Professional Documents
Culture Documents
Version 3.2
Page i
©Copyright 2006
OLI Systems, Inc.
All rights reserved.
The enclosed materials are provided to the lessees, selected individuals and agents of OLI Systems,
Inc. The material may not be duplicated or otherwise provided to any entity with out the expressed
permission of OLI Systems, Inc.
Page ii
Contents
Contents .......................................................................................................................... 3
Chapter 1 Introduction to OLI Electrolytes ........................................................................ 4
Overview......................................................................................................................... 4
Understanding Aqueous Electrolyte Systems................................................................. 5
Aqueous Electrolytes Primer ........................................................................................ 11
The OLI Thermodynamic Framework.......................................................................... 13
Other Physical Phases in Equilibrium With the Aqueous Phase .............................. 14
What OLI Provides ....................................................................................................... 15
Range of Applicability of the OLI Model..................................................................... 15
The OLI Databank .................................................................................................... 17
The OLI Engine ........................................................................................................ 18
Chapter 2 Getting Started with Hysys OLI....................................................................... 19
Introduction................................................................................................................... 19
Assumptions.................................................................................................................. 19
Starting HYSYS OLI (HEO) ........................................................................................ 19
Starting a new case.................................................................................................... 19
Entering the Simulation Environment. ..................................................................... 24
Chapter 3 Aqueous Thermodynamics............................................................................... 27
Overview....................................................................................................................... 27
The Equilibrium Constant............................................................................................. 27
Question: ................................................................................................................... 28
Principal Thermodynamic Properties ........................................................................... 29
HKF (Helgeson-Kirkham-Flowers)Equation of State,.................................................. 29
The Helgeson Equation of State ............................................................................... 32
What is the standard State? ....................................................................................... 33
Excess Properties .......................................................................................................... 34
Ionic Strength............................................................................................................ 34
Definition of Aqueous Activity Coefficients............................................................ 35
Long Range Terms.................................................................................................... 37
Short Range Terms ................................................................................................... 38
Modern Formulations................................................................................................ 39
Neutral Species ......................................................................................................... 41
Multiphase Model ......................................................................................................... 43
Solid-Aqueous Equilibrium ...................................................................................... 43
Solid Phase Thermodynamic Properties ................................................................... 44
Mixed EOS Model .................................................................................................... 44
Limitations of the Current OLI Thermodynamic Model .............................................. 46
Scaling Tendencies ....................................................................................................... 46
What is a scaling tendency?...................................................................................... 46
Summary ....................................................................................................................... 49
Chapter 4 Examples of OLI Prediction Accuracy ............................................................ 51
Solubility Prediction ..................................................................................................... 51
Speciation in Sour Water .............................................................................................. 53
Page 3
Chapter 1 Introduction to OLI Electrolytes
Overview
A great many industrial processes cannot be designed and optimized effectively without
comprehensively and accurately addressing electrolyte chemistry and phenomena. The
same statement can be made with regard to many production and environmental problems
as well. Electrolyte chemistry plays an important role in many chemical operations,
including:
OLI has developed a theoretical framework, database, data regression techniques, and
applications software that comprehensively and accurately simulate and predict
electrolyte systems. The OLI electrolyte approach is based on and distinguished by the
following unique elements:
• Complete speciation. The OLI model predicts and considers all of the true
species in solution, and accounts for these in the computations.
• Robust standard state framework. Based on the Helgeson equation of state
and parameter regression and proprietary estimation techniques, the OLI
model provides accurate equilibrium constants and other standard state
properties over the broadest possible aqueous range of conditions.
• Activity coefficients for complex, high ionic strength systems. Based on
the combined work of Bromley, Zemaitis, Meissner, Pitzer, and OLI
technologists, OLI models can predict behavior under real world conditions.
Page 4
• Comprehensive databank. The OLI Databank covers 79 inorganic elements
and their associated compounds and complexes, and over 3000 organic
chemicals. OLI Data Service provides customized coverage of clients’
chemistry and private databanks.
• Thermo-physical properties. OLI has developed unique chemical-physical
based models to compute thermodynamic and transport properties for
complex aqueous environments.
OLI clients have used this unique and powerful electrolyte capability to provide hundreds
of millions of dollars benefit through a host of applications in the oil and gas, chemicals,
government research, paper, metals and mining, pharmaceutical, petroleum, and energy
industries. Example applications are listed in Table 1.
Many chemicals react with water to form one or more new species via
chemical equilibrium reactions. The resulting process is often termed
speciation. A particularly complex system occurs when a single chemical
compound, like ferric chloride (FeCl3) is dissolved in water. This actually
results in 14 different species within the aqueous phase. The list is as follows:
H+, OH-, Cl-, Fe+3, FeCl+2, FeCl2+1, FeCl30, FeCl4-1, FeOH+2, FeOH2+1,
FeOH30, FeOH4-1, H2O, HCl0
Page 5
And, the following independent equilibrium reactions are occurring in the
aqueous phase:
H2O = H+ + OH-
HCl0 = H+ + Cl-
Fe+3 + Cl- = FeCl+2
FeCl+2 + Cl- = FeCl2+
FeCl2+ + Cl- = FeCl30
FeCl30 + Cl- = FeCl4-
Fe+3 + OH- = FeOH+2
FeOH+2 + OH- = FeOH2+
FeOH2+ + OH- = FeOH30
FeOH30 + OH- = FeOH4-
This process of aqueous speciation via reaction, together with the physical
equilibria with other phases can produce results, which are quite counter-
intuitive. For example, one mole of ferric chloride dissolved in water
produces a solution with pH of about 2, making ferric chloride in water a
fairly strong acid. This non-intuitive result occurs because of the reactions,
shown above, wherein OH- combines with Fe+3 via a series of stepwise
reactions. This depletion of hydroxide ions then causes the water dissociation
equilibrium reaction to move to the right liberating more hydrogen ions.
