson (PR) model is ideal for VLE calculations for hydrocarbon systems. Several
enhancements to the original PR model were made to extend its range of applicability and to improve its
predictions for some non-ideal systems. However, in situations where highly non-ideal systems are
encountered, the use of Activity Models or advanced equation of state such as CPA or PC-SAFT is
recommended.
The PR property package rigorously solves any single-, two-, or three-phase system with a high degree of
efficiency and reliability and is applicable over a wide range of conditions:
The PR property package also contains enhanced binary interaction parameters for all library
hydrocarbon-hydrocarbon pairs (a combination of fitted and generated interaction parameters), as well
as for most hydrocarbon-non-hydrocarbon binaries, such as N2, CO2, H2S, H2, and H2O. For non-library or
hydrocarbon hypocomponents, HC-HC interaction parameters are generated automatically by HYSYS for
improved VLE property predictions.
6
Peng-Robinson property package also offers further tuning options.
Enthalpy:
EoS (Default)
Lee-Kesler
Density:
EoS
Costald (Default)
Rackett
Viscosity:
HYSYS (Default)
Indexed Viscosity
PR Options:
HYSYS (Default)
Standard - 1976
Phase Identification:
Surface Tension:
HYSYS (Default)
API 10A3.2 Method
Thermal Conductivity:
For Oil, Gas, or Petrochemical applications, the PR EOS is the generally recommended property package.
The PR property package can be used for the following simulations:
7
Atm Crude Towers
Vacuum Towers
High H2 Systems
Reservoir Systems
Hydrate Inhibition
Crude Systems
Nitrogen & Helium Rejection
Methanol Injection
Mercury Removal
GOSP
Fractionation
LNG Processes
Light Ends Separation
PR can also be used for TEG dehydration with and without aromatics, but the Glycol property package is
preferred.
Sour PR
The Sour PR model combines the Peng-Robinson equation of state and Wilson's API-Sour Model for
handling sour water systems and can be applied to:
In the Sour PR model, the K-values for the aqueous phase are calculated using Wilson's API-Sour
method. This option uses Wilson's API-Sour model to account for the ionization of the H2S, CO2, and NH3
in the aqueous water phase. The aqueous model employs a modification of Van Krevelen's original
model with many of the key limitations removed. The K-value of water is calculated using an empirical
equation, which is a function of temperature only.
The original Wilsons API-Sour model is applicable for temperatures between 20C (68F) and 140C
(285F), and pressures up to 50 psi. Use of the PR equation of state to correct vapour phase non
idealities extends this range, but due to lack of experimental data, exact ranges cannot be specified. The
acceptable pressure ranges for the HYSYS model vary depending upon the concentration of the acid
gases and H2O. The method performs well when the H2O partial pressure is below 100 psi.
SRK
In many cases, the Soave-Redlich-Kwong (SRK) model provides comparable results to Peng-Robinson,
but its range of application is significantly more limited:
8
The SRK EOS should not be used for non-ideal chemicals such as alcohols, acids or other components.
These chemicals are more accurately handled by the Activity Models (highly non-ideal) or the advanced
EOS model such as CPA or PC-SAFT.
The SRK property package is generally used for the following simulations:
TEG Dehydration
Cryogenic Gas Processing
Air Separation
Atm Crude Towers
Vacuum Towers
High H2 Systems
Reservoir Systems
Hydrate Inhibition
Crudes systems
Nitrogen & Helium Rejection
Methanol Injection
Mercury Removal
GOSP
Fractionation
LNG Processes
Light Ends Separation
Sour SRK
The Sour SRK model combines the Soave Redlich Kwong and Wilson's API-Sour Model and can be
applied to:
This model is analogous to the Sour PR model, except that SRK is used instead of PR.
CPA
The CPA method represents the Cubic-Plus-Association EOS model developed by Kontogeorgis and co-
workers (Kontogeorgis, Voutsas, Yakoumis, Tassios, IECR 1996). The model combines the SRK cubic EOS
with the Wertheim association term similar to that of SAFT, as present in the PC-SAFT model. The
models applicability covers the VLE and VLLE of mixtures containing hydrocarbons and polar/associating
chemicals such as water, alcohols, glycols, esters, and organic acids.
The CPA property package uses the EOS model for all thermodynamic properties.
9
In the absence of associating (hydrogen-bonding) species, the CPA model reduces to the SRK model. The
addition of the association term for components capable of exhibiting hydrogen-bonding (water,
alcohols, glycols, acids) enables CPA to better model the V(L)LE of complex systems.
The CPA property package has built-in pure component and binary parameters for use in modeling
processes such as Methanol gas cleaning. The built-in parameters are stored in the AP-EOS databank.
For species not in the databank, the required parameters are estimated. We recommend however, and
necessary for associating components, that you use the Aspen Properties Data Regression System (DRS)
to determine the pure component parameters from vapor pressure and liquid density data, and the
binary parameters from experimental phase equilibrium data (binary VLE and LLE data).
