United States
            Environmental Protection
            Agency
           Robert S. Kerr Environmental Research
           Laboratory
           Ada OK 74820
EPA-600/2-80-067
April 1980
            Research and Development
&EPA
A New Correlation of
NH3,  COa, and H2S
Volatility Data from
Aqueous Sour Water
Systems

-------
                RESEARCH REPORTING SERIES

Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology.  Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:

      1   Environmental  Health  Effects Research
      2.  Environmental  Protection Technology
      3.  Ecological Research
     ' 4.  Environmental  Monitoring
      5.  Socioeconomic Environmental Studies
      6.  Scientific and Technical Assessment Reports (STAR)
      7.  Interagency Energy-Environment Research and Development
      8.  "Special" Reports
      9.  Miscellaneous Reports

This report has been assigned  to the ENVIRONMENTAL PROTECTION TECH-
NOLOGY series. This series describes research performed to develop and dem-
onstrate instrumentation, equipment, and methodology to repair or prevent en-
vironmental degradation from point and non-point sources of pollution. This work
provides the new or improved technology required for the control and treatment
of pollution sources to meet environmental quality standards.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.

-------
                                                   EPA-600/2-80-067
                                                   April 1980
         A NEW CORRELATION OF NH,, C09, AND H9S
    VOLATILITY DATA FROM AQUEOUS JSOUITWATER SYSTEMS
                            by

                    Grant M. Wilson
               Thermochemical Institute
          and Chemical Engineering Department
               Brigham Young University
                  Provo, Utah  84602
               EPA Grant no. R804364010
                    Project Officer

                    Fred fL Pfeffer
              Source Management Branch
 Robert S. Kerr Environmental Research Laboratory
               Ada, Oklahoma  74820
                   Sponsored by the
             American Petroleum Institute
      Committee on Refinery Environmental Control
ROBERT S. KERR ENVIRONMENTAL RESEARCH LABORATORY
       OFFICE OF RESEARCH AND DEVELOPMENT
      U.S. ENVIRONMENTAL PROTECTION AGENCY
              ADA, OKLAHOMA  74820

-------
                                 DISCLAIMER
     This report has been reviewed by the Robert S. Kerr Environmental
Research Laboratory, U.S. Environmental Protection Agency, and approved for
publication.  Approval does not signify that the contents necessarily reflect
the views and policies of the U.S. Environmental Protection Agency, nor does
mention of trade names or commercial products constitute endorsement or
recommendation for use.

-------
                                  FOREWORD


     The Environmental Protection Agency was established to coordinate admin-
istration of the major Federal programs designed to >protect the quality of our
environment.

     An important part of the Agency's effort involves the search for infor-
mation about environmental problems, management techniques and new technologies
through which optimum use of the nation's land and water resources can be
assured and the threat pollution poses to the welfare of the American people
can be minimized.

     EPA's Office of Research and Development conducts this search through a
nationwide network of research facilities.

     As one of these facilities, the Robert S. Kerr-Environmental Research
Laboratory is responsible for the management of programs to:  (a) investigate
the nature, transport, fate and management of pollutants in ground water;
(b) develop and demonstrate methods for treating wastewaters with soil and
other natural systems; (c) develop and demonstrate pollution control tech-
nologies for irrigation return flows; (d) develop and demonstrate pollution
control technologies for animal production wastes;  (e) develop and demonstrate
technologies to prevent, control, or abate pollution from the petroleum re-
fining and petrochemical industries; and (f) develop and demonstrate technolo-
gies to manage pollution resulting from combinations of industrial wastewaters
or industrial/municipal wastewaters.

     The use of inplant processes to remove undesirable components of a
wastewater stream prior to discharge to a wastewater treatment plant can often
effect significant improvements in treatment plant effluent quality.  This
report contains the findings of a study to utilize new correlations between
sour water constituents so as to improve the ammonia removal efficiency of
sour water scrubbers in petroleum refineries.
                                        W. C. Galegar
                                          Director
                      Robert S. Kerr Environmental Research Laboratory
                                      111

-------
                                  ABSTRACT


     A new correlation model has been developed for calculating sour water
equilibrium data at temperatures from 20 C to 140 C.  The correlating
equations in this new sour water equilibrium model (SWEQ) have been used to
obtain a computer program capable of handling the various chemical and physical
equilibria of NH~, C0?, and H?S in sour water systems including the effects
of carboxylic acfds on ammonia (NhL), Carbon Dioxide (C02), and Hydrogen
Sulfide  (HpS) in sour water systems including the effects of carboxylic acids
on ammonia fixation and release by caustic addition.

     This new SNEQ correlation model has been used to evaluate published and
new vapor-liquid equilibrium data, and comparisons are made with the Van
Krevelen prediction equations as published by Van Krevelen.  Average errors
between  calculated and measured partial pressure data can be summarized.

     Both models predict low temperature data quite well, but at high temp-
eratures the Van Krevelen model deviates considerably from measured data,
and errors between the SWEQ model and measured data increase from about 11%
to about 29%.  Comparisons with variations of the Van Krevelen model as
published by other authors have not been made.

     The basic NFL-FLS-H^O equilibrium program has been inserted into a tray
by tray  program by CoNOCQ.  Two brief example problems have been run to date.
The calculated stream requirements appear to be approximately 30 percent
greater  for a refluxed tower and 20 percent more for a non-refluxed unit com-
pared to Van Krevelen - Beychok procedures.  Definite conclusions cannot be
drawn until wider user experience is obtained.

     Details of the SWEQ correlation model, correlating equations, the com-
puter program, and evaluations of experimental data are given in this report.
This report covers a period from March 15,"1976, to March 17, 1977, and work
was completed as of November 30, 1977.
                                     IV

-------
                               CONTENTS
Foreword	  .  .    iii
Abstract	     1y
Figures	     vi
Tables	    vii
Abbreviations and Symbols  ... 	

     1.  Introduction  	      1
     2.  Project Objectives. ,  	      2
     3.  The SWEQ Model	      3
     4.  Computer Program Based on the SWEQ Model	     20
     5.  Sample Problem Using the SWEQ Model	     51
     6.  Comparisons and Evaluations Between Calculated and
           Measured Data	     59
              Evaluation of Van Krevelen Model  	     59
              Evaluation of SWEQ Model	     75
              Evaluation of New BYU Data	     85
              Ammonia Fixation by Acids and Release by Caustic
                Addition	     86

     7.  Accuracy of Correlation	     91
     8.  Summary	     92

References	     94
Appendix   	     97

-------
FIGURES
Number
1
2
3
4
5
6
7
8
9
10
11
12
13
14

Sample plot of the volatility of NH3> C02, and H2S versus



Flow diagram of option 3 of SWEQ computer program ......
Ammonia mean ratio of measured over calculated partial
Carbon dioxide mean ratio of measured over calculated partial
pressures based on SWEQ correlation 	
Hydrogen sulfide mean ratio of measured over calculated
partial pressure based on SWEQ correlation 	
Ammonia mean ratio of measured over calculated partial
pressures based on Van Krevelen correlation 	
Carbon dioxide mean ratio of measured over calculated partial
pressures based on Van Krevelen correlation 	
Hydrogen sulfide mean ratio of measured over calculated
partial pressures based on Van Krevelen correlation ....
Free ammonia versus pH adjustment by caustic addition at 25°C
Free ammonia versus pH adjustment by caustic addition at 80°C
Sample plot of pH versus caustic addition showing variation of
oH at 25 C and at column temperature 	
Page
19
23
24
25
26
79
80
81
82
83
84
87
88
90
  VI

-------
TABLES
Number
1
2

3

4

5
6
7
8
9
10
11
12
13
14
15
16
17


Summary of Equations Used to Calculate Temperature and
Composition Effects on Henry's Law Relations ' 	
Summary of Chemical Equilibria Involved in Calculating
NH3-C02-H2S-H20 Vapor-liquid Equilibria 	
Effect of Composition and Ionic Strength on Chemical
Equilibrium Constants ' 	
Summary of Temperature Parameters, Used to Calculate Chemical
Equilibrium Constants in Table 1 	

Subroutine KREAC 	
Subroutine HENRY 	
Subroutine YFX 	
Subroutine CFX 	
Subroutine PHST 	
Subroutine SPECV 	
Subroutine NPH 	 , 	
Subroutine SPECL 	
Subroutine PRESY 	
Subroutine NTEMP 	
Input data for Sample Problem with SWEQ Computer Program . . .
Computer Output from Data in Table 16 with Computer Program
Based on the SWEQ Model 	
Page
6

12

14

17
27
32
33
33
34
34
35
37
38
39
39
48

50
  vii

-------
Number                                                                  Page

  IS      H?S-hLO, NH^-H?S-H?0, and NH--H20 Systems, Comparison of
            Calculated and Measured7Data of Miles and Wilson  and
            of Clifford and Hunter17	      60

  19      NH3-H?S-H?0 System, Comparison of Calculated and Measured
            Data of Terres29	      61

  20      NH2-H?S-H?0 System, Comparison of Calculated and Measured
            Data of Van Krevelen, et al.1	      62

  21      H?S in Aqueous Buffer Solutions, Comparison of Calculated
            and Measured Data of Shih, et al.2B	      63

  22      NH3-C02-H20 and C02-H20 System, Comparison of Calculated
            and Measured Data of Van Krevelen, et al J and Data From
            Lange's Handbook20  	      64
            1--CO?-H?0 System, Comparison of Calculated and Measured
            Data of Otsaka, et al.22	
23      NH^-C
                                                                       65
  24      NH^-COp-HpS-HpO System, Comparison of Calculated and
            Measures Data of Cardon and Wilson	      66

  25      NHo-C02-H2S-H20 System, Comparison of Calculated and
            Measurea Data of Badger and Silver^	      67

  26      NH-j-hLO System, Comparison of Calculated and Measured
            Data of Breitenbach and Permanl6,23	      68

  27      NH.,-C02-H2S-H20 System, Comparison of Calculated and
            Measurea Data of Van Krevelen, et al.'	      69

  28      Summary of References of Experimental Data	      70

  29      Summary of Deviation Errors Between Calculated and
            Measured Ammonia Partial Pressures  	      72

  30      Summary of Deviation Errors Between Calculated and
            Measured Carbon Dioxide Partial Pressures 	      73

  31       Summary of Deviation Errors Between Calculated and
            Measured Hydrogen Sulfide Partial Pressures 	      74


  Al       Computer Program Used for Calculating Vapor-Liquid
            Equilibrium Data From the Van Krevelen Correlation  . .      99
                                    vm

-------
                                   SECTION 1

                                 INTRODUCTION


     Previous design calculations of vapor-liquid equilibrium compositions
in sour water strippers have primarily been based on a correlation by Van
Krevelen (1) as outlined in Aqueous Wastes from Petroleum and Petrochemical
Plants by M.R. Beychok (2).  The Van Krevelen correlation has proved suffi-
ciently reliable and many sour water strippers have been designed and built
using his correlation as a basis.  New vapor-liquid equilibrium measurements
have been made since Van Krevelen's correlation published in 1949 including
new measurements at Brigham Young University sponsored by the API Technical
Data Committee.  Although used considerably,  the Van Krevelen correlation has
been previously recognized to be deficient in the following areas:

     1.  Only data to 60°C were correlated; thus the use of the correlation
         at sour-water stripper temperatures  of 100 to 120 C represented an
         extrapolation of existing data.

     2.  The calculation method outlined by Van Krevelen did not allow for
         mixtures containing ammonia over hydrogen sulfide ratios less than
         1.5 in the liquid phase.

     3.  The calculation did not take into account reduced volatilities of
         hydrogen sulfide and ammonia at low parts per million concentrations
         due to the ionization constants of the two compounds in water.

     Subsequent sections of this report give details of a new sour water
equilibrium model (SWEQ) which is based on new higher temperature data and
which avoids deficiencies mentioned above.  This new correlation model also
permits the addition of caustic for release of NH3 held by carboxylic
acids or stronger acids.

-------
                                   SECTION 2

                              PROJECT OBJECTIVES


     The development of a new correlation for ammonia, carbon dioxide, and
 hydrogen sulfide volatilities from aqueous sour water systems has required
 the completion of the following project objectives:

     1.  Compare new NhL-HpS-HpO experimental vapor-liquid equilibrium data
         developed by Bfignam Young University with previously published data
         by Van Krevelen.

     2.  Check and "fine tune" (if necessary) the new vapor-liquid equilibrium
         equations developed by Brigham Young University to the measured
         experimental data.

     3.  Compare BYU equations to equilibrium expressions previously published
         by Van Krevelen and Beychok.

     4.  Modify the BYU equilibrium equations to allow calculations with or
         without external pH adjustment (i.e., using caustic).

     5.  Modify the existing BYU computer program to allow equilibrium
         calculations with or without adjustment.

     These objectives have been achieved by first developing a correlation
model in which literature data of Van Krevelen, new BYU data, and other liter-
ature data have been used to develop equations capable of predicting data
over wide ranges in concentration and temperature.  Based on these equations
a new sour water equilibrium computer program has been developed which is
capable of handling the various chemical and physical equilibria of sour
water systems including the effects of carboxylic acids or stronger acids on
        fixation and release by caustic addition.
     This new sour water equilibrium correlation has now been used to evaluate
published and new vapor-liquid equilibrium data.  Details of the correlating
equations, computer program, and data evaluations are given in subsequent
sections of this report.

-------
                                   SECTION 3

                                THE SWEQ MODEL


     The SWEQ correlation model developed from this project is very similar
to the model used by Van Krevelen (1) except that some of the limitations
imposed by that model have been removed.  Van Krevelen assumed that FLS and
COo only exist in aqueous solutions as ionized species.  This is virtually
true at concentrations where NHL is in excess, but such an assumption would
not be true when these acid gases are present in the absence of NH- or other
basic components.  The method used here, therefore, avoids this problem by
considering the chemical equilibrium between ionic species of H?S or C02 and
undissociated H9S or C09 in the liquid as follows.    ,

                      -    +              k - (HS"} (" }           (1)
           H2SU) "• HS  +H                      (H2S)             UJ
           H2C03(£) + HC03- + H+          k -     3-          (2)
The SWEQ model not not take into consideration the equilibrium between dissolved
C02 and carbonic acid (H2C03) according to the following reaction


           C02 + H20 * HC03                                        (3)


because the presence of other acidic or basic component does not affect this
equilibrium.  This reaction is apparently slow enough that the kinetics of
absorption of C02 into basic aqueous solutions is slower than for H2S.  In
spite of this slower reaction rate, the assumption is made here that sufficient
contact time or catalyst is used to achieve chemical equilibrium.  '  By this
method, the partial pressure of H2S or C02 in the vapor phase above a solution
can be calculated from the concentrations of the undissociated species as
follows.
      'Because of the slower absorption of C02 into water and because of the
possibly slow conversion of bicarbonate ion to carbonate ion by excess ammonia,
a warning is given that actual plate efficiencies could be low compared to
expected efficiencies when C02 is present^

-------
           PH2S = H^


           PC02 = HC02 CH2C03                            (5)
where
             H2S> PC02   = partial pressure of H2$ or C02


            HH2S, HCQ    = Henry's constants for H2$ and C02



              H2S> CH,,CO-  = liquid phase concentrations of H2S and
                       6    H2C03' mo1es/K9 of solution

 The  Henry's  constant used here must apply at finite concentrations as well
 as infinitely dilute concentrations, so, in general, Hn2S and Hco2 become
 dependent on the composition of the solution.  This method of calculating
 HpS  and  CCL  partial pressures is analogous to Van Krevelen's method for
 calculating  ammonia partial pressures which a composition dependent Henry's
 constant is  used.  The addition of Henry's constants and undissociated H2$
 or H^CO., species concentrations makes possible the calculation of vapor-
 liqufd equilibria at acid gas concentrations in excess of ammonia or of other
 basic components; thus the Van Krevelen restriction to compositions with
 excess ammonia  is avoided.

     This method for calculating vapor-liquid equilibrium data under condi-
 tions of simultaneous chemical equilibrium requires two properties that
 must be  correlated in terms of analytical equations as follows.

     1.  Analytical equations for the effect of temperature and composition
         on  Henry's Law constants so that component partial pressures in
         the vapor phase can be calculated from calculated concentrations
         of  undissociated NH3, C02, and H2S in the liquid phase.

     2.  Analytical equations for the effect of temperature and composition
         on  chemical equilibrium constants so that the concentrations of
         undissociated NH3> C02, and H2$ in the liquid can be calculated.

     Rather  than do an exhaustive recorrelation of existing literature data
for these properties, an attempt has been made to use existing correlations
where possible.  Modifications to these existing correlations have been made
when necessary to improve the representation of multicomponent data studied
in this project.  Fortunately, the Henry's constants for NH3, C02, and H?S
can be based  primarily on binary data in water.  This simplifies the corre-
lation because these properties are fairly well known.  Multi-component
vapor-liquid  data thus serve primarily to establish the effects of high
concentrations of the various compounds in solution on these Henry's constants,

-------
By this method, the Henry's constants for ammonia and carbon dioxide at low
concentrations of each compound have been taken directly from the literature.