Now, imagine the complexity of not just a single chemical compound in water
but, rather, several or even many. The opportunity for equilibrium reactions
abounds in such systems. One of literally uncountable examples of multi-
component systems is the four-component system water-ammonia-carbon
dioxide-hydrogen sulfide (See Appendix 1 for a more detailed discussion and
comparison of OLI predicted and experimental results for this system). In this
case, the list of species and reactions in the aqueous phase is as follows:
H2O, H+, OH-, CO20, CO3-2, HCO3-, NH30, NH4+, NH2CO2-, H2S0, HS-, S-2
H2O = H+ + OH-
CO20 + H2O = H+ +HCO3-
HCO3- = H+ +CO3-2
NH30 + H2O = NH4+ + OH-
NH2CO2- + H2O = NH4+ + CO3-2
H2S0 = H+ +HS-
HS- = H+ + S-2
If one were to take the standard thermodynamic properties for the VLE
between H2O, CO2, NH3, and H2S in the aqueous and gas phases and ignore
the seven aqueous speciation equilibrium reactions shown above, the errors in
Page 6
predicting VLE would be as high as several orders of magnitude. This is
shown rather dramatically in Figure 1 below, where for some points, errors
exceed three orders of magnitude. Figure 2 illustrates the accuracy of the OLI
prediction based on complete speciation in the aqueous phase, and a robust
thermodynamic framework and supporting databank.
100000
10000
CO2 (60 C)
1000
Calculated
H2S (60 C)
100 NH3 (60 C)
10 CO2 (20 C)
H2S (20 C)
1
NH3 (20 C)
0.1 Diagonal
0.01
0.01 0.1 1 10 100 1000 10000 100000
Experimental
100000
10000
1000
Calculated
CO2 (60 C)
100
H2S (60 C)
10 NH3 (60 C)
CO2 (20 C)
1 H2S (20 C)
NH3 (20 C)
0.1
Diagonal
0.01
0.01 0.1 1 10 100 1000 10000 100000
Experimental
And, it is not just inorganic salts (e.g., FeCl3) and gases (e.g., ammonia),
which dissolve in water to form electrolytes. Many organic acids (e.g., formic
acid) as well as other organic compounds dissociate in water forming ions. In
Page 7
addition, the resulting ions may form new species (called complexes, ion
pairs, or organo-metallic complexes) by combining with metal ions. Many
other organic compounds dissolve in water without dissociating.
For almost all chemical compounds, which dissolve in water, there are limits
to their solubility. And, this solubility varies when other chemical compounds
are present as well. Any attempt to dissolve more than this solubility limit,
results in partitioning of the chemical compound to another physical phase.
At room temperature, for example, this may be a solid phase in the case of
dissolving sodium chloride, a gas phase in the case of carbon dioxide, or a
second liquid phase in the case of benzene.
0
0 1 2 3 4 5 6
At high concentration, interactions of ions and NaCl (moles/kg)
Observed
27.5
concentrations, the formation of complexes, Predicted
27.0
as well as multi-component chemical
interactions make accurate prediction 26.5
Page 8
predictive activity coefficient model that can accommodate these severe non-
idealities. And, the OLI in-place Databank includes coefficients resulting
from thousands of data regressions based on the best available thermodynamic
data for binary and many ternary systems. Figure 4 illustrates the effect of
temperature, pressure, and concentration, and the accuracy of OLI predictions.
Until the culmination of more than 20 years of applied R&D at OLI, no one
had put together a complete, speciation-based model for prediction of the
thermophysical properties of aqueous-based systems over a broad range of
conditions (e.g., -50C to 300C, 0 to 1500 bar, 0 to 30 molal). Key elements of
the OLI model were developed over the past 100 years by Debye and Huckel,
Helgeson, Pitzer, Meissner, and others. However, no one has put all of this
together with key extensions and synthesized a model such as the OLI model.
Thus, the OLI model is not generally known and understood at the university
level.
The plant operators increased the sodium hydroxide concentration from 10%
by weight to 20% by weight. In addition, the Waste Liquid from the scrubber
was desired to be at pH > 8.3 for alkalinity control. The result of the change in
concentration was frequent fouling of the scrubber and an inability to control
the pH.
Using the OLI Electrolytes package, the plant operators found that increasing
the concentration of sodium hydroxide to 20 % by weight did little to increase
the scrubbing efficiency.
Page 9
The amount of chlorine %Cl2 Removal v. Concentration of NaOH
removal seems to be
limited to approximately
%Chlorine Removal
50% at sodium 60.00
hydroxide 40.00
concentrations greater 20.00
than 15% by weight. The
plant operators also 0.00
0.00 10.00 20.00 30.00
noticed that the column
wt% NaOH
fouled frequently at the
high concentrations of
sodium hydroxide.
The off-gas contained carbon dioxide as well as chlorine. Using the OLI
Electrolytes package, the plant operators found that sodium bicarbonate was
created at the increased sodium hydroxide concentrations.
hydroxide, sodium
NaHCO3 solid
6
bicarbonate (a solid)
(moles)
4
begins to form. The
2
plant operators
0
confirmed the presence
0.00 10.00 20.00 30.00
of this solid in their
scrubber. w t% NaOH
Another problem was that the pH could not be maintained at values greater
that 8.3. The OLI Electrolytes package determined the pH response of the
chemistry underlying the process.
4
aqueous chemistry, is
2
due the presence of solid
0
NaHCO3 forming inside 0.00 5.00 10.00 15.00 20.00 25.00 30.00
the scrubber. As long as wt% NaOH
this solid continues to
form, the pH cannot
Page 10
increase.
A typical example is scaling in tower units. The formation of salts can result
in scale formation within a tower, which will reduce a tower's effectiveness
and, eventually, require shutdown of the unit and, perhaps, even the plant.
With the OLI model such salt formation in readily predicted leading to
avoidance by altering operating conditions, the level of use of a reagent, or the
addition of a scale inhibitor all of which can be predicted with the OLI model.