You must add the following components to your case in order to use the Acid Gas - Chemical
Solvents property package:
The Acid Gas - Chemical Solvents property package supports the following amines and amine blends:
MEA
DEA
MDEA
PZ
DGA
DIPA
TEA
MDEA + PZ
Sulfolane + MDEA
Sulfolane + DIPA
MDEA + MEA
MDEA + DEA
Guidelines:
10
CO2, H2O, H2S must be part of the component list. You are prompted to add them if they are not
present.
No salts or ions can be added to the component list. (The exception is Heat Stable Salts.)
Physical solvents cannot be added to the component list.
Only vapor-liquid models are supported.
The side stripper and side rectifier are not supported.
Different stages of an "Acid Gas" column cannot be associated with different fluid packages.
You must add the following components to your case in order to use the Acid Gas - Physical Solvents
property package:
DEPG
CO2
H2S
H2O
Unlike the Acid Gas - Chemical Solvents and the Acid Gas - Physical Solvents property packages, the Acid
Gas - Liquid Treating property package does not use a rate-based method.
11
You must add the following components to your case in order to add the Acid Gas - Liquid Treating
property package:
The Acid Gas - Liquid Treating property package supports the following amines and amine blends:
MEA
DEA
MDEA
DGA
MDEA + PZ
It also supports the main LPG and NGL components (ethane and propane), other light and heavy
hydrocarbons, BTX, COS, CS2, and light mercaptans.
Assays
Salts or ions (except for Heat Stable Salts)
Physical solvents
Solid components
Activity coefficients of the TEG-water solutions within the average absolute deviation of 2%
Water content of gas within the average absolute deviation of 1%
The Glycol property package should be applicable over the range of temperatures, pressures, and
component concentration encountered in a typical TEG-water dehydration system: between 15C to
50C and between 10 atm to 100 atm for the gas dehydrator, and between 202C to 206C and 1.2
atmospheres for the glycol regenerator.
The accuracy of predicted solubility of hydrocarbons in aqueous phase is expected to be within the
experimental uncertainty.
12
Sulsim (Sulfur Recovery) Property Package
The Sulsim (Sulfur Recovery) property package incorporates properties developed by Sulphur Experts for
the purpose of simulation the modified-Claus process, and uses the same Gibbs free energy, enthalpy,
and viscosity correlations. These parameters have been refined over years of industry experience to
ensure that simulation results match plant performance.
Aspen Technology has worked with Sulphur Experts to perform robust validation and ensure equal
accuracy in results compared to legacy versions of the Sulsim software.
When you add a Sulsim (Sulfur Recovery) property package prior to editing the new component list, the
component list will automatically include the required and supported components listed below. This
enables you to easily integrate your sulfur recovery simulation with other gas processes.
The following required components must be part of any component list that is associated with a Sulsim
(Sulfur Recovery) property package:
13
GERG 2008
The GERG2008 property package is available in Aspen Properties. It is based on the 2008 extension of
the GERG-2004 equation of state model for calculations of thermodynamic properties and phase
equilibrium of natural gases and other mixtures consisting of natural gas components.
This model was developed under the auspices of the Groupe Europen de Recherches Gazires (GERG)
in cooperation with the German technical and Scientific Association on Gas and Water and European
natural gas companies (E.ON Ruhrgas, Germany; Enagas, Spain; Gasunie, The Netherlands; Gaz de
France, France; Snam Rete Gas, Italy; and Statoil, Norway).
This is a highly accurate, wide-range equation of state that has been adopted as a standard (ISO-20765)
international reference equation suitable for all natural gas applications, including processing,
transportation, and storage of natural gas. The model is applicable to 21 natural gas components and
their mixtures, including Methane, Nitrogen, Carbon dioxide, Ethane, Propane, n-Butane, Isobutane, n-
Pentane, Isopentane, n-Hexane, n-Heptane, n-Octane, n-Nonane, n-Decane, Hydrogen, Oxygen, Carbon
monoxide, Water, Helium, Argon, and Hydrogen sulfide.
The GERG2008 model can be applied to natural gases, rich natural gases (natural gases that contain
large amounts of ethane and heavier alkanes), liquefied natural gases, liquefied petroleum gases, highly
compressed natural gases and hydrogen-hydrocarbon mixtures, such as Hythane in the homogeneous
gas, liquid, and supercritical regions as well as in the vapor-liquid equilibrium (VLE) region.
The equation of state is based on pure substance equations of state for each applicable component and
correlation equations for binary mixtures consisting of these components. This allows predictive
description of multi-component mixtures over a wide range of compositions. The equation of state was
developed and tested using an extensive database of binary mixtures, natural gases, and other multi-
component mixtures data of over 100,000 data points for the thermal and caloric properties. The data
used cover the homogeneous gas, liquid, and supercritical regions as well as vapor-liquid equilibrium
(VLE) states. The types of data used include: PVT, isochoric heat capacity (Cv), speed of sound (w),
isobaric heat capacity (Cp), enthalpy difference, excess molar enthalpy, saturated liquid density, and VLE
data. About 75% of the data are for binary mixtures, while the remaining 25% are for multi-component
mixtures. Almost 70% of the mixture data describe a PVT relation, more than 20% of the data are vapor-
liquid equilibrium state points, and less than 10% account for caloric properties. About 50% of the
available binary mixture data were used for model development while the remaining data were used for
model validation.