     Van Krevelen's correlation was made in terms of component concentrations
in moles per liter (£") of solution or molarity.  This method introduces an
unnecessary variable which is the density of the solution.  This occurs
because the density is needed to calculate the molarity when the number of
moles or number of pounds or grams of each component in a mixture are speci-
fied.  The SWEQ model avoids this problem by using concentrations in moles/Kg
of solution.  At low concentrations of the solutes the density of the solu-
tion is about one, so the low concentration parameters of Van "Krevelen's
correlation still apply.  However at conditions where the density deviates
significantly from unity, then parameters in the two correlations cannot be
directly compared.  At these conditions, the parameters in the SWEQ model
have been determined by directly fitting available phase equilibrium data
using concentrations in moles/Kg of solution.  By this method there is no
ambiguity in the correlation because concentrations in moles/Kg of solution
have only been used in the correlation, and the method avoids the need for
density at the various concentrations and temperatures of the correlation.
At low concentrations of the components, published Henry's constants and
chemical equilibrium constants have been used in units of moles/Kg of water
because the two sets of units are the same at the zero concentration unit.

     Table 1 summarizes the various equations used in the SWEQ model for
calculating Henry's constants for NH3, C02, and H2S.  The Henry's constant
for ammonia at low ammonia concentrations has been taken directly from the
equation of Edwards, Newman, and Prausnitz (3) rather than from Van Krevelen
because their correlation is more recent and includes data which was not
available to Van Krevelen.  Exisiting literature data for the volatility of
ammonia from aqueous solutions scatters considerably, but the equation of
Edwards ert aj_ appears to correlate the data of greated precision.

     The Van Krevelen model does not require Henry's constants for CCL and
HpS, so these have been obtained from another source.  Kent and Eisenoerg
have recently published correlations (4) on H2S and CCL partial pressures
from aqueous monoethanol amine and diethanol amine solutions which appear
to correlate these systems quite well.  In their correlation they adjusted
the amine equilibrium constant for reaction with hydrogen ions to obtain
agreement with published data on H2S and CCL partial pressures.  By this
method they obtained a model capable of accurately predicting equilibrium
in HoS-CCL-amine systems.  Their equations for the Henry's constant for CCL
has Been used without any changes as it is given in Table.!.  Their Henry's
constant for H?S however was increased about 12% in order to improve the
represenation of multicomponent data by a change in the first constant as
noted at the bottom of Table 1.

     The use of Henry's constants to correlate volatility data introduces
two methods for calculating concentration effects.  One method is to assume
that the Henry's constant varies with the concentration of the various com-
pounds in solution, and the other method is to assume the various compounds
in solution.  In some cases, the choice of a concentration parameter in the
Henry's constant or of using a concentration parameter in the equilibrium

                                      5

-------
Water
                  TABLE 1.   SUMMARY  OF EQUATIONS  USED TO  CALCULATE  TEMPERATURE
                              AND COMPOSITION EFFECTS ON HENRY'S LAW  RELATIONS*)

Compound
Ammonia
Carbon
Dioxide
Hydrogen
Sulfide
Lit.
Ref.
3
4
4
Fortran
Symbol
HA
HC
HS
Equation '
ln(HA) « 178.339 - 15517. 91/T - 25.6767
+ (.06)(2CC + CS)
ln(HC) = 18.33 - 24895. 1/T + .223996 X
ln(HS) * 100.684* - 246254/T + 2.39029
ln(T) +
108/T2 -
X 108/T2
.019660T + (131.4/T
.090918 X 101]/T3 +
- 1.01898 X 1011/!3
- .1682) (CAS)
.12601 X 1013/T4

HW
ln(HW)
   + 1.59734  X 1013/T4 - .05(CAS) + (.965 - 486/T)(CC)
c^ = 14.466 - 6996.6/(T-77.67)
"' T » temperature in  °R
   CAS » free ammonia  concentration, gram-moles/Kg of solution.
   CC a total C02  1n solution, gram moles/Kg of  solution.
   CS • total HgS  in solution, gram-moles/Kg of  solution.
 ' Henry's constant in psia/(gram-moles/Kg of solution).
c' Water vapor pressure in psla; the partial pressure in water 1s calculated from Raoult's Law
 Constant adjusted from 100.573 to 100.684 in order to fit new H2S solubility data; and
 multicomponent NH,-CO«-H9S-H«0 data.

-------
constant has been arbitrary.  A summary of methods used in the SWEQ model is
given in the following.
Compound
Affected

   NH
   C0
   H2S
  Concentration Effects on
Volatility Data Correlated by
Henry's Const.   Equil. Const.

free NH
absorbed C02

and H2S
free NH
                 absorbed H2S

                 ionic strength
absorbed C02     absorbed C02
                                                      Principal  Data
                                                        Correlated

                                                      NH3-H20

                                                      H2S-C02-NH3-H20
                                                      H2$-C02-NH3-H20

                                                      C02-NH3-H20

                                                      H2S-NH3-H20

                                                      H2S-C02-NH3-H20
     Van Krevelen used a Henry's constant for ammomia which he assumed to
be only dependent on free ammonia concentration.  Additional effects of ab-
sorbed H2S and CCL were found necessary in the SWEQ model in order to corre-
late more recent RpS-COp-NHo^O, so any concentration effects for these
compounds in his model correlated in the equilibrium constant.

     The effects of free ammonia, and of absorbed H2S or CCL on the Henry's
constants used in the SWEQ model are given in Table 1.  In this table, the
Henry's constant of ammonia is proportional to a constant times CAS, free
NH3, and to a constant times'(2 CC + CS), absorbed C02 and H2S.  No concen-
tration effects were introduced in the SWEQ model on the Henry's constant of
C0?, but effects for free ammonia and absorbed C00 were introduced to corre-
late H9S volatility data as shown by terms proportional to CAS and CC in
Table f.

     An equation for water is also given in Table 1.  Water generally exists
as the principal component even in concentrated solutions of electrolytes so
that liquid-phase non-ideality effects on the partial pressure of water are
small.  For this reason, the partial pressure of water in the vapor phase can
be calculated from Raoult's Law where the moles of each ionized and unionized
species in solution is considered in calculating the mole fraction of water.
The partial pressure of water is then calculated from its vapor pressure
according to the following equation.
                    = p
                  H20
          H0
                                                                 (6)

-------
where

             o
            p H20 = vapor pressure of water


            XH20  = liquid phase mole fraction of water


By  this method, the constants for water in Table 1 are simply the vapor
pressure  of water fitted over the range from 25°C to 1500C from data in the
steam  tables.

     No attempt has been made in the SWEQ model to correct for non-ideal
behavior  in the vapor phase.  At low pressures, errors from assuming ideality
are probably less than +_ 5%; but at pressures of 50 psia or higher, the
errors will be greater than this and serious consideration should be made
to  correct for non-ideal in the vapor phase.

     Besides Henry's constants, one must correlate the chemical equilibria
of  reactions occurring in the liquid phase as mentioned above.  The Van Kre-
velen  correlation is limited because the effects of other acidic or basic
components cannot be readily taken into account.  This problem is avoided
in  the SWEQ model by assuming that the various chemical equilibria are depen-
dent on the concentrations of either the ionized or undissociated species of
a component and the hydrogen ion concentration.  For an acid, the general
form of the equilibrium equation is as follows:

            AH + A" + H+                    k = (
while for a base the equilibrium can be written as follows:

            BOH + H+ -> H+ * B+ + H,0 .         k = ID(^]U+,           (8)
In principle, the assumption of equilibria according to these equations makes
possible the calculation of the equilibrium species concentration of each
component knowing only the total concentration of that component and the pH.
If the pH is not known it can be calculated by trial and error until electrical
neutrality is achieved in a given mixture of compounds.  This method of
calculation permits the development of generalized calculation methods, so
that new compounds can be added as necessary.  In many respects the method
is similar to an equilibrium flash calculation where the feed composition and
equilibrium K-values of individual components are known.  In a flash calcu-
lation, the concentrations of each component in the vapor and liquid phase
is known.  But in general, the fraction as vapor or liquid is not kftown so
an iterative calculation is made until the concentrations in each phase add
to 100%.  For acid-base equilibria, the problem is nearly as simple except
that the iteration parameter is pH instead of fraction as vapor or liquid.
The picture for H^S and COp is slightly more complicated because both com-
ponents have second ionization constants so that two chemical reactions must
be simultaneously solved at a given pH value.  In this case, a calculation
example is given as follows.

                                      8

-------
              AH" + H+                     k  =  AH                 (9)
        AH  -> A  + H                       k2 =   (AH-}             (10)
     To solve these equations, it is assumed that the total concentration
of both ionized plus undissociated species concentrations is known by
chemical analysis; but that the concentration of each individual species is
not known.  In this case, the concentrations of individual species can be
related by the following equations.

        (AH2) = nA -a-3                                             (11)

        (AH~) = a                                                   (12)


        (A=) = 6                                                    (13)


From these equations, the following equations are obtained for k-, and k2.

        k  _  >)(H+j
        kl ~  (nT-a-B)                                               (14)
         2       (a)

These simultaneous equations can be algebraically solved for a and 3 to
give the following equations.
                         k1k2/(H+)                                  (16)
If no second ionization occurs, then k2 is zero and a becomes as follows*
        a =
                    k]                                              (18)
After a and 3 have been calculated from equations 16 and 17, then the con-
centration of undissociated species can be calculated from equation 11.
Because of computer round-off error due to subtracting two large numbers to
get a small number, it has been found better to calculate the undissociated
species concentration from equation 14 instead of equation 11.  This is done
by rearranging equation 14 to the following equation in which the round-off
error ia avoided.

          nA-a-3= (AH2) = (

-------
     The calculation of chemical equilbria in mixtures containing both
ammonia and carbon dioxide requires allowance for the reaction of bicarbamate
ion with free ammonia to produce carbamatc ion as follows.
           HC03" + NH3 -> H2NCOO" + H20


                       i, - (MCOQ")
                       K " (HC03 )("H3}

 This  introduces a third simultaneous reaction for CCL and a second simultan-
 eous  reaction for NH~.  This added complexity makes necessary a second itera
 tive  calculation procedure to calculate individual species concentrations at
 a  specified pH value.  This calculation is made by assuming various bicarbo-
 ate concentrations from which the concentrations of the other species can be
 algebraically calculated.  The resulting concentrations of individual CCL
 species  are then added compared with the specified moles of C02 in the solu-
 tion  as  follows.
          n
           CO,
= (C02) + (KC03~)  + (C03~)  +  (H2NCOO~)          (21)
             'Z(calc)

The amount of carbonate is then adjusted up or down by the following ratio.

                                    (nrn )
           (HC03  ^    = (HC0~>          2 actua1
                                          old

Fortunately, this iteration method appears to converge after only town or
three iterations.

     This discussion of chemical equilibria involving HLS, C02, NH-, and
water outlines the details of various steps used in the SWEQ model  to calcu-
late the concentration of each individual species in solution.   Table 2 gives
a summary of the various reactions which are accounted for by the model.
There are a total of eight reactions listed in this table.  First ionization
constants are involved in reactions 1, 3, 5, 7, and 8; and second ionization
constants are involved in reactions 2 and 6.  In addition, bicarbonate ions
react with ammonia to produce carbanate ions in reaction 4.  The corresponding
equilibrium equations based on the extent of each reaction occurring are
given to the right for each reaction in Table 2.  Except for reaction number
4 for carbamate formation, the equilibrium concentrations of each species
are shown to be proportional  to the pH of the solution.  If the pH is known,
it becomes a rather easy matter to compute the equilibrium concentration of
each species in solution.  If the pH is not known, an iterative method has
to be devised as discussed above whereby an initial pH is assumed.   Then as
steps in the iteration loop,  the concentration- of each species  is calculated
by calculating the extent of each chemical reaction.  From the  calculated
species  concentrations the sum of all electronic charges HT can then be cal-


                                     10

-------
culated by the equation shown at the bottom of Table 2.  Generally this sum
will not be zero, but the assumed pH can then be adjusted to bring the sum
closer to zero thus forming a closed iteration loop.  Iteration can then be
formed until the two valves agree within a small tolerance.  This calculation
method is very convenient and powerful because it can be readily expanded to
include other basic compounds as future needs arise.

     A rigorous thermodynamic approach to the problem of calculating chemical
equilibria in electrolyte solutions involves the use of activity coefficients
for each species in solution requiring interaction parameters between each
species.  These activity coefficients are then used to calculate the effect
of composition and ionic strength on the chemical equilibrium constants.  Such
a method has been proposed by Edwards, Newman, and Prausnitz (3) for aqueous
solutions of volatile weak electrolytes.  However, because of assumptions in
their model, their correlation  is not suitable for concentrated solutions of
these compounds. ' To avoid this problem and to minimize computer time required
for calculating the activity coefficient of each individual species, a more
empirical method was used for the SWEQ model.

     In the SWEQ model, the equilibrium constants in Table 2 are assumed to
be given by equations of the following form.


                In K. = In K?   + aCu <- + bCrn  + C I0'4            (23)
                    1       1      HO     \j\Jy

                where K- = equilibrium constant

                      K? = equilibrium constant at infinite dilution of
                           all  species

                a,b,c = parameters

                r
                 HpS = Total moles h^S absorbed/Kg of solution


                CCQ  = Total moles C02 absorbed/Kg of solution

                                                  2
                   I = ionic strength = 1/2 ^,-C^Z^ ,  Z.. = ionic charge


The constant a and b have been  found to be independent of temperature while
c is found to be dependent on temperature.  In many respects, this empirical
       'A  oaoer was aiven  by Edwards* Newman, and Prausnitz at the 70th AICHE
Meeting,  New York Session, 13-17 November 1977, on "Vapor-Liquid Equilibria
in Multicomponent Aqueous Solutions of Volatile Weak Electrolytes."  They
report a  new correlation  similar to their first paper, 4) but the range of
application has  been extended to temperatures from 0 to 170°C (32 to 338°F)
and  total  soHtte concentrations up to 10 molal.  This new work was published
as the final report of this project was being written, so no comparisons with
the  SWEQ  model have been  made.

                                     11

-------
                  TABLE  2.  SUMMARY OF  CHEMICAL  EQUILIBRIA INVOLVED IN CALCULATING
                             NH3-C02-H2S-H20 VAPOR-LIQUID  EQUILIBRIA



          Chemical Reaction                      Equilibrium Constant*

    i .  co, + H,O -»• HCO: + H+                      k = (H+) a
       „    „ 2       3                               (np-o-B-e)
         ~   ~e
      „    „
       c~   ~e

   2.  HCO: - co" + H+
         o     J
  3.  NH« + H  •*• NH*                             If _
        3          4

      n.-6-e      6



  4.  NH3 + HCOj •»• H2NCOO" + HgO                  k =


      n^-6-e   a     e



  5,  H2S -^ HS" + H*





  6.  HS" -^ S" + H*                              k

       Y    $


  7.  H20 + H+ + OH"                              k • (H+)(o + CCAU)

                (a + CCAU)
*
  8.   RCOOH •*• RCOO" + H+

      "sA"c    *                                       nsA--c


                                                                                                        4.
The sum of all  electronic charges 1s given as follows:   (HT) = a + 2e-6  +y  +2i()  +a + e  +5  -CCAU - H

-------
method is similar to the method used by Van Krevelen.  Van Krevelen found
that the equilibrium constants for reaction of hLS and CCL with NH-, were
proportional to ionic strength, so a single correlation parameter was intro-
duced to account for this effect.  This has been changed slightly in the SWEQ
model in order to predict multi component equilibrium data at high concentra-
tions by introducing a and b as additional parameters for the separate effects
of absorbed C02 and H2S.  Actual parameters used in the SWEQ model are sum-
marized in Table 3.  This table shows that only three coefficients have been
introduced.  A multiplying factor of -.278 times the concentration of absorbed
H2S and a temperature function time the ionic strength appear for the first
dissociation constant of C02 as given by reaction 1 in Table 3-.  The effect
of ionic strength has been taken directly from Figure 3 of Van Krevelen's
papers by fitting the curves in his plot to an analytical equation of the
following form

         (Effect of ionic strength on In K) = CIn               (24)

        where C = temperature dependent parameter

              n = empirical exponent (a value of 0.4 was
                  found although a value of 0.5 would be more
                  correct from Debeye-Huckel considerations)

Van  Krevelen's correlation was made in terms of the following reaction.