There are five principal partial molal properties in each phase about which we
will be concerned. These are Gibbs Free Energy, Enthalpy, Entropy, Heat
Capacity, and Volume. In particular, we will focus on the aqueous phase for
which the thermodynamic approach is not well understood. The key to
thermodynamic equilibrium, both aqueous intraphase speciation equilibria as well
as the physical equilibria between phases is the Gibbs Free Energy (also called the
Chemical Potential). For every thermodynamic equilibrium reaction the total
Gibbs Free Energy for the species on the left-hand side of the reaction must be
equal to the total Gibbs free energy on the right-hand side of the reaction. Thus:
In this case:
Page 11
The total Gibbs Free Energy can be further defined in terms of a Standard State
Gibbs Free Energy (Go) and an activity (a) as follows:
Where R=Gas Constant and T is absolute temperature. Note that this classical
definition is in terms of an activity rather than a concentration. To be able to
write this in terms of concentration, we will have to introduce the notion of an
activity coefficient, which captures the departure from ideality or, in other words,
the relationship between the activity of a species and the concentration of that
species. Thus,
ai = γi * mi (5)
Please note that the Greek symbol γ is used to denote the activity coefficient and
the concentration unit molality (m) is usually used as the concentration unit for
aqueous systems thermodynamics. Molality is defined as the number of moles of
a species in a kilogram (approximately 55.508 moles) of water. Defining
concentration in this way makes the concentration definition independent of
density and, therefore, temperature.
The standard state used for aqueous-based systems is the one most commonly in
use in the literature. For any species other than water, the standard state is the
concentration of a hypothetical one-molal solution of the species extrapolated to
infinite dilution. This standard state is termed asymmetrical and is a function of
temperature and pressure only. Another way of writing equation (4) is:
Where the superscript E denotes the Excess Gibbs Free Energy and is therefore by
equation (4) defined as RTln(a).
The formidable challenge faced by researchers over the past century is how to
write closed form equations as a function of temperature, pressure and
concentration, which would predict the standard state and excess Gibbs Free
Energy. This challenge reduces essentially to one of how to predict the standard
state Gibbs Free Energy and activity coefficient for every (and, therefore any)
possible species in water.
Before proceeding to the specific framework used by OLI, it is useful to point out
that there are similar expressions to (6) for all five principal partial molal
properties. Specifically, for the other four (Enthalpy, Entropy, Heat Capacity, and
Volume), these expressions are:
H = Ho + HE (7)
S = So + SE (8)
Page 12
Cp = Cpo + CpE (9)
V = Vo + VE (10)
We saw earlier that the excess term for the Gibbs Free Energy was directly related
to the activity coefficient. For the other four properties, denoted above, the excess
terms all relate to various partial derivatives of the activity coefficient. For
example, the excess enthalpy directly relates to the first partial derivative of the
activity coefficient with respect to temperature.
OLI has developed a thermodynamic formulation for equations (6) - (10) above
by utilizing the framework of Helgeson and co-workers for the standard state
terms (the ones with the superscripts of "0") and the frameworks of Bromley,
Zemaitis, Pitzer, Debye-Huckel, and others for the excess terms (the ones with the
superscripts of "E").
The key to the predictive nature of the OLI model is based upon the work of
Harold Helgeson and co-workers. Helgeson, a Professor of Geochemistry at UC
Berkeley, has worked for 40 years to develop a predictive equation of state for the
prediction of the partial molal standard state properties of any species within the
aqueous phase. Co-workers such as Dimitri Sverjensky, now at Johns Hopkins
University in the Department of Earth and Planetary Sciences, have contributed
greatly to provide methods for estimating the coefficients of the equation of state
for species where it is not practical to regress for the required coefficients. It is
beyond the scope of this document to provide the specific formulas for the five
principal standard state terms, however a general description is as follows:
The seven parameters shown in each equation are equation of state coefficients
that are unique to each species. In many cases of interest to chemical engineers,
a1 through a4 can be ignored since these coefficients relate only to the effects of
pressure on the properties of interest and until pressures get above 100
atmospheres the effects of pressure are negligible. OLI has incorporated a
Helgeson framework and estimation methods that allows data banking for the ω,
c1, and c2 for virtually any species in water and, in most cases, a1 through a4 as
well. Remember that equilibrium constants depend solely on the individual G0
values for the constituent species in each reaction. Since OLI, via its Helgeson
framework, can predict these individual G0 values for any species in water, we
can, therefore, predict any equilibrium constant. This fully predictive framework
Page 13
for standard state properties relieves entirely the burden of speciation from the
activity coefficients where, historically, this burden has been placed. Thus, the
remaining compositional effects (called non-ideality) are placed, where they
belong, on the activity coefficients without the added burden of speciation
equilibrium.
For the excess properties, as noted earlier, everything relates to the activity
coefficients. OLI's model for activity coefficients of ions has the following form:
γ = DH(I) + BZ(I, T, m)
where the first term DH is the so-called Debye-Huckel term, which is a function
of I=ionic strength (this is the term that completely describes the activity
coefficient for very dilute systems). The second term is the Bromley-Zemaitis
term and is a function of ionic strength, temperature, and the individual species
concentrations represented by m the vector of species molalities. The ionic
strength is directly related to the various ionic concentrations.
The condition for physical equilibrium between phases is that the total Gibbs Free
Energy (also known as chemical potential) is equal for the phases in equilibrium.
Since we have already established the manner in which OLI calculates Gibbs Free
Energy for the aqueous phase, we need only describe how OLI does this same
calculation for solids, gases, and the second liquid phase.
Each solid in equilibrium with the aqueous phase is an independent phase. Thus,
we need only be able to calculate the Gibbs Free Energy for individual solids.
This is done via classical thermodynamics by the equation:
Where, Tr is the Reference State temperature of 298.15K, and the two integrals
are integrated from Tr to T.
Page 14
For the gas and second liquid phase, OLI relies on an enhanced SRK formulation
for the thermodynamic properties. For the free energy, the formulation is the
classical formulation, which is:
Gi = Gio + RTln(ficiP)
OLI provides all facilities, which enable a user to avoid all of the complexities
associated with aqueous electrolyte systems. This means that the user never has
to:
Essentially, from a user point of view, the description of process streams and
process units is no different than with conventional simulation. For example,
using the highly complex system of H2O/FeCl3, described earlier, the user deals
only with the two component system while the OLI software, behind the scenes
deals with the 14 true species in solution, the 10 equilibrium reactions in the
aqueous phase, and the physical equilibrium between the aqueous phase and any
solid, vapor, and second liquid phase that may occur.