The model cannot be used for components other than the ones listed above.
The GERG2008 model is valid in the homogeneous gas, liquid, and supercritical regions and in the vapor-
liquid states. Due to the vast amount of experimental data for the different binary and multi-
component mixtures, and the varying real mixture behavior, which strongly depends on temperature,
pressure and composition, it is impossible to discuss each binary and multi-component system
separately. In general, there are no restrictions concerning the composition range, but the uncertainty
is mostly unknown for the composition ranges not covered by the experimental data.
The range of validity of the model related to the calculation of thermal and caloric properties of
mixtures is divided into three parts: the normal range and the extended range.
14
In the normal range, the validity covers temperatures of 90 K T 450 K and pressures of p 35 MPa.
This range corresponds to the use of the equation in applications using natural gas and related mixtures
in pipeline transport, natural gas storage, and improved processes with liquefied natural gas.
The uncertainty for the important properties and phases are as follow:
The uncertainty in gas phase density is < 0.1% for the temperature range from 250 450 K and
pressure up to 35 MPa for various types of natural gases, including natural gases rich in
nitrogen, rich in carbon dioxide, rich in ethane, rich in hydrogen, and natural gases containing
considerable amounts of propane, and heavier alkanes, carbon monoxide, or oxygen and other
mixtures consisting of the 21 natural gas components. The great majority of experimental
densities for various rich natural gases containing large amounts of carbon dioxide (up to 20%),
ethane (up to 18%), propane (up to 14%), n-butane (up to 6%), n-pentane (up to 0.5%), and n-
hexane (0.2%) are reproduced to within (0.1 0.3)% over the temperature range 280 K to 350
K and pressure up to 30 MPa.
The uncertainty in gas phase speed of sound is < 0.1% in the temperature range from 270 K to
450 K and pressures up to 20 MPa, and in the temperature range from 250 K to 270 K at
pressures up to 12 MPa. At higher pressures, the uncertainty is < (0.2 0.3) %.
The uncertainty in gas phase isobaric enthalpy difference is < (0.2 0.5) %.
The uncertainty for isobaric and isochoric heat capacity (Cp and Cv, respectively) in the
homogeneous gas, liquid, and supercritical regions is (1 2)%
The uncertainty for liquid phase isobaric enthalpy difference is (0.5 1) %.
The uncertainty in liquid phase density is (0.1 0.5)%.for pressures up to 40 MPa for many
binary and multi-component mixtures, including LNG-like mixtures and mixtures of light and
heavy hydrocarbons.
The uncertainty for saturated liquid density for the same types of mixtures in the temperate
range from 100 K to 140 K, which is very important for processes with liquefied natural gas, is <
(0.1 0.3)%.
Vapor-liquid equilibrium data of binary and multi-component mixtures as well as dew points of
natural gases and hydrocarbon mixtures are accurately described. The most accurate vapor
pressure data for binary and ternary mixtures consisting of the natural gas main components
(methane, nitrogen, carbon dioxide, and ethane), as well as propane, n-butane and isobutene
are reproduced to within their experimental uncertainty (~ (1-3)%). Certain other mixtures have
higher uncertainty of up to 5% or more. Accurate experimental vapor phase compositions are
described to within (0.5 1) mole %.
The extended range of validity covers temperatures of 60 K T 700 K and pressures of p 70 MPa and
has the following uncertainty:
The uncertainty of gas phase density is (0.2 0.5)% at temperatures and pressures outside the
normal range of validity
Density data of certain binary mixtures are described to within (0.5 1)% at pressures up to
100 MPa or more.
The current data situation outside the normal range does not allow for well-founded estimates
of uncertainty of other thermodynamic properties.
15
PC-SAFT
The PC-SAFT property package is available in Aspen Properties. It is based on the perturbed-chain
statistical associating fluid theory (PC-SAFT) equation of state of Gross and Sadowski (2000, 2001,
2002a).1.
PC-SAFT represents an improved version of the very successful SAFT EOS. Therefore, its applicability
includes fluid systems of small and/or large molecules over a wide range of temperature and pressure
conditions. The big advantage of this EOS method is that it can represent the thermodynamic properties
of polymer systems very well2. In addition, it is better than other chain equations of state (Sanchez-
Lacombe, SAFT) in describing the properties of conventional chemicals. In fact, its accuracy is
comparable to, and often better than, the Peng-Robinson EOS or other similar cubic equations of state
for small molecules.