                                              +
C0
  2(g)
H2°(e)
NH
                             3(e)
                           HC0
                                            NH
where
where
                     =
                V.K.    (pco  ){NH3
 This  equilibrium constant can  be  rewritten  in  terms of a Henry's constant
 for C02  as  follows.
  /.K.
                      )(H2C03)(NH3)
 From Table 2,  this  represents  the  sum  of  reactions  1 and 3 as follows.
          V.K.


 In the SWEQ  model,  it is assumed  that  HCQ   and  ko are  independent of  compo-
 sition; thus any effect of ionic  strengtn2of  ionfc  strength  on  1C, K   becomes
 a similar effect on K-.
      The term of -0.278 CH2$  in  Table  3  has  resulted  from  fitting  C0?  partial
 pressures data from quarternary  FUS-COp-NHo-hLO mixtures measured  at Brigham
 Young University.  (5)   It does not  affect  ternary  C02-NH3-H20  data and only

                                     13

-------
        TABLE: 3.  EFFECT OF COMPOSITION AND IONIC STRENGTH
   _ ON CHEMICAL EQUILIBRIUM CONSTANTSA) _


         in K. = in  Ki° + aC^ + bC^ + cl°'4

    where K.° = equilibrium constant at infinite dilution of all
           1    species

        a,b,c = parameters

   CM c» Crn  = total  moles of H9S or CO, absprbed in one Kg of
    H2S   C02   solution        2       d
                          ?
            I = 1/2  £.jC.Z.  = ionic strength

           Z- = ionic charge
Chemical
Reaction
in Table 2
1
2
3
4
5
6
7
8


a
-.278
0
0
0
0
0
0
0


b
0
0
0
0
.427
0
0
0


c
-1.32 + 1558.8/T°R
0
ob)
0
0
0
0
0
     equation and constants given here are discussed in the section
 on the SWEQ model, equation 23.

^No effect of ionic strength is required for NHo because its equili-
 brium constant is used in combination with either h^S or C02«
                              14

-------
2.
becomes important when significant concentrations of both CCL and H2S are
present.  The third coefficient in Table 3 appears as a mumplying factor
of 0.427 times the concentration of absorbed C0? 'which affects the first
dissociation constant of HLS as given by reaction 5 in Table 3.  This effect
was also found necessary besides concentration terms in the Henry's constant
to correlate the quarternary H2S-C02-NH3-H?0 data.  Van Krevelen found that
a multiplying factor of 08089 times ionic strength to be necessary in log-,0
K.  In the SWEQ model, this is accounted for in the concentration-dependent
terms of the H2$ Henry's constant given in Table 1.

     The concentration effects given in Tables 1 and 3 were developed in the
following steps:

     1.  Binary NH,-H,>0 data were correlated to obtain the effect of free
         NH3 onitbi HSnry's constant of NHg.

         Binary H2S-H20 and ternary H2S-NHo-H20 data were correlated to
         obtain an adjusted zero concentration Henry's constant of H2S
         as noted at the bottom of Table 1, and an additional concentration
         parameter proportional to free NH3 concentration for the Henry's
         constant of H2$ was introduced as shown in Table 1.  The chemical
         equilibrium constant of reaction 3 in Table 2 for the combination
         of NH3 plus H  to give (NH» ) was also adjusted as an empirical
         parameter in order to fit the H2S-NH3-H20 "data.  It was also found
         necessary to introduce an effect of absorbed H2S on the Henry's
         constant of NH, in order to correlate the ternary data.  Thus
         four effects wire correlated:

              a)  The zero concentration Henry's constant of H2$

              b)  The effect of free NH~ on the Henry's constant
                  of H2S               J

              c)  The equilibrium constant of NH3 + H+ -> NH4+

              d)  The effect of absorbed H9S on the Henry's constant
                  of NH3.                 *•

     3.  Binary CCL-H20 and ternary C02-NH3-H20 data were correlated.  The
         effects of ionic strength on the first dissociation constant of
         ($2 was used directly from Van Krevelen's correlation.  The
         available data appear to be suitably correlated by this one
         effect so no new additional concentration parameters were in-
         troduced.  However, the zero concentration dissociation con-
         stant of C02 was adjusted slightly in order to obtain an im-
         proved representation of the C02-NH3-H20 data.  The equilibrium
         constant for the reaction of HCO," and NH- to produce H2NCOO~
         carbamate ions in reaction 4 of fable 1 was not changed from
         Van Krevelen's correlation.
                                 15

-------
     4.  Quarternary H?s-COo-NHq-H?0 data were correlated to obtain the
         effect of absorbed H7S on the first dissociated constant of
         H?CCL and the effect of absorbed C0? on the Henry's constant
         of H^S and on the first dissociation constant of H2$.

     Comparisons between measured and calculated data are given in a subse-
quent  section of this report.

     The chemical equilibrium constants for the reactions given in Table 2
are  dependent on temperature.  This effect is calculated in the SWEQ model
from equations of the same form given by Kent and Eisenberg (4) as follows.


           In K.° = A + B/T + C/T2 + D/T3 + E/T4                (28)


where  T is in degrees Rankine and concentrations are in gram moles or gram
ions/Kg of solution.  Actual parameters used are given in Table 4.  In many
cases  the parameters are the same as the ones used by Kent and Eisenberg (4).
Various changes were made in these constants as noted at the bottom of Table 4.
These  changes were as follows.

     1.  The reaction constant of NH, + H+ -> NhL+ was first adjusted
         empirically using available H2S-NH3-H20 volatility data and
         the equilibrium constant of HpS as puolished by Kent and
         Eisenberg.

     2.  The reaction constants for the first and second ionization
         constants of CCL were adjusted from Kent and Eisenberg's
         equations to fit available C02-NHo-H?0 data.  This was done
         so as not to affect the H^-NH^-H^O correlation.

     3.  After Parts 1 and 2 were done it was found by detailed com-
         parisons of measured and calculated data given in subsequent
         tables of this report that both the H?S and CO, volatility
         data could be adjusted slightly to improve thefr predicted
         values.  This was done by changing the first dissociation
         constants of H2S and C02.  The original constant for NHv
         was left unchanged.  The net effects of these various changes
         are noted at the bottom of Table 4.

     The equilibrium constant of NHL reacting with HCCL" to produce H?NCOO~
carbamate ion was used as published by Van Krevelen.  The dissociation con-
stant of HLO was used as published by Kent and Eisenberg t4).  The ionization
constant of carboxylic acids (RCOOH) in water (H20) are nearly independent
of temperature thus a single constant is used for RCOOH ionization according
to reaction 8 in Table 2.  The value\of -11.28 is based on a pK  of about
4.9 reported by Bomberger and Smith ' from potentiometric titrailons of actual
refinery sour water streams.  This reaction has been introduced into the
calculation method so that the effect of carboxylic acids on the volatility
of NH3 can be calculated.  A molecular weight of 60.05 is assumed in the SWEQ
calculation model, but another value could be entered if necessary.  The
amount of carboxylic acid in a given sour water stream can be obtained from a

                                     16

-------
                  TABLE  4.   SUMMARY OF TEMPERATURE  PARAMETERS,  USED TO  CALCULATE CHEMICAL
             	EQUILIBRIUM CONSTANTS IN TABLE  1	
              (Iniq = A + B/T + C/T2 + D/T3 + E/T4 for T  1n °R, and concentrations in gram
                                   moles or gram ions per Kg of soln.)
 Chemical
 Reaction
 In Table 2
     1
     2
     3
     4
     5
     6
     7
    8
  Lit
  Ref
                            Temperature Parameters
4,7,8     -241.79*
4,7,8,9   -295.60*
   *         1.587*
   1        -5.40
4,11,12   -293.88*
4,11,12   -657.965
4,.10
   6
 39.5554
-11.28
 536256*
 655893
  11160*
   3465
 683858*
1649360
-177822
    0
                                                TT
 -4.8123 X 108
 -5.9667 X 108
       0
       0
 -6.27125 X 10
-15.8964 X 101
  1.843  X 101
       0
                                                      1.94  X  10
                                                              .11
                                                      2.4249 X 10
                                                          0
                                                          0
                                                                ,11
                                                            -2.96445  X  10
                                                            -3.7192   X  10
                                                                  0
                                                                  0
                             I13
                             ,13
                                                .8*
                                                     2.555  X 10
                                                     ,11
                                                      Jl
                 -3.91757  X  101
                                                                                   ,13
                                           6.72472  X  10"  -10.6043  X 10
                                                      ,11    1 «sinSt  u> 1 A I 
-------
potentiometric titration of samples taken from the stream as performed by
Bomberger and Smith.  This information can then be used to calculate the
amount of caustic to be added in order to release the NH3.

     In the SWEQ model, the volatilities of H2S, C02, or NH3 in solution are
dependent on the H  ion concentration or pH of the Solution.  This effect
is shown in Figure 1 where the ratio of vapor over liquid concentrations on
a weights-basis are plotted at 120°C (or about 30 psia) versus pH measured
at 25 C.  These data were calculated assuming a 0.01  weight-% concentration
in the liquid phase.  From this plot, we see that H^S and CO- have greater
volatilities at low pH levels while NH-, has greater volatilities at high pH
levels.  This means that a process for simultaneous stripping of all three
components from solution must operate at an intermediate pH where all  three
have reasonable volatilities.  From this plot, the optimum pH measured at
25 C appears to be around 10, but we find that it varies depending on  the
mixture involved.  The equilibrium-constant parameters in Tables 3 and 4 and
the Henry's law equations in Table 1 give all the parameters necessary for
predicting vapor-liquid data in NI-L-CO^HL-HpO systems.   Other acidic  or basic
components could be added to the correlation simply by adding parameters for
the added components to these tables and by incorporating them into the com-
puter program.

     Details of a computer program based on the SWEQ  model  and comparisons
with literature data are given in the next sections of this report.
                                     18

-------
in
c
O
O)
   10,000
    8,000

    6,000

    4,000
2,000   -


1,000
  800
C O
O O
O O
  CM
3 -M
O- 3
•r- O
_J .Q
  (C
S ----
O)
> rd
O •r-
  (/)
i- 0.

§.0

-------
                                   SECTION 4

                  COMPUTER PROGRAM BASED ON THE SWEQ MODEL


     A computer program for calculating NH-, C02, abd HgS volatility data
from aqueous sour water systems has been developed basea on the SWEQ model.
The program is written to handle a wide range of conditions and temperatures.
The estimated ranges of applicability are as follows:

                 Property                     Range

                Temperature             20°C to 140°C
                Pressure                up to 50 psia*
                Composition             1 ppm to about 30 weight %
                                        dissolved NH3> carboxylic acid,
                                        salts, and caustic
                pH                      2 to 14

     Corrections for vapor phase non-ideality are recommended at
     pressures above 50 psia.

As presently written, the program will handle NH3, CO^, H^S, and water plus
NHg fixation effects due to carboxylic or stronger acids and the effects of
caustic addition.

     This computer program uses the new vapor-liquid equations presented
in the previous section of this report which were developed from both old
and new experimental data.  This same program was used to develop data
comparisons given in the next section of this report.

     The main features of the SWEQ model as it has been programmed are as
follows:

     1.  As shown in the next section, it is more precise than the Van
         Krevelen method of prediction.  This improvement is primarily
         due to the use of actual data at the conditions of commercial
         interest for development of the SWEQ model while the Van
         Krevelen correlation is used at extrapolated conditions.

     2.  The program will  take into account NH_ fixation effects
         due to carboxylic acids in sour water systems,,

     3.  The program will  also take into/account caustic addition
         to release fixed NhL.


                                      20

-------
     4.   The program can be readily converted to a subroutine for
         equilibrium stage calculations for various separation pro-
         cesses.  Calculations can be made going either up or down
         in a distillation process.

     5.   The SWEQ model can be expanded to additional acidic or basic
         components with only minor changes to introduce new ioniza-
         tion constants and Henry's constants.

Various ostions in the computer program are available to the user as
follows.9'

     1.  Option 1 allows the calculation of vapor-liquid equilibrium
         data at a specified temperature and liquid composition.  This
         option would be used for circumstances in which the temperature
         at liquid vapor equilibrium is known rather than the pressure.
         This was the option used in correlating available experimental
         data of this project.  It may also be useful in some process
         situations.

     2.  Option 2 allows the calculation of vapor-liquid equilibrium
         data at a specified pressure and liquid composition.  This
         option would normally be used for equilibrium stage process
         calculations going up a distillation tower.  The program
         calculates the temperature and vapor composition from a given
         stage.  The pressure change from stage to stage must be con-
         trolled by the user in specifying the pressure of the equilib-
         rium calculation.

     3.  Option 3 allows the calculation of vapor-liquid equilibrium
         data at a specified pressure and vapor composition.  This
         option would normally be used for equilibrium stage process
         calculations going down a distillation tower.  The program
         calculates the temperature and liquid composition from a
         specified vapor composition and pressure.  This option would
         normally be used for sour water stripper calculations.  The
         pressure increment between stages must be controlled by the
         user in each pressure specified to the program.  Option 3
         also calculates water in the condenser vapor at a specified
         pressure, temperature, and vapor stream composition on a
         water- free basis.  In this case, a zero water content is
         specified as input data for the calculation.  This response
         for zero water content only occurs with Option 3.

     Ammonia fixation and caustic addition effects can be calculated with
all three options given above.  Ammonia fixation effects can be calculated
by entering as data a specified wt. % of carboxylic acids in the liquid
               the three options listed here, a. fourth, option for a flash
       calculation has been completed.  This was done after this report
       was written, so the results are not in this report.  Please contact
       the author for the details.


                                     21

-------
analysis.  The amount to be entered may be determinate from a potent!ometnc
titration of the sour water under study.  The method of titration could be
the same or similar to that used by Bomberger and Smith (6).  A molecular
weight of 60.05 is assumed in the computer program.  This number was assumed
without any real basis and can be changed in the program without affecting
other parts of the program.  The effect of caustic addition can be calculated
in two ways as follows:

     1.  If a negative pH is specified as input data, then the program
         ignores the entry and calculates the pH based in the amount of
         caustic in wt % specified in the input data.  The input concen-
         tration refers to the liquid phasie even when option 3 is used
         for calculating down a distillation tower.

     2.  If a positive pH is specified as input data then the program
         computes the amount of caustic necessary to obtain the specified
         pH.  In this case, the concentration of caustic specified in
         the input data is set to zero.

    Both of thesie pH options use or compute pH data at the temperature of the
equilibrium stage.  If the pH of the liquid at 25 C is desired, the user must
specify this temperature ;and the liquid composition obtained from a higher-
temperature equilibrium stage calculation.  This would involve the use of
distillation option number 1.

     Table 5 gives a flow chart of the main program.  The format for entry of
data to the program is the same regardless of the options used.  The basic
program involves the reading of input data which then converts the data so
it can be processed by options 1, 2, or 3 in the program.  After these options,
the calculated equilibrium data are then printed by the program.

     Flow charts for options 1, 2, and 3 are given in  Figures  3,  4, and  5
respectively.  These options primarily act as executive programs which call
various subroutines necessary to perform the calculations.  Iteration loops
are involved in each of the options because of the problem of calculating
simultaneous chemical  equilibria at each condition.  The primary interation
of pH is done in each option by calculating equilibrium concentrations of
each species at assumed pH values.  Initially chosen pH values are arbitrary
so a test is made to check for electrical neutrality of the solution.   For an
arbitrary pH, neutrality will not occur; so then a new pH is chose in subse-
quent iterations until electrical neutrality within a small tolerance is
achieved.