• The temperature resulting from the equilibrium calculation must fall in the
range -50 and 300 centigrade.
• The pressure resulting from the equilibrium calculation must fall in the range
0 and 1500 atmospheres.
• The mole fraction of water within the liquid phase resulting from the
equilibrium calculation must fall in the range between 0.65 and 1.0.
Page 15
• The OLI model used for the gas phase and second liquid phase is based upon
an enhanced SRK equation of state. This means that certain non-ideal second
liquid phases may not be represented well due to limits of the SRK.
• OLI has done data regression for an enormous number of chemical systems.
Generally, all single components plus water have been regressed. Figure 5
illustrates the accuracy of typical binary system predictions at different
temperatures.
Solubility (mmol/kg)
7
binary systems. Figures 6 and 7 illustrate 6
Observed
Predicted
ternary system predictions compared to 5
4
experimental data. (Also see Appendix 1 3
2
for other examples of data fits). Note 1 0.2 m NaCl
Solubility (mmol/kg)
7 Observed
7 Observed
10%. 6 Predicted
5
4
Eventually, for all multi-component 3
2
systems with more than three components 1 1.0 m NaCl
25 oC
0
0 1 2 3 4 5 6 7
NaCl (mole/kg)
0.02
Solubility (mmol/kg)
200 oC
Anhydrite
0.01
Observed
Predicted Page 16
0
0 1 2 3 4 5 6
NaCl (mole/kg)
The OLI Databank
OLI Electrolytes includes the Full Databank and the GeoChem Databank.
• Full databank
• Special databanks
In addition to the above-noted databank, OLI has a few other databanks that
are generally distributed with the product. The Geochem databank (see list in
Appendix 4) is separated from the public since it contains many solids, which
only form via thermodynamic equilibrium after long periods of time (often
thousands of years). There solids form in nature and are often different from
the form of a solid, made up of the same elements, that may form in a much
shorter amount of time (usually instantly) within a process. Thus, the
Geochem solids tend to form through aging of the aqueous environment in
contact with a source of the elements comprising the solids.
The OLI Engine Software allows users to provide their own supplementary
databanks (Private Databanks) in addition to the Public (Limited, Full and
Special) databanks. These databanks can be prepared by users with special
training but are normally prepared by the OLI Data Service. The procedure is
that OLI is informed that certain chemistry either not covered or inadequately
covered by the Full Databank, is required. OLI will then estimate the amount
of time that it will take to do such a job, which leads to a quotation for
performing the work. The fee quoted will vary based upon whether or not
OLI can do the work based upon public domain information and whether or
not OLI is free to release the results with subsequent releases of its Public
Databanks. OLI's Data Service starts all projects with a thorough literature
search. Then all relevant literature is critically reviewed and the experimental
data, which is deemed fit is placed into computer files from which nonlinear
regression is carried out. As a result of these regressions, databank
coefficients are developed.
Page 17
The OLI Engine
The OLI Engine is the essential heart of all OLI Software. The OLI Engine is
defined as the databanks and solvers that enable the prediction and numerical
solution of the underlying chemical and physical equilibria and, therefore, the
thermophysical properties and phase separations for almost any mixture of
chemicals in water at almost any conditions (T = -50 to 300 C, P = 0 to 1500 bar,
Ionic Strength = 0 to 30 Molal) of practical interest to industry. The thermo-
physical properties available in OLI Electrolytes are summarized in Table 2. In
addition to the Thermodynamic Properties in Table 2, OLI has developed certain
transport properties to support the rates of corrosion model. These properties
include: viscosity, electrical conductivity and self-diffusivity.
Thermodynamic Properties
Gibb’s Free Energy pH
Enthalpy Osmotic Pressure
Entropy Ionic Strength
Heat Capacity Density
Volume Scaling Index
Page 18
Chapter 2 Getting Started with Hysys OLI
Introduction
The HYSYS OLI package (commonly referred to as HEO) can be used to study
electrolyte solutions in a process simulator. In this simulation we will mix an acid and a
base stream together and review the output.
Assumptions
The user is expected to have adequate knowledge on starting, stopping and running
Hysys. This includes process design and data entry.
Start > Programs > Hyprotech > Hysys 3.2 > Hysys
Wait for the program to load the splash screen and then start the main window.
Figure 2
Page 19
Click the File > New > Case menu items.
Figure 3
Hysys requires the user to first select a Fluid Package before selecting components. This
is not the first tab displayed but it is the first tab that needs to be selected.
Figure 4
Page 20
Figure 5
Figure 6
You can see that additional options are displayed when this package is selected. You may
now close the window by clicking on the X.
Page 21
Figure 7
Figure 8
Page 22
Figure 9
Figure 10
Enter H2O into the Match Box and press enter. All Electrolyte models require water
(H2O) be selected.
Page 23
Notice that the species listing scrolls to find your species. Now Press Enter.
In this demonstration we will add hydrochloric acid (HCl) and sodium Hydroxide
(NaOH). Enter both these species.
Figure 13
Page 24
Figure 14
Figure 15
We will now create a mix block with an acid feed and a base feed. Create the following
streams
Stream 1
Temperature 25 C
Pressure 101.325 Kpa
Page 25
Total Flow 100 Kg/Hr
H2O 0.9 mole fraction
HCL 0.1 mole Fraction
Stream 2
Temperature 25 C
Pressure 101.325 Kpa
Total Flow 100 Kg/Hr
H2O 0.9 mole fraction
NAOH 0.1 mole Fraction
Double-click Stream 3
Figure 16
Hysys OLI will calculate the streams as they are entered. In this case the results indicate a
temperature rise to 75 .6 C. Save the file as required.
Page 26
Chapter 3 Aqueous Thermodynamics
Overview
Understanding aqueous thermodynamics can be a daunting task. In this chapter
we will describe some of the essential topics in aqueous thermodynamics and
present them in a logical, relatively easy to understand manner.