The perturbation term in SAFT takes into account the attractive (dispersion) interactions between
molecules. In PC-SAFT, Gross and Sadowski used the Barker-Henderson second-order perturbation
theory of spherical molecules and extended it to chain molecules. The idea is that the perturbation
theory concept applies to segments that are connected to chains rather than between disconnected
segments, which is the case in SAFT3. This is equivalent to considering attractive (dispersion) interactions
between the connected segments instead of disconnected ones. For example:
This concept offers a more realistic picture of how chain molecules, such as hydrocarbons, oligomers,
and polymers, behave in a solution.
PC-SAFT expresses the attractive term of the equation as a sum of two terms (first- and second-order
perturbation terms):
1
Gross, J., & Sadowski, G. (2000). Application of perturbation theory to a hard-chain reference fluid: An equation
of state for square-well chains. Fluid Phase Equil., 168, 183.
2
Gross, J., & Sadowski, G. (2002a). Modeling polymer systems using the perturbed-chain statistical associating
fluid theory equation of state. Ind. Eng. Chem. Res., 41, 1084-1093.
3
Gross, J., & Sadowski, G. (2001). Perturbed-Chain SAFT: An equation of state based on a perturbation theory for
chain molecules. Ind. Eng. Chem. Res., 40, 1244-1260.
16
Where A denotes the Helmholtz free energy. The Helmholtz free energy is used frequently in statistical
thermodynamics to express equations of state because most properties of interest, such as the system
pressure, can be obtained by proper differentiation of A. The coefficients A1 and A2 have a dependence
on density and composition, as well as molecular size. All the necessary constants that appear in the
coefficients of the previous equation have been obtained by Gross and Sadowski (2000) by regression of
thermophysical properties of pure n-alkanes.
There are three pure-component parameters for each chemical substance. These are the:
Segment number, m
Segment diameter,
Segment energy,
These parameters are obtained by fitting experimental vapor pressure and liquid molar volume data for
pure components. Also, a kij binary interaction parameter is used to fit phase equilibrium binary data;
this parameter defaults to zero if not supplied.
The PC-SAFT method represents the copolymer PC-SAFT EOS model developed by Sadowski and co-
workers (Gross and Sadowski, 2001, 2002a, 2002b; Gross et al., 2003; Becker et al., 2004; Kleiner et al.,
2006). The model includes the association and polar terms. Its applicability covers fluid systems from
small to large molecules, including normal fluids, water, alcohols, and ketones, polymers and
copolymers and their mixtures.
Plots of methane-butane VLE at 21.1C (Figure 3.2) and ethane-decane VLE at 238C (Figure 3.3) show
the PC-SAFT results for typical hydrocarbon mixtures.
17
Figure 3.3 Ethane-Decane VLE at 238 C. Predictions usingkij=0.
PSRK
The PSRK property package is available in Aspen Properties. It is based on the Predictive Soave-Redlich-
Kwong equation-of-state model (Holderbaum and Gmehling, Fluid Phase Equilibria. 70 (1991). pp. 251-
265).
Alpha functions that can accurately represent vapor pressure of a wide range of pure
components, including hydrocarbons, water, alcohols, and other polar compounds.
AE mixing rules that can be used to fit phase equilibrium data of highly non ideal systems.
By default, AE mixing rules use a UNIFAC model, specially formulated for PSRK to handle light
gases (C1, CO2, H2S, etc) and solvents, thus making the model predictive.
18
IV. Common Upstream & Midstream Processes
Upstream Processes
a. Hydrate Formation & Inhibition
Hydrate Formation Analysis is available under the Stream Analysis Tools in Aspen HYSYS. It is
used to calculate the incipient hydrate formation conditions in a given stream. That is, at the
given stream temperature, the hydrate formation pressure is calculated. And at the stream
pressure, the hydrate formation temperature is calculated. It can also consider the effect of
inhibitor (methanol, EG, DEG, or TEG) and will determine the minimum amounts of the selected
inhibitor required to suppress hydrate formation at the stream condition.
The Hydrate Formation Analysis can handle three types of hydrates, Type I, II and H. The
hydrate type that will form is determined automatically.
The models used in predicting the hydrate formation conditions are based on fundamental
thermodynamic principles (van der Waals-Platteeuw; see e.g., W.R. Parrish, J.M. Prausnitz, Ind.
Eng. Chem. Proc. Des. Dev. 11 (1) (1972) 2635.) and use equation of state in calculating the
equilibrium conditions. Two models/methods are available for Type I and II hydrates:
1. Ng-Robinson
2. CSM
H-J. NG and D.B. Robinson, The prediction of hydrate formation in condensed systems,
AIChE Journal, 23, 4 (1977).
H-J. NG and D.B. Robinson, The measurement and prediction of hydrate formation in liquid
hydrocarbon-water system, Ind. Eng. Chem. Fundam., 15, 4 (1976).
N.-J. Ng and D. B. Robinson, A method for predicting the equilibrium gas phase water
content in gas-hydrate equilibrium, Ind. Eng. Chem. Fundam., 19 (1980)
The CSM model is based on the work of Professor Sloans group at the Colorado School of
Mines. The pertinent references are:
A.L. Ballard, E.D. Sloan, Fluid Phase Equilibria 194-197, (2002) 371-383.