     A direct listing  of the main program is given in Table §, and listings
of the various subroutines used by the main program are given in Tables 6
to 15.   These subroutines and their functions are as follows:
                                      22

-------
                              Read option and wt% carboxylic
                              acid and caustic in liquid
                              Read temperature, pressure, pH,
                              and either liquid or vapor
                              compositions in wt% depending
                              on option
                               Convert temperature iir°C to
                               absolute temperature in °R and
                               °K
                                            I
                               Test  for wt% NH. to be a
                               positive number
                                                  yes
                              Add a small value to the-com-
                              position to avoid calculation
                              problems at zero concentrations
    Option 1
Calculate pressure
and vapor compn.
from temperature
and liquid compn
      Option 2
Calculate temperature
and vapor compn.  from
pressure and liquid  compn.
      Option 3
Calculate temperature
and liquid compn.
from pressure and
vapor compn.
                                     Print Equil.
                                        Data
                    Figure 2.   Flow diagram of SWEQ main program.
                                        23

-------
          Convert input compositions to moles/Kg of
          soln in liquid - subroutine CFX
          Initialize iteration parameters for pH
          calculation - subroutine PHST
          Calculate chemical equilibrium constants -
          subroutine KREAC
  o
          Calculate the concentration of each
          species in the liquid - subroutine SPECL
          Estimate a new pH value - subroutine NPH
          Test if input pH is positive
          Iterate ten times with a new estimated
          value for caustic concentration each time
>
o
c
                                                        c
                                                        to
                                                        I/I
                                                        0)
          Test to see if electrical neutrality is
          computed within a small tolerance
                                  yes
          Calculate equilibrium pressure and vapor
          compositions - subroutine PRESY
                                                        10
 o
 c
 O)
                                                        &.
                                                        O)
          Print results
Figure 3.   flow diagram of option 1  of  SWEQ  computer program.
                             24

-------
o
      Convert  input composition to moles/Kg
      of soln.  in liquid - subroutine CFX
      Initialize  iteration parameters
      Initialize  pH  parameters for each
      temperature iteration  - subroutine
      PHST                »
      Calculate chemical  equilibrium con-
      stants -  subroutine KREAC
                                              no
      Calculate the concentration of  each
      species in the liquid  -  subroutine
      SPECL
      Estimate a  new pH value  -  subroutine
      NPH
      Test if input pH is  positive
                              yes
      Iterate ten times with a  new estimated
      value for caustic concentration  each
      time
      Test to see if electrical  neutrality
      is computed within a small  tolerance
                              yes
  Calculate equil.  pressure and
  vapor compositions  -  subroutine
  PRESY
Estimate new temp.
NTEMP
- subroutine
  Test to see if calculated press-
  ure is within  a small  tolerance
  of specified pressure
                                                                 X
                                                   Print results
r ten
ations
    Figure 4.   i?low  diagram of option 2 of  SWEQ  computer  program.
                                        25

-------
Convert input compositions  to mole
fraction in vapor - subroutine YFX
I
                                    
-------
             TABLE 5.  COMPUTER PROGRAM BASED ON THE SWEQ MODEL
C       SWEQ COMPUTER PROGRAM
C THIS COMPUTE* PROGRAM HAS WRITTEN BY GRANT M. WILSON FOR THE API CREC
C COMMUTE, RONALD G. GANTZ SOUR WATER STRIPPER PROJECT MANAGER. QUESTIONS
C ABOUT THIS PROGRAM SHOULD BE DIRECTED TO EITHER GRANT M. WILSON OR RONALD
C G. GANTZ. THIS PROGRAM IS WRITTEN IN FORTRAN FOR OPERATION ON A TIME
C SHARE TERMINAL CONNECTED TO A DIGITAL EQUIPMENT CO.MODEL 10 COMPUTER.
C IT CAN BE CONVERTED FOR USE AS A SUBROUTINE OR FOR BATCH OPERATION. FOR
C OPERATION AS A SUBROUTINE THE ERROR MESSAGES NOW PRINTED ON THE TERMINAL
C WOULD HAVE TO BE CHANGED SO THAT THE EXECUTIVE PROGRAM WOULD TAKE
C CORRECTIVE ACTION. ERROR MESSAGES ARE NOW PRINTED BY STATEMENTS 320, 420,
C AND 515 IN THE MAIN PROGRAM; AND BY STATEMENTS 12 AND 35 IN SUBROUTINE
C SPECV AND PY STATEMENT 7341 IN SUBROUTINE SPECL.
C THIS PROGRAM CALCULATES EQUILIBRIUM VAPOR-LIQUID COMPOSITIONS FOR WEAK
C ELECTROLYTE MIXTURES CONTAINING NH3,C02,H2S,RCOOH,CAUSTIC, AND WATER.
C COMMENTS IN THE PROGRAM DESCRIBE VARIOUS OPTIONS POSSIBLE AND THE FUNCTION
C OF VARIOUS PARTS OF THE PROGRAM
        COMMON TC,TK,TR,PSI,P,XA,XC,XS,XW,CA,CC,CS,CAS,CCS,CSS,
     lYA,YC,YS,YW,Wft,fcC,WS,WW,PHO,PH,TOL,HT,DPH,PHA,PH8,AL,BT,GA,
     2DE,SI,HP,SD,EPS,EKS,£KCA,EKCAO,£KCB,EKA,EKrt,EKCC,EKSB,EKSA,HTU,
     3EL1,HA,HC,HS,MW,CCST,RHQ,XSA,XCAU,CSA,CCAU,ZET,WSA,«CAU,CCAUS
        COMMON ICD
        OPEN (UNI T = 20, DE V ICEs ' DSK', ACCES'Ss'SEQIN', FILEr'SWSD1)
        DATA WA,WC,WS,WW,WSA,WCAU/17.03,«4.01,34.08,18.02,60.05,
-------
                           TABLE 5   (continued)
 2       RtAD(20,100(>)  TC,PSI,PHO,XA,XC,XS,XW
        IF(XK)  10,10,20
 10      ICO  =  1
        GO TO  30
 20      ICO  =  0
 30      XCAU =  XCAUO
 C CONVERT  TO ABSOLU1E  TEMPERATURES  IN  DEC.  K  AND  OEG.  R.
        TK = TC+273.15

 C A  SMALLKVALUE*ISBAOOED  10  THE COMPOSITIONS  IN ORDER  TO  AVOID  CALCULATION
 C PROBLEMS AT  ZERO  CONCENTRATIONS.
        XA = XAtlfc-12
        XC - XC+1E-12
        XS = XS-HE-12
        Xw = XWtlt-12
 C NEGATIVE AMMONIA  CONCENTRATION  SIGNALS  NEW  OPTION,  RCOOH,  OR  CAUSTIC
 C DATA.
        IF(XA)  1,200,200
 200     GO TO  (300,400,500).NOOPT
 C OPTION  1 CALCULATES  PRESSURE AND  VAPOR  COMPOSITION  FROM SPECIFIED
 C TEMPERATURE  AND LIQUID  COMPOSITION.  SUBROUTINE  FUNCTIONS ARE  LISTED WITH
 C THE  SUBROUTINES
 300     CALL CFX
        PH = 7
        HTO  =  1
        CALL PHST
        CALL KREAC
        ELI  = CA
 C ITERATION  LUOP TO CALCULATE EITHER EQUILIBRIUM  PH OR EQUILIBRIA  AT
 C SPECIFIED  PH.
        DO 310  1C = 1,100
        CALL SPECL
        CALL NPH
        IF(PHO) 305,305,302
 302     CCAU =  (CCAU+HT)/2
        IF(IC-IO) 310,310,330
 C TEST FOR PH CONVERGENCE
 305     IF(ABS(HT/TOL)-.0001) 330,330,310
 310     CONTINUE
 320     WRITE(5,1010)
 330     CALL PRESY
 C EQUILIBRIA CALCULATED;  TRANSFER TO PRINT  OUT OF  RESULTS.
        GO 10 900
 C OPTION 2 CALCULATES TEMPERATUKE AND  VAPOR COMPOSITION FROM SPECIFIED
 C PRESSURE AND LIUU1D COMPOSITION.
 400      CALL  CFX
        P = PSI
        ELI = CA
        PH =  7
        HTO = 1
C FIRST 00 LOOP ITERATES TO DETERMINE  TEMPERATURE.
                                                       (continued)
                                     28

-------
                           TABL'E 5  (continued)
        DO 440 IEsl.100
        CALL PHST
        CALL KREAC
C SECOND 00 LOOP ITERATES TO FIND EQUIL. PH AT ITERATION TEMPERATURE.
        DO 410 1C = 1,100
        CALL SPECL
        CALL NPH
        IF(PHO) 405,405,402
402     CCAU = (HT+ccAu>/2
        IF(IC-IO) 410,410,130
C TEST FOR PH CONVERGENCE
405     IF(ABS(HT/TOL)-.0001) 430,430,410
410     CONTINUE
420     WRITE(5,J010)
430     CALL PRESY
        CALL NTEMP
C TEST FOR PRESSURE CONVERGENCE.
        IF(ABS(PSI/P-1)-.001) 460,460,440
440     CONTINUE
        WRITE(5,1011)
C EQUILIBRIA CALCULATED? TRANSFERS TO PRINT OUT OF RESULTS.
460     GO TO 900
C OPTION 3 CALCULATES TEMPERATURE AND LIQUID COMPOSITION FROM SPECIFIED
C PRESSURE, VAPOR COMPOSITION, PLUS WTZ OF RCOOH AND CAUSTIC IN LIQUID.
500     CALL YFX
        P = PS1
        CSA s 10*XSA*RHO/WSA
        CCAU = 10*XCAU*RHO/WCA"
        ELI = 0
        CC = 0
        CA = 0
        CS = 0
        CAS s 0
        PH x S
        HTO * 1
c FIRST po LOOP ITERATES TO (^TERMINE TEMERAIURE.
        DO 530 IE s 1,100
        CALL PHST
C SECOND DO LOOP ITERATES TU FIND EQUIL. PH AT ITERATION TEMPERATURE.
        DO 510 IC=1,100
        CALL KREAC
        CALL SPECV
        CALL NPH
        IF(PHO) 505,505,502
502     CCAU = (CCAU+HT)/2
        IFUC-10) 510,520,520
C TEST FOR PH CONVERGENCE.
505     IF(ABS(HT/TOL>-.0001> 520,520,510
510     CONTINUE
515     WK1TE(5,1010>
520     CALL NTEMP

                                                     (continued)
                                    29

-------
                             TABLE 5. (continued)
C
C
C
C
C  3
900
C TEST FOR PESSURE CONVERGENCE.
        1F(A6S(PSI/P-1)".001) 550,550,530
530     CON1INUE
        hRITE(5,1011)
C EQUILIBRIA CALCULATED; TRANSFER TO PRINT OUT OF RESULTS.
550     GO TO 900
C COMPOSITION DATA IN THE LIQUID PHASE ARE USED IN THE PROGRAM IN  TERMS
  OF MOLES OF COMPOUND PER KILOGRAM OF SOLUTION. THE VAPOR PHSAE IS  IN  TERMS
  OF MOLE FRACTION. THE NEXT TEN STATEMENTS CONVERT THESE BACK TO  WEIGHT
  PERCENT. THE ORIGINAL CONVERSION OF THE INPUT DATA TO MOLES PER  KILOGRAM
  AND VAPUR MOLE FRACTION IS DONE IN SUbROUTINES CFX, YFX, AND FOR OPTION
    PARTLY IN THE MAIN PROGRAM.
        XA = 100*CA*wA/(1000*RHO)
        XC = 100*CC*WC/(1000*RHO)
        XS = 100*CS*rtS/(1000*«HO)
        XCAU = 100*CCAU*WCAU/(1000*RHO)
        XW = 100-XA-XC-XS-XSA-XCAU
        YT s YA*toA+YC*WC+YS*WS+YW*Wrt
        YA = JOO*YA*wA/YT
        YC = 100*YC*WC/YT
        YS = 100*YS*WS/YT
        YW s 100*YM*nw/YT
        XMT = XA/WAfXC/wC+XS/WStXCAU/WCAUtXW/WW+XSA/WSA
        XMT = 100/XMT
        XMA = XA*XMJ/WA
        XMC = XC*XMT/WC
        XMS = XS*XMT/*S
        XMCAU * XCAU*XMT/WCAU
        XKW s XW*XMT/WW
        XMSA s XSA*XM1/WSA
        YMT = YA/WA+YC/WC*YS/rtS+YW/WW
        YMT s 100/YMT
        YMA = YA*YMT/wA
        YMC s YC*YM1/«C
        YMS = YS*YMT/wS
        YMW = YW*YMT/WW
        TF = TC*I.8+32
        PKPA = P*6.895
        ATM s P/14.696
C OUTPUT FROM THE FOLLOWING STATEMENTS IS MORE OR LESS SELF EXPLANATORY
C IN THE FORMAT STATEMENTS.
        WRITE(5,1030) TCrTF,TK,TR,P,ATM,PKPA,PH
        hRITE(5,10?0)
        EK s YMA/XMA
                      XA,YA,XMA,YMA,EK
        WR1TE(5,1040)
        EK  s  YMC/XMC
        WRITE(5,1050)
        £K  s  YMS/XMS
        I»RITE(5,1060)
        EK  s  YMW/XMrt
        WRITE(5,1070)
                      XC,YC,XMC,YMC,EK

                      XS,YS,XMS,YMS,EK

                      XH,Ytt,XMW,YMM,EK
                                                        (continued)
                                    30

-------
                          TABLE 5   (continued)
        WRITE(5,1072)  XSA,XMSA
        WR1TE(5,107«)  XCAU.XMCAU
        XTOT  = XA+XC*XS+X«+XSAtXCAU
        XMTOT s XMA+XMC+XMS*XMKtXMSA+XMCAU
        YT01  s YA + YC + YS+Y*f
        YMTOT s YMA+YMC+YM8+YMW
        WRITE(5,1075)  XTOT,YTOT,XMTOT,YMTOT
        GO TO 2
1000    FORMAT(IOE)
1001    FORMAT(I,2E)
1010    FORMATC PH DID NOT CONVERGE IN 100 CYCLES')
1011    FORMATC TEMPERATURE DID NOT CONVERGE IN 100 CYCLES')
1020    FORMATC
     1«                     WEIGHT PERCENT
     2' COMPONENT          LIQUID     VAPOR
     3-VALUE')
1030    FORMAH/X
     I1 TEMPERATURE1,F8.2, ' C,   SF8.2,' F,  SFS.a,' K»',F8.2,' RV
     2* PRESSURE  ',Fe.2,' PSIA,',F8.3r' ATM,»,F9.2f' K-PASCALSV
     31 PH',8X,F8.3//)
                                        MOLE PERCENT1/
                                       LIQUID     VAPOR
1040
10SO
1060
1070
1072
1074
1075
1060
         AMMONIA         S5F10.5)
         CARBON DIOXIDE  *,5F10.5)
         HYDROGEN SULFIDE',5F10.5)
         WATER           SSFiO.S)
         CAR80XYL1C ACID • ,F10.5, IOX,F10.5)
FORMATC SODIUM HYDROXIDE1,F10.b,10X,F10.5>
FORMATC TOTAL           V5F10.5)
FORMAT(X,F4.0»3F7.3.7F6.2,9F6.3»F6.2)
END
FORMATC
FORMATC
FORMATC
FORMATC
FORMATC
                                 31