∆ R G o = − RT ln K
Where ∆ R G is the partial molal, standard-state Gibbs Free Energy of
o
∆ R G = ∑ vi ∆ f Gi ( PRODUCTS ) − ∑ vi ∆ f Gi ( REACTANTS )
i i
Page 27
Question:
∆ f G R ( Na + ) = −261800 J / mole
( )
∆ R G R = 2 ∆ f G R ( Na + ) + ∆ f G R ( SO42 − ) − ∆ f G R ( Na 2 SO4 )
ln K R = − R
RT
By now substituting the appropriate numbers we get:
1
25oC (298.15K) is also know as the reference temperature.
2
NBS Tables of Chemical Thermodynamic Properties - Selected Values for Inorganic
and C1-C2 Organic Substances in SI Units, Wagman,D.D., et al, 1982
3
The subscript f refers the energy of formation from the elements. The superscript R refers to the reference
state. This is a special case of the standard state normally denoted with a superscript o.
Page 28
KR=0.34
Principal Thermodynamic Properties
Each thermodynamic property is composed of two parts. The first is the standard
state part which is only a function of temperature and pressure (denoted by the
superscript o).
The second is the excess part which is a function of temperature and pressure as
well as concentration (denoted by the superscript E).
G =G +G
i i
o
i
E
H =H +H
i i
o
i
E
i S =S +S i
o
i
E
Cp = Cp + Cp
i
o
i i
E
V = V +V
i i
o
i
E
0
Note: Superscript = Standard State Property
Superscript E = Excess Property
Working since 1968, Helgeson, et. al., have found that the standard-state
thermodynamic property of any species in water can be represented by a
function with seven terms which have specific values for each species.
4
H.C.Helgeson, D.H.Kirkham, G.C.Flowers. Theoretical Prediction of the Thermodynamic Behavior of
Aqueous Electrolytes at High Pressures and Temperatures - Parts I through IV. American Journal of
Science 1974, 1976, 1981.
5
J.C.Tanger, IV Doctorial Thesis. “Calculation of the Standard Partial Molal Thermodynamic Properties of
Aqueous Ions and Electrolytes at High Pressures and Temperatures” University of California at Berkley,
1986 H.C.Helgeson Advisor.
Page 29
These seven terms (a1-4, c1-2, and ω) are integration constants for volume (a),
heat capacity (c ) and temperature and pressure properties of water ( ω). They
are independent of the data system used to obtain them.
H io = H iR + f Hi (a1 ,...., a4 , c1 , c2 ,ω )
( )
Gio = GiR − SiR T − T R + f Gi (a1 ,...., a4 , c1 , c2 ,ω )
S io = S iR + f Si (a1 ,...., a4 , c1 , c2 ,ω )
Page 30
Superscript R - Reference State Property (25C, 1 bar)
Superscript o - Standard State Property
a1...a4 - Pressure Effects
c1, c2 - Temperature Effects
ω - Pressure, Temperature Effects
-9
-9.5
-10
-10.5
-11
Log K
-11.5
-12
-12.5
-13
-13.5
-14
0 50 100 150 200 250 300 350
Temperature (C)
Figure 3-1 The logarithm of the equilibrium constant (LOG K) for the dissociation of the
bicarbonate ion as a function of temperature at saturation pressure. The symbols
represent the data taken from the references listed in the footnotes 6,7,8,9,10 but the
line was generated from the equation of state.
6
H.S.Harned and S.R.Scholes. The Ionization Constant of HCO3- from 0 to 50o. J.Am.Chem.Soc. 63,1706
(1941)
7
R.Nasanen. Zur Einwirkung der Saure und Basenzusatze auf die Fallungskurvevon Bariumcarbonate.
Soumen Kemistilehti 90,24 (1946)
8
F. Cuta and F.Strafelda. The second dissociation constant of carbonic acid between 60 and 90oC. Chem.
Listy 48,1308 (1954)
9
B.N.Ryzhenko. Geochemistry International 1,8 (1963)
10
C.S.Patterson, G.H.Slocum, R.H.Busey and R.E.Mesmer. Carbonate equilibrium in hydrothermal
systems: First ionization of carbonic acid in NaCl media to 300oC. Geoch.Cosmoh.Acta 46,1653 (1982)