M.D. Jager, A.L. Ballard, E.D. Sloan, Fluid Phase Equilibria 211, (2003) 85-107.
E.D. Sloan, C. A. Koh, Clathrate Hydrates of Natural Gases Third Edition, Taylor & Francis
Group, 2007.
For structure H hydrates, the model developed by Mehta and Sloan [E.D. Sloan, Clathrate
hydrates of natural gases, 2nd edition, 1998, 705pp] is used.
Currently, the Hydrate Formation Analysis can be used with the Peng-Robinson, SRK, Glycol
property package, and CPA.
19
CPA
When methanol is used as the hydrate inhibitor, CPA is recommended. The CPA property
method has built-in pure component and binary parameters for use in modeling processes
that has Methanol.
A quick introduction to the CPA and methanol partitioning can be accessed at:
https://www.aspentech.com/HYSYS-Methanol-Partitioning/
PR
Peng-Robinson has the largest applicable range for temperature and pressure and can be
used for Hydrate Formation Analysis.
SRK
When using the CSM model in the Hydrate Formation Analysis, the SRK property package is
recommended for the best results. The CSM model was developed based on SRK.
b. Methanol Injection
Methanol injection can be seen in the gas plant for the gas dehydration process (see
Dehydration topic for further details). Another use of methanol is as an inhibitor for hydrate
formation.
CPA
When Methanol is used as the hydrate inhibitor or in gas dehydration, CPA is
recommended. The CPA property method has built-in pure component and binary
parameters for use in modeling processes that has Methanol.
PR
Peng-Robinson has the largest applicable range for temperature and pressure.
SRK
20
hydrocarbons as well as form its own liquid phase apart from the hydrocarbon liquid and water
phase under certain conditions.
Modeling of systems containing mercury can be performed in Aspen HYSYS using the Peng-
Robinson, CPA, and SRK property packages. These packages contain binary parameters for
mercury with water, hydrocarbons (including hypothetical components), methanol, and glycols.
The mercury free phase and mercury distribution are displayed in process streams when up to
three phases are present (up to VLLE). When four phases are expected (VLLLE), the Free
Mercury Analysis is recommended to accurately predict the mercury partitioning behavior.
Free Mercury Analysis is available under Stream Analysis Tools in Aspen HYSYS. The Free
Mercury Analysis calculates the formation of free mercury phase, other mass flow of the free
mercury phase if applicable, concentration of the dissolved mercury in phases, and mercury
formation conditions at given stream temperature and pressure. Currently, Free Mercury
Analysis can be used with the Peng-Robinson, CPA, and SRK.
CPA
In systems with water, alcohol, glycols, and acids, where hydrogen-bonding is present, CPA
is recommended for more accurate prediction.
A useful example of the mercury partitioning behavior is provided in the <HYSYS Install Dir>
| Samples folder:
o Mercury Removal with Component Splitter
PR
Peng-Robinson has the largest applicable range for temperature and pressure.
SRK
21
Gas Plant Processes
a. Acid Gas Treating & Tail Gas Treating Process
Chemical solvents are suitable for low acid gas partial pressures.
Development and validation work has been performed on the amine solvent using the Acid
Gas Chemical Solvent property package. A white paper has been published and available
at:
http://www.aspentech.com/wp-acid-gas-validation-Amines-NOV14/
Several useful examples of the acid gas cleaning process are provided in the <HYSYS Install
Dir> | Samples | Acid Gas Cleaning folder:
o Acid Gas Cleaning Using MDEA
o Acid Gas Cleaning Using DEA
o Acid Gas Cleaning Using DGA
o Acid Gas Cleaning Using DIPA
o Acid Gas Cleaning Using MDEA + Piperazine
o Acid Gas Cleaning Using TEA
o Acid Gas Cleaning Using Sulfolane-MDEA
o Acid Gas Cleaning Using Sulfolane-DIPA
o Acid Gas Cleaning Using PZ
22
o Acid Gas Cleaning Using MEA
o Effect of Heat Stable Salts On Acid Gas Cleaning Using MDEA
o Acid Gas and COS-CS2-Mercaptans Cleaning Using MDEA
o Acid Gas Cleaning Using DEA with 2+1 Superclaus Process
o Acid Gas Cleaning Using DEA with 3 Stage Claus Process
o Acid Gas Cleaning Using Sulfolane-DIPA + 2 Stage Claus Process + Tail Gas Treating
o Acid Gas Cleaning Using Sulfolane-DIPA with 3 Stage Claus Process
Physical solvents are favorable for bulk removal of acid gases at high pressure.