-------
                          WBLE 6.  SUBROUTINE KREAC
        SUBROUTINE KREAC
C THIS SUBROUTINE CALCULATES CHEMICAL EQUILIBRIUM CONSTANTS AS FOLLOWS.
C       SYMBOL            EQUILIBRIUM
C        EKS        H2S FIRST IONIZAT10N
C        EKCAO      C02 FIRST IONIZATIOM AT ZERO IONIC STRENGTH
C        EKCB       C02 SECOND IONIZATION
C        EKA        NH3 PLUS PROTON GOING TO AMMONIUM ION
C        EKW        WATER DISSOCIATION
C        EKCC       BICARBONATE PLUS AMMONIA GOING TO CARBAMATE
C        EKSB       H2S SECOND IONIZATION
C        EKSA       RCOOH IONIZATION
C THE EFFECT OF IONIC STRENGTH ON EKCAO IS CALCULATED BY THE CALLING PROGRAM.
        COMMON IC,TK,TR,PSI,P,XA,XC,XS,XK,CA,CCrCS,CAS,CCS,CSS,
     lYA,YC,YS,YW,WA,ViC,WS,Ww,PHO,PH,TOL,Hr,OPH,PHA,PHB,AL,BT,GA,
     2DE,SI,HP,SD,EPS,EKS,EKCA,tKCAO,EKCB,EKA,EK*l,EKCC,EKSB,EKSA,HTO/
     3£LI,HA,HC,HS,HW,CCST,RHO,XSA,XCAU,CSA,CCAU,ZET,WSAr*CAU,CCAUS
        EKS = EXP(-293.88+68385tt/TR-6.27125E8/UR*TR)+2.5551Ell/(TR**3)
     l-3.91757El3/(TR**«)+.
        EKCAO * EXP(-2/J1.79+536256/TR-/».8123E8/(TR*TR) + l,
        EKCB = EXP(-295.64655893/TR-5.9667E8/(TR*TR)+2.4249EH/(TR**3)
     1-3.7192E13/(TR**«J)
        EKA * EXPU.587+11160/TR)
        EKW = EXP(39.5S5«-177822/TR+1.8«3E8/(TR*TR)-.85«)E11/(TR**3)
     1+1.0292E13/(TR**0))
        EKCC = EXP(-5. 40+1925*1. 8/TR)
        EKSB = EXP(-657.965+1649360/TR-15.e964E8/
-------
                           TABLE 7;  SUBROUTINE.HENRY
           SUBROUTINE HENRY
   C THIS SUBROUTINE CALCULATES HENRY'S CONSTANTS FOR  NH3,C02,  H2S,  AND  H20
   C RESPECTIVELY AS HA, HC, HS, AND Hh. HA OF AMMONIA  IS DEPENDENT  ON  THE
   C CONCENTRATIONS OF SPECIES NH3, C02, AND  H2S RESPECTIVELY BY  THE SYMBOLS
   C CASr CC, AND CS. HS OF H2S IS DEPENDENT  ON CAS  AND CC. HW  FOR *ATER IS
   C THE VAPOR PRESSURE OF WATER.
           COMMON TC,TK,TR,PSI,P,XA,XC,XS,XW,CA,CC,CS,CASrCCS»CSS,
        1YA,YC,YS,YW,WA,KC,WS,WK,PHO,PH,TOL,HT,DPH,PHA,PHB,AL,6T,GA,
        2DE,SI,HP,SD,EPS,EKS,EKCA,EKCAO,EKCB,EKA,£K*,EKCC,EKSR,EKSA,HTO,
        3ELl,HA,HC,HS,HW,CCST,RHO,XSA,XCAU,CSA,CCAU,ZET,AiSA,*lCAU,CCAUS
           TK = TR/1.8
           HA a EXP(178.339-15S17.91/TR-25.6767*ALOG(TR)
        |».01966*TR+(131.4/TR-.1682)*CAS
        lt.06*(2*CC+CS))
           HC = EXPU8.33-24895.1/1R + .223996E8/(TR*TR)-.090918E11/(TR**3)
        H.12601E13/(TR**4))
           HS = EXPUOO.684-24.6254E4/TRt2.39029E8/(TR*TR)-1.0l898EU/(TR
        l**3)+l.S973tE13/(TR**a)-.05*CAS+(.965-a86/TR)*CC)
           HW - EXP(1«.466-6996.6/(TR-77.
           RETURN
           END
                            TABLE, 8.  SUBROUTINE YFX
        SUBROUTINE YFX
C THIS SUBROUTINE CONVERTS COMPOSITIONS IN WTX TO VAPOR CONCENTRATIONS
C IN MOLE FRACTION. VAPOR COMPOSITIONS FOR NH3.C02, H2S, AND KATER RESPECTIVE
C ARE GIVEN BY THE SYMBOLS YA, YC, YS, AND Y*.
        COMMON TC,TK,TR,PSI,PfXA,XC,XSrXw,CA,CC,CS,CAS.CCS,CSS,
     lYA,YC,YS»YW,^AfwC,WS,WW,PHO,PH,TOLTHT,OPH,PHA,PHB,AL»BT,GAr
     20E,SI,HP,SD,EPS,EKS,EKCA,EKCAO,£KCB,EKA,EK*,EKCC,EKSH,EKSA,HTO,
     3ELI,HA,HC,HS,hW,CC-ST,RHO,XSA,XCAU,CSA,CCAU,ZET,^SA,rtCAU,CCAUS
        XT =
        YA = XA/(WA*XT)
        YC s XC/(rtC*XT)
        YS = XS/(wS*XT)
        Yw = XW/(WW*XT)
        CA = 1
        RETURN
        END
                                    33

-------
                             TABLE.9.  SUBROUIINE CFX
         SUBROUTINE  CFX
 C  THIS  SUB«OUI1NE CONVERTS  COMPOSITIONS  IN  WTX  TO LIQUID CONCENTRATIONS
 C  IN  MOLES  PEH  KG OF  SOLUTION.  LIQUID  COMPOSITIONS FOR  NH3,  C02,  H2S, RCOOH,
 C  AND CAUSTIC RtSPECTlVELY  ARE  GIVEN BY  THE SYMBOLS CA,  CC,  CS,  CSA, AND CCAU
         COMMON  TC,TK,TR,PSI,P,XA,XC,XS,Xrt,CA,CC,CS,CAS,CCS,CSS,
      lYA,YC,YS,Yw,KA>«C,*S,rtH,PHO,PH,TOL,HT,DPH,PHA,PHB,AL,8T,GA,
      20E,SI,HP,SD,EPS,EKS,EKCA,EKCAO,EKCB,EKA,EKW,EKCC,EKSB,EKSA,HTO,
      3ELI,HA,HC,HS,Hrt,CCST,KHO,XSA,XCAU,CSA,CCAU,2ET,«SA,WCAU,CCAUS
         F s 1000*RHO/(XA+XC+XS+XW+XSAtXCAU)
         CA  =  XA*F/wA
         CC  =  XC*F/rtC
         CS  =  XS*F/WS
         CSA = XSA*F/WSA
         CCAU  a  XCAU*F/WCA(J
         RETURN
         ENO
                                 TABLE 10.  PHST
        SUBROUTINE PHST
C THIS SUBROUTINE , INITIALIZES PARAMETER  VALUES FOR  ITERATIVE  CALCULATION
C OF PH. THIS SUBROUTINE DETERMINES THE  VALUE OF  THE  TOLERANCE  TOL TO BE USED
C IN TESTING FOR PH CONVERGENCE, AND  INITIALIZES  PH AND OTHER PARAMETERS FOH
C THE ITERATION
        COMMON TC,TK,TR,PSI,P,XA,XC,XS,XW,CA,CC,CS,CAS,CCS,CSS,
     2De,SI,HP,SD,£PS,EKS,EKCA,EKCAO,EKCB,£KA,EK«,EKCC,EKSB,EXSA,HTO,
     5ELI,HA,HC»HSrHw,CCST,«H(l,XSA,XCAU,CSA,CCAU,Z£T,WSA,hCAU,CCAUS
        IF(PHO) 40,30,30
30      PH = PKO
40      PHA = 0
        PHB = 14
        HTA = -CA
        HTB s 2*CC+CS
        IFCHTB+HTA) 50, $0,60
50      TOL = -HTA
        GO TO 70
60      TOL = HTB
70      *L = 0
        TOL s TOLtlE-f
        DPH s 1
        EPS s 0
        HIO s 1
        RETURN
        END
                                      34

-------
                           TABLE IV.  SUBROUTINE SPECV
        SUBROUTINE SPECV
C THIS SUBROUTINE CALCULATES EQUILIBRIUM SPECIES CONCENTRATIONS  IN  THE
C LIQUID PHASE FROM A SPECIFIED VAPOR COMPOSITION, TEMPERATURE,  AND  AN
C ASSUMED PH. SPECIES CONCENTRATIONS IN THE LIQUID ARE GIVEN BY  THE  FOLLOWING
C SYMBOLS.
C       SYMBOL          SPECIES
C        CAS             NH3
C        CCS             C02
C        CSS             H2S
C        AL              HC03-
C        BT              C03--
C        DE              NH4+
C        EPS             CARBAMATE ION
C        GA              HS-
C        SO              S—
C        SI              OH-
C        2ET             RCOO-
C        CCAUS           NAt
C THIS CALCULATION IS PERFORMED AT AN ASSUMED PH SO THAT SPECIES CONCENTRATIO
C CAN BE SOLVED FROM DIRECT ALGEBRAIC EQUATIONS. SOME ITERATION  IS REQUIRED
C BECAUSE OF THE EFFECT OF IONIC STRENGTH ON THE FIRST IONIZATION OF C02
C AND THE EFFECT OF SPECIES CONCENTRATIONS ON THE VOLATILITY OF  NH3  AND H2S.
C ELI = IONIC STRENGTH.
        COMMON TC,TK,TR,PSI,P,XA,XC,XS,XW,CA,CC,CS,CAS,CCS,CSS,
     lYA,YC,YS,Y«,KA,WC,WS,WW,PHO,PH,TOL,HT,t>PH,PHA,PHB,AL,8T,GA,
     2DE,SI,HP,SD,EPS,EKS,EKCA,EKCAO,EKCB,EKA,EM,EKCC,EKSB,EKSA,HTO,
     3ELI,HA,HC,HS,HW,CCST,RHO,XSA,XCAU,CSA,CCAU,ZET,«SA,wCAU,CCAUS
        COMMON ICO
        CALL HENRV
        XW x .9
        PW s 1E-19
        PA s 0
        PC s 0
        PS s 0
        EKCA s EKCAO
        DO 30 1=1,100
        EKAP s EKCA
        EKCA s EKCAO*EXP((-1.32tl558.8/TR)*ELI**.
-------
                          TABLE 11. (continued)
        P a PW
        ICO s 0
        GO TO 6
5       Yw = Pw/PSl
        YTOT = U-Yw)/(YAtYC + YS)
        YA s. YA*YTOT
        YC = YOYTOT
        YS = YS*YTOT
        PA = PSI*YA
        PC = PSI*YC
        PS = PSI*YS
        60 TO 7
6       PA s (YA*Prt/YW+PA)/2
        PC = (YC*Prt/Yrt+PC)/2
        PS s (YS*PW/YW+PS)/2
        PBA s p
        P = PW+PAtPO+PS
7       CAS = PA/HA
        CCS = PC/HC
        CSS = PS/HS
        HP = tXP(-?.30259*PH)
        Ai. = EKCA*CCS/HP
        8T s £KCB*AL/HP
        OE s EKA*HP*CAS
        EPS s EKCC*CAS*AL
        GA * EKS*CSS/HP
        SO = £KSB*GA/HP
        SI s EKW/HP-CCAU
        2ET = EKSA*CSA/(HP+EKSA)
        CCAUS = CCAU-SI
        CA = CAS*OE+EPS
        CC ~ CCS+ALtBTtEPS
        CS = CSS+GA+SD
        ELI s ((AU+4*6T*OE*EPS+GA*fl*SDtSl4HPtZET*CCAUStCCAU)/2+ELI)/2
        TN1" s (IOOO*RHO-CA*«A-CC*«C-CS*WS-CSA*WSA-CCAU*WCAU)/WW
     1-ALtEPS-SI
        TNM s CA+CCtCS+BT-OE+GA+SD+SI+TNrt+ZET42*CCAU+CSA
        XiMO v- XW
        XW = TNrt/TNM
        XH s (XMtXwO)/2
8       JF(XW)  80,80,9
80      XW s XW+.l
        GO TO 8
C TEST FOR ITERATION CONVERGENCE
9       IF(ABS(P«/PwO-l.)-.OOl) 10,10,30
10      IF(ABS(PA/PAO-1.)-.001) 15,15,30
15      IF(ABS(PC/PCO-1.)-.001) 20,20>30
20      IF(ABS(PS/PSO-l.)-.OOl) 50,50,30
30      CONTINUE
35      WRITE(5,«0)
40      FOKMATC  OIDNT CALCULATE LIQUID IN  100 CYCLES')
50      RETURN
        END
                                 36

-------
                        TABLE 12.  SUBROUTINE NRH
        SUBROUTINE: NPH
C THIS SUBROUTINE CALCULATES A NEW ESTIMATED PH FROM  A PREVIOUS PH.
C CRITERIA USED ARE THE SUM OF IONIC CHARGES 10 ELECTRICAL NEUTRALITY, HT;
C AND ANY CHANGES IN SIGN OF HT FROM A PREVIOUS ITERATION.
        COMMON TC,TK,TR,PSI,P,XA,XC,XS,Xrt,CA,CC,CS,CAS,CCS,CSS,
     !YA,YC,YS,YW,WA,WC,i«S,WW,PHOfPH,TOL,HT,DPH,PHA,PHB,AL,BT,GA,
     2D£,SI,HP,SD,EPS,EKS,EKCA,EKCAO*EKCB,EKA,EK'Al,EKCCrEKSB,EKSA,HTO,
     3ELI,HA,HC,HS,HW,CCST,«HO,XSA,XCAU,CSA,CCAU,ZET,«SA,WCAU,CCAUS
        HT - AL+2*BT+GA-DE+S1-HP+2*SD+EPS+ZET-CCAU
        TOL s 2*CCtCS
        IF(TOL-CA) 60,70/70
60      TOL = CA
70      IF(PHO) 81,81,80
80      HT = HT+CCAU
        GO TO 88
81      IF(HTXDPH) 8«,8«,82
82      HTO 8 ,5*HTO
84      DPH s. -HT*HTO/(ABS(HT)+1E-19)
        PH = PHtOPH
        GO TO (8fi,88re5),NOOPT
85      REF = HP*(HT+EKA*CAS*HP)
        PH = SQRT(REF/(EKA*CAS))
        PH = -,5*ALOGlPH*HP)/2.30259
88      RETURN
        END
                                    37

-------
                           TABLE 13.  SUBROUTINE SPECL
c  THIS suoRouHNE CALCULATES EQUILIBRIUM species CONCENTRATIONS IN THL LI«UIO
C  PHASE FROf. A SPECIFIED LIUUID COMPOSITION,  TEMPERATURE,  AND  AN ASSUMED
C  Ph. SYMBOLS USEU FOk SPECIES CONCENTRATIONS  ARE  THE  SAME  AS  FOR SPECV.
C  ITERATION IS NECESSARY BECAUSE  THE CARBAMATE CONCENTRATION CANNOT BE
C  SOLVED DIRECTLY, AND BECAUSE EKCA IS DEPENDENT ON  IONIC  STRENGTH.
        COMMON TC,TK,TR,PSI,PfXA,XCfX5,XW,CA,CC,CS,CAS,CCSrCSS,
      lYA,YC,YS,Yi»,WA,wC,WS,WW,PHO,PH,TOL,HT,DPH,PHA,PHB,AL/BT,GA»
      2DE,SIrHP,Sr),EPS,EKS,fcKCA,EKCAO,EKCB,EKA,EKrt,EKCC,EKSH,EKSAfHTO,
      •*ELI,HA,HC,HSfHW,CCST,*HO,XSA,XCAU,CSA,CCAU,2ET,«SA,WCAU,CCAUS
        HP = EXP(-2.30259*PH)
        AL = CC
        DO 734 1AL = 1,100
        EKCA = EKCAO*EXP((-1.32+155a.8/TR)*ELI**.«)
        CCS = HP*AL/EKCA
        BT = EKCB*AL/HP
        DE = EKA*hP*CA/((H-AL*EKCC)*U+EKA*HP/(ltAL*EKCC))>
        EPS = AL*EKCC*(CA-DE)/(ltAL*EKCC)
        CCSI = CCS^AL+BTfEPS
C  TEST FOR ITERATION CONVERGENCE
        IF(ABS(CCST/CC-1)-.0001) 736,736,733
732     IF(CCST-1E-16) 733,733»73«0
733     AL = CC/2
        CCS = CC/2
        BT = IE-19
        DE = EKA*HP*CA/(1+EKA*HP)
        EPS = IE-19
        60 TO 736
73<»0    AL s. AL*CC/CCST
        6A s. CS*EKS/(HP*(1 + (1*EKSB/HP)*EKS/HP))
        SD = GA*EKS6/HP
        SI s EKW/HP-CCAU
        ZET = EKSA*CSA/(HPtEKSA)
        CCAUS = CCAU-SI
7>fl     ELI = (ALt1*BT+DE+EPStGA+1*SD+SItHP+ZET+CCAUStCCAU)/2
7341    WRITE(5,735)
735     FORMAT(•  CARBAMATE 01DN1 CONVERGE IN 100 CYCLES')
736     RETURN
        END
                                     38

-------
                            TABLE 14.  SUBROUTINE PRESY
        SUBROUTINE PRESY
C THIS SUBROUTINE CALCULATES EQUILIBRIUM  VAPOR  COMPOSITION  AND  PRESSURE
C FROM TEMPERATURE AND CALCULATE!) SPECIES CONCENTRATIONS  OF NH3,  C02,  AND  H2S
C IN THE LIQUID
        COMMON TC,TK,TR,PSI,P,XA,XC,XS,XW,CA,CC,CS,CAS,CCS,CSS,
     lYA,YC,YS,Yrt,WA,WC.«S,WW,PHO,PH,TOL,HT,DPH,PHA,PHB,ALr8T,GAr
     2DE,Sl,HP,SD,EPS,EKS,E*CA,EKCAO,EKCB,EKA,EKrt,EKCCffcKSB,EKSA,HTO,
     3ELl,HA,HC,HS,HW,CCST,RHO,XSA,XCAU,CSA,CCAU,ZET,rtSA,WCAU,CCAUS
        CCS = HP*AL/EKCA
        CSS s HP*GA/EKS
        CAS s DE/(EKA*HP)
        CALL HENRY
        PA s CAS*HA
        PC s CCS*HC
        PS s CSS*HS
        TNW s UOOO*RHO-CA*WA-CC*WC-CS*WS-CSA*WSA-CCAU*WCAU)/KW
     1-AL+EPS-SI
        TNM s CA+CC+CS+BT-DE+GAtSD+SI+TNW+ZETt2*CCAU+CSA
        P« a TNW*HW/TNM
        P * PA+PCtPStPW
        YA = PA/P
        YC a PC/P
        YS s PS/P
        YW = PW/P
        RETURN
        END
                             TABLE IS.  SUBROUTINE NTEMP
                    NTEMP
 C  THIS  SUBROUTINE  ESTIMATES A NEW TEMPERATURE IN AN  ITERATIVE  CALCULATION
 C  TO  AGREfhITH  A SPECIFIED PRESSURE.  THE ONLY CRITERION  USED  IS  THE  CALCULAT
 C  PRESSURE  OF A PREVIOUS ITERATION VERSUS THE SPECIFIED  PRESSURE.  AN ASSUMEl
 C  EFFECTIVE  HEAT OF  VAPORI/ATION OF 9000X1.987 BTU PER POUND  MOLE  IS USED
 C  TO  ESTIMATE A NEW  TEMPERATURE.
         COMMON  TC,TK,ltf,PSI,P»XA/XC,XS,XW,CA,CC,CS,CAS/CCSfCSS,
      IYA,YC,YS.YW,WA,WC,WS,W«»PHOfPH,TOL,HT,DPH,PHA,PHB,AL.BT,GA,
      2DE,SI,HP,SD,EPS,EKS,EKCA,EKCAO,EKCB,EKA,EKrt,EKCCftKSB,EKSA,HTO,
      3ELI,HA,HC,HS,HW,CCST,RHO,XSA,XCAU,CSA,CCAU,Z£T,rtSA,WCAU,CCAUS
         TRi  r TR
         TR  s -ALOG(PSI/P)/9000+1/TR1
         TR  s 1/TR
         1C  = TR/1.8-273.15
         TK  s TR/1.8
         RETURN
         END
                                     39

-------
                  Table                   Function
     Subroutine    No.                 of Subroutine

         KREC        6       Calculates chemical equilibrium constants from
                            parameters in Tables 3 and 4

         HENRY       7       Calculates Henry's constants from parameters in
                            Table 1

         YFX         8       Converts vapor compositions in wt % to vapor
                            concentrations in mole fraction

         CFX         9       Converts liquid compositions in wt % to
                            liquid concentrations in moles/Kg of solution

         PHST       10       Initializes pH iteration parameters

         SPECV      11       Calculates pressure and equilibrium species con-
                            centrations in the liquid phase from a specified
                            vapor composition, temperature, and asssumed pH

         NPH        12       Calculates a new estimated pH from a previous pH

         SPECL      13       Calculates equilibrium species concentrations in
                            the liquid phase from a specified liquid composi-
                            tion, temperature, and an assumed pH

         PRESY      14       Calculates equilibrium vapor concentrations and
                            pressure from temperature and calculated species
                            concentrations in the liquid

         NTEMP      15       Estimates a new temperature in an iterative cal-
                            culation so that the calculated pressure will
                            agree with a specified pressure

A discussion of each of these subroutines in the order listed above is given
in the following text of this report.