Page 31
The Helgeson Equation of State
Enthalpy
⎡⎛ 1 ⎞ ⎛ 1 ⎞⎤ ⎛ Ψ+P ⎞
∆ H Po ,T = ∆ H of + c1 (T − Tr ) − c 2 ⎢⎜ ⎟ − ⎜⎜ ⎟⎟⎥ + a1 (P − Pr ) + a 2 ln⎜⎜ ⎟⎟
⎣⎝ T − Θ ⎠ ⎝ Tr − Θ ⎠⎦ ⎝ Ψ + Pr ⎠
⎛ ⎡ Ψ + P ⎤ ⎞ ⎡ 2T − Θ ⎤ ⎛1 ⎞ ⎛ 1 ⎞⎛ ∂ω ⎞
+ ⎜⎜ a3 (P − Pr ) + a 4 ln ⎢ ⎥ ⎟⎟ ⎢ 2 ⎥
+ ω ⎜ − 1⎟ + ωTY − T ⎜ − 1⎟⎜ ⎟
⎝ ⎣ Ψ + P r ⎦ ⎠ ⎣ (T − Θ ) ⎦ ⎝ ε ⎠ ⎝ ε ⎠⎝ ∂T ⎠P
⎛ 1 ⎞
− ω Pr,Tr ⎜⎜ − 1⎟⎟ − ω Pr,Tr Tr Yr
⎝ ε Pr,Tr ⎠
Gibbs Free Energy
⎡ ⎛T ⎞ ⎤ ⎛ Ψ + P ⎞⎛ Ψ + P ⎞
∆GPo,T = ∆Gof − SPr,
o
(T − T ) − c1 ⎢T ln⎜
⎜T ⎟⎟ − T + T ⎥ + a1 ( P − P ) + a ln
2 ⎜
⎜ ⎟⎟⎜⎜ ⎟⎟
Ψ + Ψ +
Tr r r r
⎣ ⎝ r ⎠ ⎦ ⎝ Pr ⎠ ⎝ Pr ⎠
⎡ ⎛ Ψ+P ⎞⎤⎛ 1 ⎞ ⎡⎛ ⎛ 1 ⎞ ⎛ 1 ⎞ ⎞⎛ Θ − T ⎞ T ⎛ Tr (T − Θ ) ⎞⎤
+ ⎢a3 (P − Pr ) + a 4 ln⎜⎜ ⎟⎟⎥⎜ ⎟ − c 2 ⎢⎜⎜ ⎜ ⎟ − ⎜⎜ ⎟⎟ ⎟⎜ ⎟ − 2 ln⎜⎜ ⎟⎟⎥
⎝ T (Tr − Θ ) ⎠⎥⎦
⎟
⎣ ⎝ Ψ + Pr ⎠⎦⎝ T − Θ ⎠ ⎢⎣⎝ ⎝ T − Θ ⎠ ⎝ Tr − Θ ⎠ ⎠⎝ Θ ⎠ Θ
⎛1 ⎞ ⎛ 1 ⎞
+ ω ⎜ − 1⎟ − ω Pr,Tr ⎜⎜ − 1⎟⎟ + ω Pr,Tr YPr,Tr (T − Tr )
⎝ε ⎠ ⎝ ε Pr,Tr ⎠
Volume
⎛ 1 ⎞ ⎡ ⎛ 1 ⎞⎤⎛ 1 ⎞ ⎛ 1 ⎞⎛ ∂ω ⎞
V o = a1 + a 2 ⎜ ⎟ + ⎢ a3 + a 4 ⎜ ⎟⎥⎜ ⎟ − ωQ + ⎜ − 1⎟⎜ ⎟
⎝Ψ+P⎠ ⎣ ⎝ Ψ + P ⎠⎦⎝ T − Θ ⎠ ⎝ ε ⎠⎝ ∂P ⎠T
Entropy
T c 2 ⎧⎛ 1 ⎞ ⎛ 1 ⎞ 1 ⎛ Tr (T − Θ ) ⎞⎫ ⎛ 1 ⎞ ⎡ ⎞⎤
2
⎛ Ψ+P
S =S + c1 ln − ⎨⎜ ⎟ − ⎜⎜ ⎟⎟ + ln⎜⎜ ⎟⎟⎬ + ⎜ ⎟ ⎢a3 (P − Pr ) + a 4 ln⎜⎜ ⎟⎟⎥
o o
Tt Θ ⎩⎝ T − Θ ⎠ ⎝ Tr − Θ ⎠ Θ ⎝ T (Tr − Θ ) ⎠⎭ ⎝ T − Θ ⎠ ⎣
Pr,Tr
⎝ Ψ + Pr ⎠⎦
⎛ 1 ⎞⎛ ∂ω ⎞
+ ωY − ⎜ − 1⎟⎜ ⎟ − ω Pr,Tr YPr,Tr
⎝ε ⎠⎝ ∂T ⎠ P
Where
H = Enthalpy
G = Gibbs Free Energy
Page 32
V = Volume
Cp = Heat Capacity at constant Pressure
S = Entropy
T = Temperature
P = Pressure
Θ = 228 K
Ψ = 2600 Bar
ω = Temperature and Pressure dependent term for electrostatic nature of the
electrolytes
Q = Pressure functions of the dielectric constant
ε = Dielectric constant of water
a1,…,a4 = Pressure dependent terms
c1 , c2 = Temperature dependent terms
Aqueous:
The hypothetical 1.0 molal solution extrapolated from infinite dilution.
P
r
o
p
e
r
t
1.0
Molal
Vapor:
The Ideal Gas Pure Component (mole fraction = 1.0)
Organic Liquid:
The Ideal Gas Pure Component (mole fraction = 1.0)
Solid:
The pure component solid.
11
M.Rafal, J.W.Berthold, N.C.Scrivner and S.L.Grise.”Chapter 7:Models for Electrolyte Solutions”,
Models for Thermodynamic and Phase Equilibria Calculations. Stanley I Sandler, ed. Marcel-Dekker, Inc.
New York: 1994. pp. 686.
Page 33
Excess Properties
Excess properties are a function of temperature, pressure and composition. It is
with the excess properties that we begin to introduce the concept of activities
and activity coefficients.
The excess property that we are most concerned with is the excess Gibbs Free
Energy:
The activity of a species in solution can be defined as:
ai = γ i mi
Gi = G o
+ RT ln ai
G i = G + RT ln mi + RT ln γ i
o
G i = RT ln γ i
E
δ ln γ i ⎤
Hi = RT
E 2
δT ⎥⎦ P
Ionic Strength
Ionic Strength is defined by the following equation:
nI
I = 1/ 2 ∑ (z
i=1
2
i mi )
Page 34
For Example, a 1.0 molal solution of NaCl has 1.0 moles of Na+1 ion
and 1.0 moles of Cl-1 ion per Kg H2O.
I=
1
2
(
( 2 2
) ( 1 2
2
2
)
Z Na +1 ) (m Na +1 ) + (Z Cl −1 ) (mCl +1 ) = (1) (1) + (− 1) (1) = 1
For Example, a 1.0 molal solution of CaCl2 has 1.0 moles of Ca+2 ion
and 2.0 moles of Cl-1 ion per Kg
H2O.
I=
1
2
(
( 2 2
) (
Z Ca+2 ) (mCa+ 2 ) + (Z Cl −1 ) (mCl +1 ) = (2) (1) + (− 1) (2) = 3
1 2
2
2
)
There for the ionic strength is 3.0 molal, or we can say that a 1.0
Molal solution of CaCl2 behaves similar to a 3.0 molal Solution of
NaCl
Definition of Aqueous Activity Coefficients
Page 35
interact with themselves
(oppositely charged species
attract, like charged species
repel) in addition to the
interactions with the solvent.