Development and validation work has been performed on the DEPG physical solvent using
the Acid Gas Physical Solvent property package. A white paper has been published and
available at:
http://www.aspentech.com/wp-acid-gas-validation-DEPG-NOV14/
Several useful examples of the acid gas cleaning process are provided in the <HYSYS Install
Dir> | Samples | Acid Gas Cleaning folder:
o Acid Gas Cleaning Using a Physical Solvent (DEPG)
The Acid Gas Liquid Treating property package supports the following amines and amine
blends:
MEA
23
DEA
MDEA
DGA
MDEA + PZ
It also supports the main LPG and NGL components (ethane and propane), other light and
heavy hydrocarbons, BTX, COS, CS2, and light mercaptans.
Several useful examples of the acid gas cleaning process are provided in the <HYSYS Install
Dir> | Samples | Acid Gas Cleaning folder:
o Liquid-Liquid Treating Using MDEA
Some of the Sulfur Recovery process produces elemental sulfur from gaseous hydrogen sulfide
removed from sour natural gas and from the byproduct gases containing H2S derived from
refining crude oil and other industrial processes. H2S can originate in gas treatment units in
refineries and natural gas plants, sour water stripping facilities, and coal gasification and syngas
plants. Since H2S is harmful to the environment, extremely flammable, and can dissolve in
petroleum products and degrade quality, sulfur removal is a highly important process.
Sulsim
Sulsim (Sulfur Recovery) property package is recommended when modeling a sulfur
recovery process. Several useful examples of the sulfur recovery process are provided in the
<HYSYS Install Dir> | Samples | Sulsim (Sulfur Recovery) folder:
o 2 Stage Claus Process
o 2 Stage Claus Process with Incinerator
24
o 2 Stage Claus Process with Tail Gas Treating Unit
o 3 Stage Claus Process
o 3 Stage Claus Process with Incinerator
o 3 + 1 Superclaus Process
o 2 Trains 2 Stage Claus Process with Common Incinerator
o Acid Gas Cleaning & Refining Integration
c. Dehydration
The purpose of dehydration process is to remove water from natural gas or gas and liquid
products in the process plant. Dehydration process is typically useful in the area of gas
gathering, product dehydration, and hydrocarbon recovery. In upstream oil and gas pipelines,
dehydration is important to prevent hydrate formation or corrosion. In most products
specification, no free water should be present. This specification means a maximum water in
sales gas at 4-7 lbm/MMscf and 10-20 ppmw in liquid.
Previous success has been observed with using Glycol Property Package with TEG. A
customer success story has been published and available at:
http://www.aspentech.com/process_ecology_case_study.pdf
An online seminar on the TEG dehydration process is available through the support center:
25
http://support.aspentech.com/webteamasp/KB.asp?ID=144747
CPA
In the dehydration process where methanol is used, CPA should be used. . The CPA property
method has built-in pure component and binary parameters for use in modeling processes
that has Methanol.
Sour SRK
Sour PR
Both Sour SRK and Sour PR adapt Wilson's API-Sour Model for handling sour water
components, either the PR or SRK model is used for the remaining components. In general,
Sour PR may give slightly better results comparing to Sour SRK because more binary
interaction parameters are available.
Figure 4.4 Generic midstream process (highlighting Nitrogen and Helium removal)
Nitrogen Rejection Unit (NRU) is employed for removing nitrogen from natural gas when
excessive nitrogen level is observed in the natural gas feed stream. Cryogenic Distillation is a
method commonly deployed when the feed gas rate is above 15 MMscfd4. An example file
4
Kidnay, Arthur J., Daniel G. McCartney, and William Rutledge. Parrish. Fundamentals of Natural Gas Processing.
Boca Raton, FL: CRC, 2011. Print.
26
Nitrogen Rejection Using Cryogenic Distillation.hsc is available under <HYSYS Install Dir> |
Samples folder to demonstrate the use of cryogenic distillation with single-column configuration
to remove nitrogen.
PR
Peng-Robinson property package is recommended for this process because of its
applicability over wide range of conditions (Temperature Range > -271C or -456F and
Pressure Range < 100,000 kPa or 15,000 psia).
f. NGL Fractionation
NGL (Natural Gas Liquids) fractionation is a process used to separate NGLs from mixed feeds.
The location of the fractionation unit can vary depending on the feed composition, process
conditions, and desired products. Typically, a fractionation unit will be located after the
dehydration and nitrogen removal steps.
PR
Peng-Robinson property package is recommended for this process because of its
applicability over wide range of conditions (Temperature Range > -271C or -456F and
Pressure Range < 100,000 kPa or 15,000 psia).
27
g. LNG Compression
PR
Peng-Robinson property package is recommended for this process because of its
applicability over wide range of conditions (Temperature Range > -271C or -456F and
Pressure Range < 100,000 kPa or 15,000 psia).
h. LNG Regasification
PR
Peng-Robinson property package is recommended for this process because of its
applicability over wide range of conditions (Temperature Range > -271C or -456F and
Pressure Range < 100,000 kPa or 15,000 psia).
You can use experimental property data to determine the physical property model parameters
you need for an Aspen Physical Property System calculation.