KREAC

     Equations used in KREAC come from Tables 3 and 4.  The symbols used in
the subroutine relate to the various chemical reactions in Table 2 as follows.
                Fortran
                Symbol

                 EKS
                 EKCAO
                 EKCB
                 EKA
                 EKW
                 EKCC
                 EKSB
                 EKSA
Chemical Reaction
   in Table 2

          5
          1
          2
          3
          7
          4
          6
          8
     40

-------
The effect of ionic strength of EKCAO of reaction 1 is not computed in the
subroutine because it changes each pH iteration.  This effect is therefore
computed in subroutine SPECL for each pH iteration cycle where the ionic
strength from the previously computed cycle is used for the next iteration.

HENRY

     Equations used in HENRY come from Table 1.  The symbols used in the
subroutine relate to the Henry's constant parameters given in Table 1  as
follows.

              Fortran                 Henry's
               Symbol             Constant for

                HA            free NH,
                HC            free CO^  (or H,CO-)
                HS            free H^S      *  J
                HW            vapor pressure of water

YFX

     The conversion of wt.% in the vapor to mole % in the vapor from sub-
routine YFX is fairly straight forward.  The fortran symbols and associated
molecular weights entered  by means of a data statement at the beginning of
the main program are as follows.

               Component       Symbol      Molecular Height

                 NH,             WA             17.03
                 CO-             WC             44.01
                 H9S             WS             34.08
                 HfO             WW             18.02
                 RCOOH           WSA            60.05
                 NaOH            WCAU           40

CFX

     Subroutine CFX is similar to YFX except that the concentrations in wt.
 % are converted to liquid  concentrations in moles/Kg of solution.  To  do this,
the sum of all wt.% given  as input to the program are summed and divided into
1000 x RHO to obtain the normalizing factor.  A value of RHO = 1 has to be
used or the concentrations will not come out in moles/Kg of solution.   This
assignment is made in the main program as the first executable statement.  The
number of moles of each component is then computed from the normalizing factor
times its concentration on a weight basis divided by its molecular weight.

PHST

     Subroutine PHST initializes parameters used in the pH iteration pro-
cedure.  If a positive PHO (for pH) is specified to the subroutine, then the
subroutine assigns PH = PHO and the other parameters have no effect.   If a
negative PHO is specified  then it means that the program must compute  the  pH.

                                     41

-------
 In  this  case,  it  assigns  the  limits over which the pH can be varied which
 are from 0 to  14  and  assigns  a  tolerance to be used by subroutine NPH  to
 test for convergence.   The  tolerance variable is assigned to either the sum
 of  acid  gas concentrations  if they are  in excess or to the Nl-L concentration
 if  it is in excess.   Carbon dioxide reacts with two moles of NH~, so its con-
 centration is  multiplied  by two in computing the acid gas^oncentration.   If
 the tolerance  assigned  by this  method is less than 1 x 10   then a tolerance
 of  1 x 10   moles/Kg  of solution  is assigned.  The variables DPH and HTO are
 iteration parameters  used by  NPH.  For  their use, see subroutine NPH.

 SPECV

      Subroutine SPECV is  the  main subroutine used in option 3 to calculate
 temperature and liquid  composition from specified pressure, vapor composition,
 plus RCOOH and/or caustic in  the  liquid.  The steps of this subroutine are not
 too obvious, so details of  the  calculation procedure will be discussed here.
 Temperature iteration and pH  iteration  are done outside of the subroutine,
 so  the subroutine calculates  pressure and liquid composition from temperature,
 pH, and  a specified vapor composition.  This is d ne by first estimating the
 partial  pressure  of water in  the  vapor  phase using the vapor pressure of
 water and Raoult's law  as follows.

               PW  = (HW) x (XW)                                (29)

               where PW  =  water  partial  pressure
                    HW  =  vapor  pressure of water
                    XW  =  mole fraction  water in liquid phase; initially
                          assumed  to be  1.0

 The partial  pressures of  the  other components are then calculated from the
 mole ratio of  the components  over water times the partial pressure of water.

                                  (YA, YC. or YS)
               (PA, PC,  and  PS)  =        [YW]x


               where PA, PC, and PS = partial pressures of NH-, C0?,
                                     and H2S, respectively

                    YA, YC, or  YS = vapor mole fractions
                                YW = water mole fraction
                                PW = water partial pressure

 The  total  pressure P  is then  calculated as the sum of the partial pressures;
 and  the  concentrations  of free  NH3 (CAS), C0? (CSS), and H?S (CSS) are calcu-
 lated from their  partial pressure divided by the Henry's constant of each
 component.  These  Henry's constants depend on the composition of the liquid
 phase so this  computation involves an iterative procedure where  Henry's con-
 stants computed from the  liquid composition.  This procedure could diverge
 instead  of converge, so each  new partial pressure is assumed to be the average
 of the new computed partial  pressure and the old computed partial pressure.
 This  technique requires a minimum of ten iterations to achieve an accuracy of
j^..0.1%;  so it  uses more computer time in order to avoid possibility of

                                      42

-------
diverging instead of converging.  A maximum of 100 cycles  is specified in
the subroutine for convergence; if this number is specified in the subroutine
for convergence; if this number is exceeded,  the subroutine writes to unit 5
a warning signal that 100 cycles are exceeded.  If this occurs, one may want
to give the old partial pressure more weight than the new one so as to improve
convergence.

     Once the concentrations of free NH-, C02, and H?S in the liquid have
been calculated for each iteration cycle, then the concentrations of all
species concentrations in the liquid phase can be computed according to the
chemical equilibria summarized in Table 2.  Symbols used by the subroutine
for each species present are summarized as comment statements at the beginning
of the subroutine in Table 11.  Once these concentrations have been computed,
then the mole fraction of water can be recomputed and then the iteration cycle
is repeated,  iterations are continued until the new computed partial  pressure
of each component equals the old computed partial pressure within a tolerance
of +_ 0.1%.  In each iteration cycle, the subroutine allows for any RCOOH or
caustic specified to be in the liquid phase as input data to the subroutine.

NPH

     Subroutine NPH estimates new pH values based on information gained from
previous pH iterations*   This subroutine uses the requirement of electrical
neutrality  as the determining equation for either "increasing or decreasing
the pH.  The equation for electrical neutrality can be written as follows.

              I  C. 2,. = 0


              where C. = concentration of component i in moles/Kg
                         of solution

                    Z. = electronic charge

In general, for a randomly selected input pH, the electrical  neutrality
summation will not equal zero.  In this subroutine, this summation is  repre-
sented by the symbol HT.  In order to bring to zero, the step length for a
new pH value is computed from the following equation:

              ApH = 0(k) x (HT)/|HT|                             (32)

              where  pH = computed pH increment, DPH
                      k = a proportionality constant, HTO
                     HT = electrical neutrality summation
                   |HT| = absolute value of HT

If HT changes sign compared to a previous iteration, then the proportionality
constant k  is increased by a factor of two, and the iteration is continued.
By this procedure, the pH increments are only determined by the algebraic
sign of HT compared with previous iterations.  Thus, when HT changes sign,
then the increments are reduced by a factor of two.  DPH and HTP are intially
set to unity by subroutine PHST.  This convergence method is slow, but

                                      43

-------
dependable.  Other faster methods could probably be devised to speed up  this
calculation.

SPECL

     Subroutine SPECL  is similar to SPECV in that the pH, temperature, and
composition of one of  the phases is given and the pressure and composition of
the  other  phase is calculated.  In the case of SPECL, the total amounts  of
NH3, CCL,  H S, H?0, RCOQH, and caustic in the liquid phase are given; and the
composition of tne vapor phase is calculated.

     This  subroutine is used for both options 1 and 2 of the main program.
The  method of computing the concentrations of each individual species in the
liquid  requires a knowledge of the chemical equilibrium constants which  in
turn are dependent on  the concentrations of the individual species present.
Thus, an iterative calculation procedure is required where the ionic strength
ELI  is  initially set to equal the total NH- concentration CA in the main pro-
gram.   Subsequent iterations then give better values for the ionic strength.
The  calculation method used in this subroutine is based on the calculation
method  discussed in the previous section of this report on the SWEQ model;
equations  7 to 22.  Because of H2NCOO~ formation, equations 16 and 17 are not
used to solve for a and &; instead a is used as an iteration parameter along
with ionic strength.   Initially a (Fortran symbol AL) is assumed to be the
total C02  concentration in the liquid; equations 14 and 15 are then used to
calculate  the concentration of free C02 (CSS) and of hLNCOO" ions (BT).  These
are  also listed as equations 1 and 2 in Table 2.  The equilibrium NH.  con-
centration in solution is obtained by simultaneously solving equations 3 and
4  in Table 2 by algebraic methods to obtain the equations for NHd~ concen-
tration (DE) and H2NCOO~ concentration (EPS) used in this subroutine.  Iter-
ation is continued until the sum of all C0? species equals the amount of CCL
in the  liquid from the starting composition.  If the sum of the species
concentrations is higher or lower than the starting composition, then AL is
proportionately changed by multiplying the old AL by the ratio of starting
composition over the sum of the species concentration as follows.
                                                               (33)

           CC = C02 starting concentration

         CCST = sum of CC"2 species concentrations

After this calculation, the concentrations of (HS~) and (S=) ions are cal-
culated using equations similar to equations 16 and 17 in the section on
the SWEQ model.  The actual equations involved are equations 5 and 6 in
Table 2.  These can be solved to give the following:
                                     44

-------
=

                                                (34)
                      +) + k5 +  W(H+0
                                                                  (35)
In the subroutine, these have the following  symbols:

                                       Fortran Symbol

                              Y              GA
                              4-              SD
                              k5             EKS
                              ke             EKSB
                              Mr             CS
                              H*             HP

     After this calculation, the only  species left are from RCOOH and from
water dissociation; these are calculated from equations 7 and 8 in Table 2
where SI represents the extent of water dissociation and AET represents the
extent of RCOOH dissociation.  From  the calculated species concentrations,
the ionic strength can be calculated and iteration is then continued until
the sum of CO- species equals the COp  in the starting composition within
+_ 0.01%.  When this test is satisfiea, the subroutine returns to the main
program.

PRESY

     This subroutine computes the partial pressure of NHo, COp, H2S, and
water from equations 4, 5, and 6 given in the section on the 5WEQ model.
To do this, the individual species concentrations of C02, H2S, and NH3 repre-
sented by CCS, CSS, and CAS are computed from equations similar to equation
19 in the section on the SWEQ model.   In order to calculate the partial
pressure of water, two quantities are  first  calculated in the subroutine.
These are the total number of moles of water, TNW, present in 1 Kg of solu-
tion (RHO = 1) and the total moles of  all components, TNM, in 1 Kg of solu-
tion.  TNW is computed from the residual weight left after subtracting the
weight of NHV C0?, H?S, RCOOH, and caustic  respectively from the 1000 grams
of solution divided by the molecular weight  of water.  TNM is calculated by
summing the moles of all species present including water in 1000 grams of
solution.  The partial pressure of water is  then computed from the vapor
pressure of water, HW, times the moles of water over the total moles.  The
total pressure is then calculated as the sum of the partial pressures, and
the vapor mole fraction of each component is calculated from its partial
pressure divided by the total pressure.

NTEMP

     Subroutine NTEMP is used to estimate the correct temperature for  an
equilibrium calculation where the total pressure is specified.  This occurs

                                     45

-------
in options 2 and 3.   For this purpose  a  simple  equation  is  used  as  follows


                                                                    (36)
              1n/PSpecified V _gooo  /  1_
                l^calculatedy          \J Rnew
                                                      Nold

where -9000 corresponds approximately to  the heat  of  vaporization  of  water.


              -9000 = AHvap   =  18,000 Btu/lb  mole                  (37)
                      1.987           1.987

The above equation can be solved for T°Rneu)  to  give the  following:



               I_   „  _I_   -lAsM /9000                 (38)
                                      pcalc
In the subroutine, these have the following symbols:
              TR - f Rnew                                           (39)


             PSI ' P                                                (40)
             TRI = T°RQld                                           (42)


This subroutine also computes the temperature in °C from TR before returning
to the main program.

     Tables 5 to 15 represent a total  of ten subroutines used by the main
computer program.  A large bumber of subroutines are used in order to make
it possible to devise various options  in the main program.   Many options  are
possible; an attempt was not made to develop programming for each possible
option because of the large amount of  programming required.  Instead, three
options were programmed which demonstrate the use of the subroutines.  Thus,
flow charts for options 1, 2, and 3, given in Figupeg 2, 3, and 4, primarily
involve the use of subroutines with some programming done in between to satis-
fy the requirements of the option.  In option 1, the main iteration is to
calculate the pH.  When the pH is specified, then the iteration changes
slightly to calculate the amount of caustic necessary to achieve the speci-
fied pH.  This method of calculation occurs in all  three options.  Distilla-
tion options 2 and 3 involve a second  iteration loop besides the pH iteration
loop.  This is necessary to find the correct temperature at a specified
pressure.  An example of input and output data for the computer program
listed in Table 5 is given in Tables 16 and 17.
                                      46

-------
     Table 16 explains the data format to be used in entering data to the
computer program.  The information in this table must be studied carefully
before using the computer program.  Table 17 gives an example of computer
output from data specified in Table 16.  This listing is self explanatory.