Page 36
Long Range Terms
− z 2 A(T ) I
ln γ i =
1 + Å B (T ) I
Å ion size parameter
At 25 oC and 1 Atmosphere12:
12
H.C.Helgeson and D.H.Kirkham. American Journal of Science Vol. 274, 1199 (1974)
Page 37
Short Range Terms
no
∑ (b
j=1
ij (T,I) m j )
11 = NH4:HCO3
12 = NH4:CO3
13 = NH4:OH
14 = NH4:NH2CO2
Page 38
Modern Formulations
Bromley
− A | Z+Z− | I (0.06 + 0.6 B ) | Z + Z − | I
Log γ ± = + 2
+ BI
1+ I ⎛ 1 .5 I ⎞
⎜⎜1 + ⎟
⎝ | Z+Z− | ⎟⎠
Where
A = Debye-Huckel Constant
I = Ionic Strength
B = Bromley parameter
γ = Mean activity coefficient
Z+ = Charge of the cation
Z- = Charge of the anion
Bromley-Zematis
Joseph Zematis, formerly of OLI Systems, Inc., now deceased, extended the
work of Bromley in that he added two new terms17. The Bromley-Zematis
activity model is:
13
L.A.Bromley. J.Chem.Thermo.,4,669 (1972)
14
H.P.Meissner. AIChE Symp.Ser.No. 173,74,124 (1978)
15
K.S.Pitzer,et.al. J.Soln.Chem,4,249(1975); J.Phys.Chem.,81,1872(1977); J.Soln.Chem. 7,327(1978);
J.Am.Chem.Soc.96,5701(1974)
16
H.C.Helgeson, D.H.Kirkham and G.C.Flowers. Am.J.Sci. 281,1249(1981)
17
Zemaitis, J.F., Jr, Clark, D.M., Rafal, M. and Scrivner, N.C., Handbook of Aqueous Electrolyte
Thermodynamics, American Institute of Chemical Engineers, New York, 1986.
Page 39
− A | Z + Z − | I (0.06 + 0.6 B ) | Z + Z − | I
Log γ ± = + 2
+ BI + CI 2 + DI 3
1+ I ⎛ 1.5 ⎞
⎜⎜1 + ⎟⎟
⎝ | Z+Z− | I ⎠
Where C and D are new terms. Each of the B, C, and D terms have the following
temperature functionality.
Where
fγ = The “Debye-Huckel” term.18
ν+ = Stoichiometric coefficient for the cation
ν- = Stoichiometric coefficient for the anion
ν = ν+ + ν-
m = Concentration in molal
B±γ = Pitzer B term, containing the adjustable parameters
C ±γ = Pitzer C term, containing adjustable parameters
Helgeson
− Aγ | Z i Z l | I ⎛ω ν i ,k b Y I ν bil Yi I ⎞⎟
Log γ ± = + Γγ + ⎜ k ∑ bk Yk I + ∑l il Ψl + νl ,k ∑i Ψ ⎟
1 + a0 Bγ I ⎜νk νk
⎝ k l k i ⎠
Where
Aγ = Debye-Huckel constant according to Helgeson
Zi = Charge on the cation
Zl = Charge on the anion
Page 40
Γγ = Conversion of molal activity to mole fraction activity
ωk = Electrostatic effects on the solvent due to the species k
νk = moles of electrolyte (summation)
νi,k = moles of cation per mole of electrolyte
νl,k = moles of anion per mole of electrolyte
bi,l = adjustable parameter for the ion-ion interaction.
Yi = fraction of ionic strength on a true basis attributed to the
cation
Yl = fraction of ionic strength on a true basis attributed to the
anion
Ψi = ½ the cation charge
Ψl = ½ the anion charge
Zi = the cation charge
Zl = the anion charge
Neutral Species
Neutral molecules in water are affected by other species in solution. The salting
in and out of a gas is a typical example. When Oxygen is dissolved into pure
water, it has a typical solubility. When salt is added, the solubility decreases.
This is most-likely due to an interaction between the sodium ions and the neutral
oxygen molecule and the interaction between the chloride ions and the neutral
oxygen molecule.
Page 41
Setschenow19
19
J.Setchenow., Z.Physik.Chem., 4,117 (1889)
Page 42
Pitzer20
(a function of temperature)
β0(m-s) = The adjustable parameter for molecule
– ion interactions.
(A function of temperature)
mS = The concentration of the neutral species.
Multiphase Model
Solid-Aqueous Equilibrium
Si = p1 P1 + p2 P2 + ... pp Pp
Examples
NaCl(cr) = Na+ (aq) + Cl-(aq)
CaSO4.2H2O(cr) = Ca+2(aq) + SO4-2(aq) + 2H2O
20
K.S.Pitzer,et.al. J.Soln.Chem,4,249(1975); J.Phys.Chem.,81,1872(1977); J.Soln.Chem. 7,327(1978);
J.Am.Chem.Soc.96,5701(1974)
Page 43
Solid Phase Thermodynamic Properties
T P
Cp = a1 + a2T + a3T-2
Vapor-Aqueous
GV i = G Aqi
a Aqi = γ i mi
f Vi = φVi yi P
Page 44
The reference state for the GVio + RTln( f Vi ) = G oAqi + RTln( a Aqi )
vapor is the ideal gas.
Non-Aqueous Liquid-Aqueous
G Li = G Aqi
Page 45
Limitations of the Current OLI Thermodynamic Model
Aqueous Phase
XH2O > 0.65
-50oC < T < 300oC
0 Atm < P < 1500 Atm
0 < I < 30
Non-aqueous Liquid
Currently no separate activity coefficient Model (i.e., no NRTL,
Unifaq/Uniqac)
Vapor critical parameters (Tc, Pc, Vc, and ω22) are correlated to find a
Fugacity coefficient.
Scaling Tendencies
What is a scaling tendency?
The Ion Activity Product (IAP) is defined as the product of specific ions (in this
case the ions resulting from the dissociation of a particular solid).
IAP = γNamNaγHCO3mHCO3
21
G.Soave.Chem.Eng.Sci.27,1197(1972)
22
This is the acentric factor which is not the same as Helgeson’s ω term)
Page 46
Assuming Ideal Solution Activities
IAP = (1.0)(1.0)(1.0)(1.0)
IAP = 1.0
The Solubility Product (KSP) is the thermodynamic limit of ion availability
Ksp = γNamNaγHCO3mHCO3
KSP = 0.403780
The Scaling Tendency is then the ratio of available ions to the thermodynamic
limit.