The Data Regression System fits parameters of physical property models to measured data for
pure component or multicomponent systems. You can enter almost any kind of experimental
property data, such as:
28
Vapor-liquid equilibrium
Liquid-liquid equilibrium
Density
Heat capacity
Activity coefficients
You can use Data Regression for all property models in the Aspen Physical Property System,
including electrolyte and user models.
Setting Up a Regression
To set up a Data Regression:
1. Start Aspen Plus or Aspen Properties and create a new run from a template.
2. On the Home tab of the ribbon, in the Run Mode group, click Regression.
3. Define components on the Components | Specifications | Selection sheet.
4. Select a property method on the Methods | Specifications | Global sheet.
5. Enter or estimate any supplemental property parameters on the Methods | Parameters
and Estimation forms.
6. Enter experimental data on the Data forms.
7. Specify the regression case on the Regression | Input form. See Formulating a
Regression Case.
For example, suppose you are regressing binary VLE data using the WILSON property method
and a component is not in the databank. You must enter or estimate the following parameters:
Molecular weight(MW), critical temperature(Tc), critical pressure(Pc), critical compressibility
factor(Zc), heat of vaporization(DHVLWT), Vapor pressure(PLXANT), and ideal gas heat
capacity(CPIG).
You can also enter values of the parameters to be determined on a Methods | Parameters
form. Data Regression will use these values as initial guesses.
Use the Data | PURE-COMP form to enter the experimental data as a function of
temperature.
Use the Regression | Input form to specify the property method, experimental data, and
parameters to be regressed.
29
Use the Data | PURE-COMP form to enter experimental data for pure component properties as
a function of temperature. For example, you can enter vapor pressure versus temperature data.
If you want the Aspen Physical Property System to ignore some data or standard deviations that
have already been entered, in the Usage column on the row, select IGNORE. The Aspen Physical
Property System will not use the data point in any subsequent regressions.
For example, you can generate VLE data using the UNIFAC predictive property method and use
the generated data to determine the binary parameters for the WILSON property method.
1. In the navigation pane, click Data. The Data object manager appears.
2. To create a new Data ID, click New. In the Create New ID dialog box, enter an ID or accept
the default. Choose MIXTURE in the Select Type list box, and click OK.
To edit an existing ID, select the Data ID from the object manager, and click Edit.
3. On the Setup sheet, choose the type of property data in the Data Type list box:
Select To generate this data
TXY, PXY, or TPXY VLE
TXX or TPXX LLE
30
Note: Do not select the GEN-TPXY or GEN-TPXX data type
4. Select the components from the Available Components list and click the right arrow button
to move them to the Selected Components list.
5. In the Temperature and Pressure fields, if active, specify a constant temperature or
pressure at which the data will be generated.
6. Click the Data tab.
7. On the Data sheet, click the Generate Data button.
8. In the Generate Binary VLE or LLE Data dialog box, select a property method, and a Henry's
Components ID and Chemistry ID, if applicable.
9. Click the Generate button.
The Data sheet displays the liquid phase compositions for which data are to be generated for
the regression. When you run the regression, these parameters will be generated and displayed
on the Regression | Results | Profiles sheet.
Note: The order in which the regression cases are run may be significant. The regressed parameter
values from a regression case are used automatically in all subsequent regression cases, thus
affecting their results. The Aspen Physical Property System will execute the regression cases in the
order they appear in the Run area.
Example of Regressing Vapor Liquid Equilibrium Data for Ethanol and Ethyl-Acetate
For an ethanol-ethyl acetate system, the following vapor liquid equilibrium data are available.
40C and 70C data of Martl, Collect.Czech. Chem. Commun. 37,266
(1972):
T=40C T=70C
P MMHG X ETOAC Y ETOAC P MMHG X ETOAC Y ETOAC
136.600 0.00600 0.02200 548.600 0.00650 0.01750
150.900 0.04400 0.14400 559.400 0.01800 0.04600
163.100 0.08400 0.22700 633.600 0.13100 0.23700
183.000 0.18700 0.37000 664.600 0.21000 0.32100
191.900 0.24200 0.42800 680.400 0.26300 0.36700
199.700 0.32000 0.48400 703.800 0.38700 0.45400
208.300 0.45400 0.56000 710.000 0.45200 0.49300
31
210.200 0.49500 0.57400
712.200 0.48800 0.51700
211.800 0.55200 0.60700
711.200 0.62500 0.59700
213.200 0.66300 0.66400
706.400 0.69100 0.64100
212.100 0.74900 0.71600
697.800 0.75500 0.68100
204.600 0.88500 0.82900
679.200 0.82200 0.74700
200.600 0.92000 0.87100
651.600 0.90300 0.83900
195.300 0.96000 0.92800
635.400 0.93200 0.88800
615.600 0.97500 0.94800