     The next section of this report gives a numerical example of calculations
necessary for an actual design problem and a subsequent section gives data
comparisons and evaluations between calculated and measured data.
                                      47

-------
     TABLE 16.  INPUT DATA FOR SAMPLE  PROBLEM
                  WITH SWEQ COMPUTER PROGRAM
         Parameter

Option numbers for calculating
liquid composition and temp-
erature from a specified vapor
composition and pressure

Weight percent carboxylic acid
in liquid

Weight percent caustic in
liquid

Temperature, °C (For Option 3
this is used as a starting
temperature)

Pressure, psia; specified
pressure

pH, a positive entry specifies
the pH for the calculation.
The computer program will
adjust the amount of caustic
in the liquid independent, of
the concentration entered
above when a positive pH is
entered

Weight percent concentrations
in the vapor phase

NH
co
Symbol

NDOPT
XSA


XCAUO


TC



PS I


PHO
 Value
Entered
    .05


    .05


 100



  20


   8.5
XA                .01

XC                .01

XS                .01

XW             100

    (continued)
                         48

-------
                        TABLE 16.   (continued)
Format for data <
3
NDOW
100
sntry:
.05
XSA
20
PST

.05
XCAUO
8.5 .01
PHO" "XT


.01 .01 100
"XT ~XS~ "XT
Additional lines of temp., pressure, etc. can follow

111—1  111   This entry will signal a new option line to
                  follow this one.
NDOTT     XSA"    XCAUO


Then lines of temp., pressure, etc.
                                    49

-------
          TABLE 17.  COMPUTER OUTPUT FROM DATA IN TABLE 16 WITH
                    COMPUTER PROGRAM BASED ON THE SWEO MODEL
TEMPERAIURE
PRESSURE
PH
lOfl.88
20.00
8.500
c,
PSIA,

287.99
1.361

f,
ATM,

382
137

.03
.90

K,
687.66
R
K-PASCALS



COMPONENT
AMMONIA
CARBON DIOXIDE
HYDROGEN SULflDE
WATER
CARBOXYL1C  ACID
SODIUM HYDROXIDE
TOTAL
WEIGHT
LIQUID
0.00091
0.00017
0.00073
99.93077
0.05000
0.01742
100.00000
PERCENT
VAPOR
0.01000
0.01000
0.01000
99.97001


100.00000
MOLE PERCENT
LIQUID
0.00096
0.00007
0.00039
99.97572
0.01501
0.00785
100.00000
VAPOR
0.01058
0.00409
0.00529
99.98004


100.00000
K-VALUE
11.02807
57.78305
13.62491
1.00004



                                50

-------
                                  SECTION 5

                     SAMPLE PROBLEM USING THE SWEO MODEL
     Information given in the two prior sections of this report on the SWEQ
model and on the computer program based on the Sk'EQ model can probably be
better understood by giving a numerical example which shows the calculations
necessary in an actual sour water stripper design case.  For this purpose
the following sample problem  ' is given.

     A refluxed sour water stripper operates at a condenser temperature of
212 F at a pressure of 8.7 psig (23.4 psia).  To achieve the desired removal
of HpS and NH3 from the stripper feed, the overhead gas from the condenser
must contain 48 Ib/hr of NH3 and 49.7 Ib/hr of H?S.  Determine the amount of
water in the exit gas, and the reflux composition.
Gas rates
                     NH3 = 17.03
                                            = 2'82 mole/hr
                                       mole
                     HS = 49.7/34.08 = 1.46 mole/hr
From Raoult's Law, the partial pressure of water in the vapor phase is:

                     p.p.(H20) -  (V.P.iyj) . X^


Assume that X    = ^          At ^ v.p     _
                     p.p.(H20) = 0.9 (14.7) = 13.2 psia

The partial pressure of (H9S = NH-) = 23.4 - 13.2 5,10.2 psia,  therefore
the total moles of overhead gas =J(2.82 + 1.46) x  irr = 9.82 moles/hr.
Assumed water rate is 9.82 - (2.82 + 1.46) = 5.54
the assumed vapor composition is
                                                    le/hr.  In summary,
     a^This sample calculation is given through the courtesy of Ron Gantz
and co-workers of CONOCO who did the numerical calculations and wrote this
sample problem.  It has been checked at Brigham Young University and found
to be correct.

                                     51

-------
                      1 b/hr   mole/hr  mole/fr.  p_.p. , psia

         NH3          48      2.82      .287         6.72

         H2S          49.7    1.46      .149         3.49

         H00          99.8    5.54      .564       13.19
           *                   "9782    T7DT50
Calculation of the  liquid composition in equilibrium with  the assumed
vapor composition involves simultaneous solution of the appropriate  chemical
equilibria and phase equilibrium equations.  The chemical  reactions  (Tables
2 to 4)  are:
         NH, + H+ t NH.+            kNH
           3      *   4              NH3
                    H+
+                        JP J
         HS" ->• S  + H               k  - =
         Mi  j 5  f H               KHS-


The chemical equilibrium constants are correlated in the general form  (Table
4).


         In ki = A + B/T + C/T2 + D/T3 + E/T4

where T is the temperature in °R.
At 100°C (= 671. 76R),

        In k    =39 5554 - ^ 77822   1..843-108  .8541  -IP11,  1.4292 • IP13
        in k      39.5554   -r-  +           ~  (671 .7)3   +   (671.7)4
        In kNH  = 1.587 + 11160/671.7
              »J


           k    = 8.032 • 107
        Ink    - -293 88 + ^83§5g   6!?7T25/J°8+ 2-5551 ' ^^   3.91757 • IP13
        in ic       ^3.88 +                      +  (671.7)3 --- (671.7)4
           kH $ = 2.805 • 10"7
                                     52

-------
       In k    -  fi<57 QfiR + IlilM  15.8964-IP8.6.72472- 1011  10.6043-I O13
       in KHS     ot>/.yb& + -^7T-7      (671.7)2      (671~7)J	(67i;7)4
          kHS" = 9.06 • 10"13
   For phase equilibria, the Henry's Law coefficients from Table 1  are used:





        In  (HNH  ) =  178.339 -  1591 - 25.6767 In (TR) + .01966 • (TR)

              
-------
From the Henry's Law coefficients  for H?S  and  NFL,  and  the  assumed  vapor
partial  pressures, the free H2S and  NH3  concentrations  in the  liquid  can
be calculated:
                        3

              C$s(free  H2S) = ^^ = 7.89 •  10~3 gm-mole/Kg


A pH must now be assumed - use 8.5


               A pH must now be assumed  -  use 8.5

                        pH = -log10[H+]

                      [H+] . e-2-303 - PH  .  3J5 . 1Q-9


The chemical equilibrium equations can now be solved for  the concentrations
of all other species in solution:

                  . .
               H3   [NH3JLH*J


                ] = [NH3][H+]


           [NH4+] = (1.85) (3. 15 • 10"9) (8.032 • 107) = 0.468 gm-ions/Kg


             k
             K
              H2S
            [HS"] = kH2SH2S                         [H9S] = C
                       IJFT
               -i - (2.805- 10"7)(7.89- 10"3)   n 7n
                ] =  - 3.15 .10-9 - = °-70 gm-ions/Kg
              HS
                                    54

-------
                 [S=]  =
                          LHJ
                                           )=  2.02.10-4gm.1ons/Kg
If the assumed pH and values of C.., (L- are correct,  the  solution should
be electrically neutral, that is:      b5


                [H+]  + [NH4+] = [MS'] + 2[S=]  + [OH

                3.15  • 10"9 + 0.468 = 0.68 +  2(2.02 - 10'4} + 1.6 • 10"4
                0.468  * 0.68
A trial and error procedure for pH, with successive  substitution for species
concentrations at each pH level, must be used  to  reach a converged solution.

First, adjust the Henry's Law coefficients  for the current values of H«S
and NH3 concentrations:


                                         ""'i*.  J682 (CAS) + .06 Cs
     In (HNH ) = In (3.640) +

where GS = total H2S = C$s + [HS~] + [S=]
                               » .00789 + 0.68  +  1.96 - 10'4 - 0.688 9
               In (HMM ) = 1.292 + N!TT -  .16821(1.85) + .06(.688)
                    wi,             671.7         I
            '3


          HM,,  = 3.99
                                gm-mole/Kg
               In (HH s) = In (442.5)  =-.05  CA$

                    H    = 403.4 psia
                    HH2S   gm-mole/Kg


          Calculate new free H2$, NH3  concentrations


                c   = P-P-(NH3l = |^| = 1 .68 gm-mole/Kg
                 •       H S       J-yy

                c   = fiiEiiifei) = 3.49  = 8i65 . 1Q-3  g
                 55    H,i q       H-UO.f


                                    55

-------
 The  total  anmonia  concentration, C^,  is


          CA =  CA$  +  [NH4+]  =  1.68 + 0.468 = 2.15 gm-moles/Kg


 The  total  H^S  concentration,  C<., is
     CS = CSS + tHS~J + CSJ = 8'65 ' 10"  + °'70 + 1<96 ' 10~  = -709

A new water mole fraction  in the  liquid  should be  calculated for use with
Raoult's Law to provide a  new vapor composition.

A free water concentration,  Cu n


     C
      H20        f -  CS ' ^ • wt- H2S> - CA '
          C  0 =  [1000 - 0.709(34 08) - 2.15(17.03)] _^
            £•                    lo.Od

          r    -  co o gm-mol e
          LH20 ~  ^--   K^


The free  water mole fraction in the liquid is
         XH90 =  H2°
           L     ZCi

          SCi = [NH3] + [NH4+] + [H+] + [H2S] + [HS~] + [S=] + [HgO]

              = 1.68 + .468 + 3.15 • 10"9 + 8.65 • 10"3 + 0.70 + 1.96 • 10"4

                + 52.2 + 1.6- 10"4

              = 55.06
         p.p.(H20) = (0.948)(14.7) = 13.9 psia

         p.p..(H2S + NH3) = 23.4 - 13.9 = 9.5 psia


Total moles in the vapor = 4.28-^t= 10.54
                                     56

-------
Moles of H20 vapor = 10.54 =-4.28= 6.26
The new vapor partial pressures are:
         P-P.(NH3) = 1^- • 23.4 = 6.26 psia
         P.P.(H2S) = i4jj_ - 23.4 = 3.24 psia
               t—      \ \J • 
-------
increasing the value of  H   would increase  NH.    and decrease  HS  .
Thus, the correct pH must be lower than the initial  assumption of 8.5.

If [H+] = 4.15 • 10"9, and [NH3], [H2S] are assumed constant, then
This is close enough to use for the next guess
           nH = ln (4.15-IP"9) _ R
           PH     -2.303	8'38

Using this pH and the current values of free H?S,  NH~, begin again the
successive substitution procedure for species concentrations and continue
until a final solution is reached (achieving electrical  neutrality).   In
most cases, several more pH trials may be required.

The final solution is

         pH = 8,38

                   Vapor Composition         Liquid Composition
                   Ib/hrmole/hr           wt. fr. mole fr.

         NH3       48          2.82              .036    .0384

         H2S       49.7        1.46              .018    .0096

         H20      114          6.33              .946    .9520
                                    58

-------
                                   SECTION  6

                     COMPARISONS AND  EVALUATIONS  BETWEEN
                          CALCULATED AMD  MEASURED  DATA


     Information in this  section will  be discussed  in the following order:

          a)  Evaluation  of Van Krevelen prediction model
          b)  Evaluation  of SWEQ prediction model
          c)  Evaluation  of new NH3-H2S-H20 and NH3-CO?-H2S-H?0 data
          d)  Ammonia fixation by  carDoxyfic acids  and release of
              NH3 by addition of caustic

These subjects will be discussed by frequent referral to data summarized in
Tables 18 to 27 which contain comparisons  between calculated and measured
vapor-liquid equilibrium  data.  Not all  literature data were examined in
this project because of the limited scope  of the project.  However, an attempt
was made to examine as much data as possible.  Table 28 summarizes various
references collected during the project.   This table also indicates the type
found in each reference and whether the  data were used for modeling purposes.
Tables 29, 30, and 31 give summaries  of  deviation errors between calculated
and measured partial pressures in  Tables 13 to 27 for NH,, C0?, and HpS
respectively.

     In developing the SWEQ model, some  individual experimental points have
been ignored and some entire data  sets have been ignored.  As a general policy,
individual experimental points in  a given  set of measurements have been ig-
nored in developing the correlation model  when deviation errors from these
points appeared to be radically difference from the main set of data.  Entire
sets of published data were ignored when deviations appeared to have little
or no definite pattern and were also  very  large.  When this has occurred, the
data ignored and reasons  for ignoring are  noted at the bottom of the table.

Evaluation of Van Krevelen Model
     The Van Krevelen prediction model which applies when NhL/hLS ratios are
greater than 1.5 was derived by Van Krevelen et al'' from low temperature
data.  These are compared with Tables 20 to 27 where columns headed VK repre-
sent predicted partial pressures from the Van Krevelen model ' and columns
headed MEAS represent measured partial pressure data.  The following is a
summary of the various comparisons.
         listing of the Van Krevelen computer program is given in the Appendix.

                                      59

-------
TABLE 18.  H2S-H20, NH3-H2S-H20, AND NH3-H20 SYSTEMS,  COMPARISON  OF  CALCULATED AND
           MEASURED DATA OF MILES AND WILSON21  AND OF  CLIFFORD  AND HUNTER17

Temo.
•c
89,
88,
8B,
80,
120,
120,
120,
150.
150,
150,
80,
80,
*e,
•»!




Holes/Kq of Soln.
HH,
— 3-
.900
,VB0
.000
.HUB
\Hl
.009
.900
,009
,010
.994
»339
I"'
!s2s
*«', 7.399
80, 4,535
C$2.
9,008
0,000
0.000
9,000
0,000
8,000
0.000
!00o
,000
,000
,000
,0t>0
.000
0,000
0.000
9,1100
0.000
80, 13,7*5 0,fl(i0
80, 14.358
120, (
120, 1
1,515
,B'2
120. 0.959
120. 0.415
120. «
120, 1
,519
.933
120, 8.H99
120. 0.011
60, 3
,452
69, 8.B55
60. 13.155
60, IS
• 3*2
60, 23.*00
60, 27.066
ice, 3
,452
100, 8.675
tee. 12.358
100. 17,386
140, 0,555
l«0, 3.285
140, !
.692
I8B. 7,829
.000
,000
,000
.000
,009
,{100
,000
,000
,«0a
,c 00
0,000
0,000
8,000
0,000
8,000
0,000
0,009
0)008
8,000
0,000
0,000
0,000
H£
9,054
»,297
0,5119
0.515
0,047
9,2190
9,341
0,044
0.136
9.232
1,358
9.925
4,157
1,066
5,553
4,250
4,340
1,015
2,831
3,528
0,696
0,946
0.077
0,178
0,108
0,413
1.698
0.001
0,000
8.000
0.000
B.flBfl
0,000
0.000
0,000
0,800
0,4)00
0,000
0,080
0,000
0,000
0,000


VK API MEAS



Partial

YK


t
Pressure
H,S
API
IB68.6
5932,6
K1M.I
10297^7




0,0 5,
lit
85,
137,1 |27,
200.
296,
342,
384.
776,
775,


170,
91 ,
145,
505,
1605,
3.
181,
495'

367)
365,
6611,
1229,
HI6.
1491 ,
1530|
399,
479,
442,
2023,
2099,
24,
106,
174,
87,
135.
521.
2583.
3,
214.
634,
1342,
2705,
4178,
3610,
711,
2065,
3255,
5258,
294,
r 1788,
2389,1 3166,
2808,2 4435,
12.1
72,4
136,5
208,9
465,3
597,7
574,4
2373.9
2285,1
23,8

18215
81,2
177,3
568,7
3179,5
2.
241.
627.
1442.
2957.
4559,
6079,
770,
2311,
3867,
6979,
299,
1825,
3376,
«92T,9






349,0
1225,
1056,
99,
172,
233,


1024,
285,
65,
246,
762,
8,











1166,7
4932,2
8422,4
1177,5
3637,5
6223,5
6160,6
2399,0
8670,7
3««>.8
7652,1
2244, T
19b
-------
TABLE 19.  NH3-H2S-H20 SYSTEM, COMPARISON OF CALCULATED AND MEASURED DATA OF  TERRES
                                                                                   29


Temp.

IS'
20,
20,
20.
20,
40.
40.
40.
60,
60,
60,



Holes/Kg of Soln.
NH3_
0.M0
1.574
3.975
4.T27
5,526
1.356
3.453
4,234
5,484
1,891
3.253
3.991
5.205
C0j_
*,0f)0
B.PVU5
0,ian0
0.000
0.000
e|0s>0
0.000
0,000
H2i
0.411
0,783
1.998
2,356
2,758
1.719
2,139
2,746
0,954
1,631
1.998

NH3
VK API.
a
8,
22!
25.
52)
4,2
a. a
26,1
33.0
40,7
52,4
66.7
65, 93,9
57,0 55. 1
94, 8 1B2.6
114,1 131,7
145,0 186.7
Partial

HEAS
0.0
0,0
0.0
5,6
13.0
8.2
22,2
26,9
30,4
25,5
48,5
76,1
145,0
Pressure,

VK
4,9
8,3
16,8
18.1
19,6
59,7
69,4
123,'z
181,1
205,1
240,2
mm Hg
H,S
API
5,6
10.1
24,6
28,0
32.3
67,8
84,2
104,2
130,4
213.2
255,4
131,2


WAS
5J
65,
UK.
190,
till'.
293,
134,
2«5,
250.
365,





-------
TABLE 20.  •NH,-H«S-0,SYSTEM, COMPARISON UF CALCULATED^.MEASURED
n3""2
                           DATA Of  VAN  KkEVELEN, ET AL.