ST = IAP/KSP
ST = 1.0/0.403780
ST = 2.48
Was assuming ideal conditions valid??
IAP= (0.598)(0.894)(0.596)(0.866)
IAP=0.276
Page 47
NaHCO3o = 0.101 molal
CO32- = 0.012 molal
NaCO3- = 0.004 molal
Consider Na2CO3/H2O
There are 4 solids of interest in this system. They are:
Page 48
LOG K v Temperature for Na2CO3 solids
LOG K10
-5 LOG K7
LOG K1
Log K
LOG KPPT
Series5
Na2CO3.10H2O Na2CO3.1H2O Na2CO3ppt
Series6
-10
Series7
-15
Na2CO3.7H2O
-20
273 293 313 333 353 373 393 413
Temperature (K)
Figure 3-3 Plot of Log K vs. Temperature. The Log K's have been extrapolated.
It can be seen that the deca-hydrate species does not extrapolate well to high
temperatures. If we concern ourselves with 350K, we can see that if the solid
was allows to be in the model, the equilibrium based solver will attempt include it
over the actual solid which is the mono-hydrate.
Since the deca-hydrate species is outside its temperature range, it will be
mathematically eliminated from the equations.
Summary
We have seen that the evaluation of the equilibrium constant is an important
factor in simulating aqueous systems. A rigorous thermodynamic framework has
been developed to support the simulation
Page 49
Page 50
Chapter 4 Examples of OLI Prediction Accuracy
Solubility Prediction
0
0 1 2 3 4
Barite Data NaCl (mole/kg)
OLI Electrolytes accurately predicts the barite solubility
over broad ranges of temperatures, pressures, and salinities.
5
Solubility (mmol/kg)
25 oC
Solubility (mmoles/kg)
Page 51
Calcite Solubility versus Temperature
The following graphs illustrate, the experimental and
predicted solubility of calcite at high temperatures,
pressures, and NaCl concentrations. As shown, the
software accurately predicts calcite solubility over all
conditions. This is due to the rigorous approach to
computing the thermodynamic solubility of CaCO3 and
the activity coefficients for the complete speciation of
the ions in solution.
2 25 oC
1 0.2 m NaCl Gypsum
0.04
0
100 150 200 250 300
0.02 Observed
8 Predicted
Solubility (mmol/kg)
7 Observed 0
6 Predicted 0 1 2 3 4 5 6 7
5
NaCl (mole/kg)
4
3 0.02
Solubility (mmol/kg)
2 200 oC
0.5 m NaCl
1
0 Anhydrite
100 150 200 250 300 0.01
Temperature (C) Observed
8 Predicted
Solubility (mmol/kg)
7 Observed 0
6 Predicted 0 1 2 3 4 5 6
5 NaCl (mole/kg)
4
3
2
1 1.0 m NaCl
0
100 150 200 250 300 350
Temperature (C) Page 52
Speciation in Sour Water
For the sour water system (NH3/H2S/CO2/H2O) with Sour Water Species as follows:
Aqueous Ionic Species: H+, OH-, NH4+, HS-, S2-, HCO3-, CO32-, and NH2CO2-
Aqueous Chemical Reactions (Electrolyte Equilibrium) – Note additional electrolyte reactions are
needed.
H2O = H+ + OH-
NH3(aq) + H2O = NH4+ + OH-
CO2(aq) + H2O = H+ + HCO3-
-
HCO3 = H+ + CO32-
-
NH2CO2 + H2O = NH3(aq) + HCO3- (Notice that this is the hydrolysis of an ion)
H2S(aq) = H+ + HS-
-
HS = H+ + S2-
This example illustrates both the importance of considering the aqueous reactions and speciation, as well as
the agreement of OLI predictions with experimental data. The data contained in Table 1 compares the
experimental partial pressures of carbon dioxide, ammonia and hydrogen sulfide against a VLE only model and
the full OLI speciated model.
Liquid Concentration
(molality) Partial Pressure (mmHg)
NH3 CO2 H2S
Temp
(C) NH3 CO2 H2S Exper. OLI VLE Exper. OLI VLE Exper. OLI VLE
Page 53
1.238 0.712 0.203 1.292 1.064 12.16 19 18.62 14561.6 15.276 12.464 1345.2
1.234 0.725 0.199 0.912 0.988 12.16 20.444 20.672 14835.2 15.96 13.072 1079.2
1.235 0.771 0.2 0.912 0.836 12.16 29.184 30.324 15808 27.36 17.024 1333.04
1.126 0.794 0.095 0.684 0.684 11.4 35.188 36.252 16271.6 12.236 8.892 633.08
100000
10000
1000 CO2 (60 C)
Calculated
H2S (60 C)
100
NH3 (60 C)
10 CO2 (20 C)
1 H2S (20 C)
NH3 (20 C)
0.1
Diagonal
0.01
0.01 0.1 1 10 100 1000 10000 100000
Experimental
Figure 1 - Parity plot of the partial pressures of gases without aqueous reactions
In Figure 1, the calculated partial pressures of the gases are over predicted. In the case of ammonia (filled
and open diamonds), the over prediction may be as much a five orders of magnitude. If the OLI full
speciation approach is not followed, this data can be improved with statistical corrections or by applying a
strong activity model. These corrections, however, may work for this set of conditions. But if conditions
other than the conditions in these series of calculations are desired, then these “corrections” will not be
valid. The OLI approach using full speciation is predictive over the entire range of OLI model conditions.
Page 54
Partial Pressure of Gases (Full OLI Model)
100000
10000
1000
Calculated
CO2 (60 C)
100
H2S (60 C)
10 NH3 (60 C)
CO2 (20 C)
1 H2S (20 C)
NH3 (20 C)
0.1
Diagonal
0.01
0.01 0.1 1 10 100 1000 10000 100000
Experimental
Figure 2 - Parity plot of the partial pressures of gases with aqueous reactions
When all the equilibria are included in the calculations, the calculated partial pressure of the gases agrees
with the experimental values (Figure 2). This was done without specialized data regression to the general
range of this data. This provides confidence that the predictions will hold at other conditions.
Page 55