Atmospheric data of Ortega J. and Pena J.A., J. Chem. Eng.Data 31,
339 (1986):
T C X ETOAC Y ETOAC T C X ETOAC Y ETOAC
78.450 0.00000 0.00000 71.850 0.44700 0.48700
77.400 0.02480 0.05770 71.800 0.46510 0.49340
77.200 0.03080 0.07060 71.750 0.47550 0.49950
76.800 0.04680 0.10070 71.700 0.51000 0.51090
76.600 0.05350 0.11140 71.700 0.56690 0.53120
76.400 0.06150 0.12450 71.750 0.59650 0.54520
76.200 0.06910 0.13910 71.800 0.62110 0.56520
76.100 0.07340 0.14470 71.900 0.64250 0.58310
75.900 0.08480 0.16330 72.000 0.66950 0.60400
75.600 0.10050 0.18680 72.100 0.68540 0.61690
75.400 0.10930 0.19710 72.300 0.71920 0.64750
75.100 0.12160 0.21380 72.500 0.74510 0.67250
75.000 0.12910 0.22340 72.800 0.77670 0.70200
74.800 0.14370 0.24020 73.000 0.79730 0.72270
74.700 0.14680 0.24470 73.200 0.81940 0.74490
74.500 0.16060 0.26200 73.500 0.83980 0.76610
74.300 0.16880 0.27120 73.700 0.85030 0.77730
74.200 0.17410 0.27800 73.900 0.86340 0.79140
74.100 0.17960 0.28360 74.100 0.87900 0.80740
74.000 0.19920 0.30360 74.300 0.89160 0.82160
73.800 0.20980 0.31430 74.700 0.91540 0.85040
73.700 0.21880 0.32340 75.100 0.93670 0.87980
73.300 0.24970 0.35170 75.300 0.94450 0.89190
73.000 0.27860 0.37810 75.500 0.95260 0.90380
72.700 0.30860 0.40020 75.700 0.96340 0.92080
72.400 0.33770 0.42210 76.000 0.97480 0.93480
72.300 0.35540 0.43310 76.200 0.98430 0.95260
72.000 0.40190 0.46110 76.400 0.99030 0.96860
71.950 0.41840 0.46910 77.150 1.00000 1.00000
71.900 0.42440 0.47300
1. Start Aspen Plus and create a new run.
2. In the Home tab of the ribbon, in the Run Mode group, click Regression.
3. Enter the components on the Components | Specifications | Selection sheet:
32
Note: A complete backup file with results for this example is available in the Aspen Plus Online
Applications Library. The filename is DRS1.
In this example three activity coefficient models will be fitted to the VLE data, each in a separate
case.
Use the Methods | Specifications | Global sheet to choose a property method. This example
compares fitting results for the Wilson, NRTL, and UNIQUAC property methods. Select one of
the three on the Global sheet and the remaining two on theReferenced sheet. In this example,
the Wilson model is chosen on the Global sheet.
Use the Data | Mixture form to enter the vapor liquid equilibrium data. Three data sets are
required, one for each set of VLE data. The following Setup and Data sheets are for the 40C
isothermal data set. On the Setup sheets be sure to add ETOAC first since our data is written in
terms of that component.
33
6. Specify the regression case.
Use the Regression form to formulate a regression case. In this example, Aspen Plus has already
completed the Regression | Input form. Since the WILSON property method is the Global
property method, it is used as the default in the regression. All the VLE data groups you entered
in Step 4 are on this form. Aspen Plus will test the data for thermodynamic consistency, using
the Area test.
34
Since the VLE data cover a wide temperature range, Aspen Plus selects elements 1 and 2 of the
Wilson binary parameters for regression. Aspen Plus uses the databank values for the binary
parameters as initial guesses in the regression.
Use the Regression Object Manager to specify two additional cases. Use the same set of
experimental data, but with the NRTL and UNIQUAC property methods. Again Aspen Plus
completes the Regression | Input form. However, since the WILSON property method is the
global property method, it is the default. Specify NRTL in the Method list box on the Setup
sheet, so that the NRTL property method is used in the second case. When the NRTL property
method is used, the NRTL binary parameters must be regressed. Specify the NRTL binary
parameter elements 1 and 2 as the regression parameters.
35
Repeat this step for the UNIQUAC property method.
Run all three cases. Click OK on the Data Regression Run Selection dialog box. You can also run
selected cases. Move the cases you do not want to run into the Don't Run area, using the left
arrow.
36
9. Examine the results on the Regression | Results form.
Use the Regression | Results | Parameters sheet to examine the final parameter values.
Use the Regression | Results | Sum of Squares sheet to examine the weighted sum of squares
and residual root mean square errors.
37
Use the Regression | Results | Consistency Tests sheet to examine the results of
thermodynamic consistency tests. All data groups passed the Redlich Kister area test.
Use the Regression | Results |Residual sheet to examine the residual for the fit of pressure,
temperature, and composition.
10. On the Home tab of the ribbon, in the Plot group, open the plot gallery and click Residual to
plot the residual of pressure for case VLE-1.
38
You can also plot the residual of other variables.
It is most useful to compare experimental data with calculated results. From the Plot group,
click P-xy to generate a P-xy plot for the first data group.
Add the results from the NRTL and UNIQUAC cases to the plot for WILSON.
39