                           VIC     API    WAS     «     API    fCAS
21, *,}fll 8,01)8
24, 8,5*5 4.0P1
21, l.7»8 », t»a
24, 2.358 D.reo
21, 2.121 8. nee
21, 1.215 t.atiB
21, 1,565 8. ewe
28, l.|18 9, line
21, 1.788 4.01)8
21, 2,1'B 8, nee
28, 2,110 0,8(19
21, 0.268 0,008
21, 1.548 0,888
28, 1,175 8.0(10
21, 1,781 1.000
21, 2,268 1,880
21, 2,661 1,101
28, 1.260 8,888
28. 0.575 1.808
26, |,|58 1,888
21, 1,730 8,8ft8
II, 2,318 0,880 I
20, 2,120 8,080
21, 1,231 1,0(10
21, 1,545 1.888 f
28, 1,068 8.080
21, 1,720 8,001
21, 2,271 1,188 <
21, 2,870 6,*B0 I
28, 1,468 8,000 I
40, 1,198 0.001 1
41, 1,515 8. Bill f
41, 1,718 1.888
48r 2,350 0,000
41, 2,120 0,000
48, 1,215 0,000 t
41, 0,595 0.0B0 i
41, 1,111 0,881 i
48. 1,748 1,880 <
41, 2.4B0 B,8B1
41, 2,131 0,000
41, 1,241 8,000 1
41, 1,561 1,000 1
40, 1,175 1,100 1
10, 1,710 0,080 I
41, 2,261 1,000 1
41, 2,460 1,300
10, 0.260 0,000 1
II, 0,575 8,008 1
48, 1,150 0,800 1
41, 1,138 0,880 1
40. 2,310 1,080 )
40, 2,120 0.000 (
40, 1.260 1,888 I
41, 1,595 1,800 1
40, 1,161 1,881 I
41. 1,728 1,008 1
44, 2,270 0,000 1
10, 2.471 1,0m 1
10, 1,468 0.000 1
tl, 1,388 1,103 (
tl, 1,515 1,180 I
II, 1,710 1,880
tl, 2.351 1.080
t8, 2.128 8,001
68, 1,215 0,«B0 (
t8. 1,565 1,0110 1
tl, 1.1'" 1,8(8 1
10, 1,768 l,Bjigl 1
tl, 2,4*8 0,808
tl, 2,13* 0.8110
68, 1,26* 0,001 !
10. 1,548 0.RB8
II, 1.175 8.001 I
tl, 1,710 1,808
tl, 2.261 a.naa t
tl, 2,468 0,804
tl, 1,290 4, ana 1
tl, 1.575 *.9e8
tl, 1.158 0.8311
t8, 1.718 0.»«0
tl, 2.318 .Kxa
tl, 2,128 ,8fl«
11, 1,248 .""'
64, 1,545 ,•»*
tl, 1,168 ,tet
it, l.»2* ,841
tl. l.2*« API
tt, i.«'i ;«»»
tl, !.•«• «.««
1.145 1.
8,575 2,
1.165 1,
1.508 |,
,»«8 10,
1.148 1,
1,215 3,
1,541 6,
1,861 1,
1.158 12.
l.«48 15.
1. 118 1.
1.221 3.
>,455 7,
1.618 It,
).675 14,
1,118 17,
1.878 2.
«.145 4.
1,218 9,
>.4JS t3,
1. 565 17,
>.738 21.
1.868 2,
1,128 S,
1,220 1,
1.355 13,
l.«48 16,
1.558 22,
1.518 26,
>. 145 3.
1,375 t.
.165 16,
,548 21,
.618 27,
,148 4,
1,245 1,
1,568 16,
1,868 24,
.150 32.
.448 34,
1,118 4,
1,228 9,
1.455 19,
1,619 21,
1.475 35,
.118 44,
1.870 5,
1.145 It,
1.210 23.
1.415 11,
1.565 41.
1,710 54,
1.868 6,
1.128 12,
.228 22,
1,355 35,
1.448 16.
1,550 56,
1,510 6*.
I.US 7,
>.37S 13,
.165 38,
,540 49,
.410 62.
1,140 9,
1.245 18,
1.560 17.
1,460 56.
,15.0 74.
,448 17,
1,110 10,
1.220 22.
1,455 44,
1,610 65.
1.175 42,
I. 110 101.
1,870 11,
1.145 26.
1,218 52,
l,4!5 77,
1,565 |0*.
I.7I0 12).
>.«l«,8 1).
>.I28 21,
'.228 51.
1.155 18,
1 • *•)• 1 uC
• ™ !•* » *
••**• 118,
».'•"• 156,
1.2 0, 3,6 1.1 4.1
2,1 1, 7,4 1.5 1.
7,1 0, 21.5 21.1 11.
4,5 1. 26.1 37.6 27,
12.5 8. 24,6 44,| it.
1,4 0, 1,5 1,1 1.
3,1 8, 1,2 l.t 1.
4,6 8, 6,1 7,2
18, 0, 1,5 10,4
14. 0, 10. S 13,4
11, 0. 11.0 17,4
1. 1. 0,9 1.0
3, 0, 1,6 1,1
7, 8. l.S 1,7
12, 0, 4,1 5,1
16, 1, S.I 7,1
21. 1. 7.8 1,7
2, 0. 1,1 8,3
4, 0, 1,6 0,1
9, «. 1.2 1,3
14, 0 1,7 1,9
2«. 0 2.2 2.5
27,
2.
4.
9,
15.
21,1
21,1
37,'
2.1
5,1
17, i
23, (
31.3
s,<
7.<
16,1
25,
35,
43,
4,
9,
18.
29,
38,
58.
5,
11.
22.
35,
4l!
64,
5,
11.1
-I.I
J6,'
51,!
67,1
46,1
t,
12, <
37, <
50,
t5.
1.
16. '
34. <
53.1
75.
92.
9,
19.'
40, <
61, ,
12.'
107.1
II, i
23.
41. «
74,"
102,1
134,,
11.
25,
46.1
'4l«
la. •
'•4.1
»«l.1
14*, <
• 2,6 1,1
t.
1.
12.
IS.
1.
1.
1.
4.
6,
t.
1.
1.
1.
1.
2,
2,
8 8,2 1,2 1.
* 1.1 0,4 0.
8 0,7 0,7 1,
8 1,1 1,2 1,
1 1.2 1,1 1.
1 1 1,5 l.t 1,
I 1 1,1 1,1 I,
8 11,1 12.7 13.
> • 29,5 27,1 26,
! 8 15,2 11,1 1.
1 1 106,1 111,9 1,
0 117.2 140.7 125,
1 8 1,1 5,5 0,
» 12,7 11,6 12,
> • 24.4 23,2 21,
8 i;,6 11,3 32,
0 41,6 41,2 43.
8 51,6 56,1 54,
0 1,5 1,1 1,
8 6,4 5,1 6,
8 11,8 12,1 12
8 11,9 11,8 11
0 21.8 22,5 22.
1 27.9 21,1 11
8 1.1 1,8 1
8 2,4 2,1 I
8 4,6 4,1 4
0 k-7 6.8 6
r, • •, • ™
0 s.f, 0,f 1
0 10,4 9,5 19
0 1,1 8,7 1
* ',5 1,1 1
0 ;,7 2,4 2
8 1.1 1.1 4
> 0 4,8 4,2 5
' * 5,1 5,1 5
0 5,4 4,6 1
I • 44,2 40,1 19
• * 11.3 a*> T •
* • t * •*» « f 9
8 261,9 246,4 182
• 129,5 181,7 0
' 8 163,1 449,6 11*
1 • 11,8 11,1 IS
8 11,4 17,7 15
1 < 7S.6 75,6 66
' « 104,8 186,5 14
• 129,0 140,5 US
' • 159,9 112,1 151
• !•,' ll.l 11
! • 19.1 11.9 21
1 • 40,4 19,6 14
' • 51,5 59.8 54
1 • 71,1 71,6 64
» 0 16.5 92 • ••
• "», * ^« .
0 l.S 1
•* 31
10 714
_ »3 ••
" 14,2 11,
0 >fl Jb t •
• «»!,» It,
• 27.2 26.
0 la 5 ,,
*«•* 91,
1 > < 9
„ 2,5 2,
0 4 f t
* . * 4 .
0 • « •
° 1.5 1.
* 13.2 12.
• 14.9 11.
1. 16, « 14,
*. 16.6 |5,
Jm
m
981
I
5 M
1* M
Z 2%

































































Ji
3 *
• •
*
_ *
9 '
< i?
« It
* u
* tr
                           62

-------
    TABLE 21.  H2S  IN AQUEOUS BUFFER SOLUTIONS, COMPARISON OF CALCULATED AND MEASURED DATA OF SHIH, ET AL.
                                                                                                          26
Co
Partial Pressure, mm Hq
„ _ Holes/Kg of Soln.
pH
7.00
7 00
/ " on
7.00
7.00
7.00
7,90
7.83
7.80
7.77
•c NH,
60
S00.
IZ0,
140.
tes.
•0.
140'
.l>00
.HUB
,080
,B0B
,000
,1100
,000
|fe«. 0,000
115, 0.M00
CO,
.000
,000
,0P0
.000
,000
0,000
0,000
0.000
0,000
0,000
0,000
0,0(10
H,S
0,010
0,010
0,010
0.010
0,010
0,010
0.010
e,ei0
8,010
0.010
0.010
0,010
H,S
VK API
31.3
60,1
61, B
6a, a
86,4
136.7
10.1
11,5
12,4
15,0

0.0 42.2

MEAS
84. e
98. 0
104. e
ur.0
125.
133,
16.
22!
25,

32.1

-------
TABLE 22.  NH,-C09-H90 AND C09-H,0 SYSTEMS, COMPARISON OF CALCULATED AND MEASURED
             3   C.  C        C.  f.      -i                               yn

           DATA OF VAN KREVELEN, ET AL.  AND DATA FROM LANGE'S HANDBOOK^
PtrtUl Pressure, m Hi
Tump MoUi/Kfl of Soln. NH3 CO,
*C NH3_ COj, MjS n API. MEAS VK API HEAS
II, 0,410
20. 1.960
20, 1,990
20, 2, «il>
«•, a, tie
48, 0,*R0
B0 0,liU0 B,
,440 0,B«I0 II,
,543 0,1100 |0,
,)I6 0,0*0 3,
,257 0,000 4,
,t«9 0,000 5,
,623 0,000 S,
,5U 0,000 I,
,026 0,000 |2,
,S«0 0,000 2»,
,0«8 0,000 22,
,319 0.0U0 »,
,19! 0.000 43,
,3b4 0,000 10,
,670 0.0U0 tl,
,J7B 0,000 ai,
,696 0,000 54,
.998 0,000 34,
k8, 2.000 1.3J6 0,008 i7,
0, 0,000 0,076 0,000 0,
20, 0,000 e,Bi» 0,000 0,
40, 0,I>U0 0,024 0,000 0,
J.
9,
12,
12,
2,
4,
S,
4,
7,
13,
30,
!«•
»,
39,
27,
10.
61,
54,
33.
•»,
a,
a.
• .
60, 0,000 0,0i6 0,000 0,0 0,
0, 0,
'. 0.
n. 0,
12, 0,
e. fa.
4, a.
5. 4t,
0, 45,
9,0 Jfl,
II, S 25,
0,0 2,
21.5 3.
7,0 101,
4«,0 «.
29,2 23,
12.0 220,
95.0 4.
61,0 25.
36,0 94,
16,0 379,
a. a 0,
0,0 0.
0,0 0,1
B..
*,
0.
0,
2'i
9,
33,
37,
!«•
21,
2,
3,
95,
4.
22.
203,
4,
23,
90,
355,
»95,
791,
» 718.
0,0 0,« 660,
I 0,
0.
«.
a.
2T.
10.
e.
46,5
JJ;s Van Krevelen
2!
3,
93.
4.
20.
215.
5 .
2'.
66.
394,
T60,
T60,
760.
T60.

Data








Lange's Handbook


-------
TABLE 23.  NH3-C02-H20 SYSTEM,  COMPARISON  OF  CALCULATED AND MEASURED DATA OF OTSAKE, ET AL.
                                                                                          22

Temp Holei/Kg of Soln.
J£_ 113_ £22_ HjS VK
20, 0,990 0,820 B.PBB 0,
20. 2.111 W.298 0,1100 |7,
20, 1.120 1,411 0,000 0,
20. 4,|S7 0.SB0 0.000 29.
20, 2,877 1,94} 0.MB0 2,
20, 5,590 2,836 0,0(10 7,
20. 7,1187 1.9|9 0,000 8,
40, 8,712 0,761 0,000 111,
40, 10.714 2.579 0,000 ||6,
•B, 2.719 1,134 B. 000 16,
68. |,al4 1,261 0,000 14,
68, 2,866 0.918 0,000 77,
6«, 5,790 2.200 0.000 111,
60, 6.060 2.004 0.HH0 117,
60, 6,794 3,611 0,000 60,
60, 7.8M 1,495 0.0H0 |B4,
»B, 8.291 2.818 0,000 166,
•0, 2,824 1,218 0.U0I 121,
10. 3.8«6 0.816 0,000 2H2,
IB, 4,a9| 0.405 0.0UI 416,
10, 4,486 1,561 0,000 224,
II, 7,152 0.564 0.000 585,
111, 2.601 1,102 0.000 176.
100, 4,891 0,407 0,000 765,
til, 4,474 1,563 0,000 407,
100, *.1M «.916 0,000 121,

NH}
API
B.
21.
B,
41,
2,
II.

30s!
277.

M!
>2.

mi
87.
159,
269.
122,
305.
527.
252,
885.
185.
924,
473,
H99,
Partial

HEAS
B.1
20,
B,
48.
3,
M,
15,
285,
322.
19.
22,
177,
351,
364,
101,
311.
485,
108,
297,
490,
21 4,
802.
157.
921,
427,
1022,
Pressure, pro

VK
> 96,
>
f

3 ',

,
,
1,
21,
447,

3'!
21.
207,
74,
22,
223.
27.
3,
115,
2,
1203,
12,
409|
«*.
Hq
CO,
API
r 45,
I B.
86.
0.
28.
5,

B,
B,
20.
410,
22,
16,
21,
198.

22!
267,
33,

118,'
3.
1757.
j9
625,
•T,J


ME/
55.
«.
IH5,
a.
51,
7.
13.
B.
1.
22.
442,
3,
4,
1*.
2T1.
31,
20.
256.
41.
10.
180,
19,
1211.
43,
645.
1 107.


«
t
S

























-------
TABLE 24.
                                 SYSTEM, COMPARISON OF CALCULATED AND MEASURED  DATA OF CARDON AND WILSON
ON
1 CHIP .
"C
50.
50.
50,
80,
88.
80,
80.
80,
80,
8B,
80.
80,
80,
110,
110,
110,
110,
110.
120.
120,
120,
128,
128,
Ma_
5, 445
10,142
1,511
5,606
1,963
4,543
4,101
1,048
0,542
Pi, 752
0.582
2,692
15,101
2.124
0,552
0,597
5.676
2,645
2.152
0,259
0,853
0.105
0,083
C0,_
1.706
2.665
8.PI55
1.485
0,027
2, IP9
1,999
0,204
0,460
C.137
0,384
0,995
2,7«0
0,754
0.541
0,072
2.511
1,566
0,094
0,231
0,426
0.011
0,828
Mjfl
2,689
3,711
0,172
3.000
0.165
1,774
1,487
0.083
0,093
0.499
0,049
1,117
3,769
1,155
0,115
0,166
1,281
0,786
0,166
0,138
0,306
0,022
0,B83
Partial Pressure, mm Hq

VK
15,5
67,8
49,8
72,4
211,7
30,7
29.5
86,6
0,8
13,5
14,3
42,2
579,9
41,6
0,0
98,4
249,6
«2.2
600,6
0,0
34,7
25,9
e,e
NHj
API
22.3
122,3
49,2
108,5
195,9
52,2
47,4
73,8
4,3
15,3
13.1
49,5
1425,6
112.3
17.7
90,5
409,7
118,8
630,1
12.0
76,6
23, a
«•*
CO, H,S
MEAS VK AP£ MEAS VK A£I_ MEAS
30,
263.
•«,
164,
164.
»5,
127,
1U2,
B.
T,
29,
102,
2176,
285,
15,
10P>,
353,
563,
734,
l«.
93,
1*.
195,1 309.0 239,9 594. 1286,2 1132,2
v 25,2 62,6 54.3 > l|4, 442,3 540,8 ,
aj 0,3 0,3 2,33; 2, 2,4 . s.sa;
, 975,4 1871,1 1613,0 x 1370. 4740,7 3960,2 N
aj 0,6 0,8 7.6aJ 7, 6,6 20,43;
4152.7 4706,4 3453,6 1215, 4267,5 2585,
4024,3 4362,3 3458,7 1117. 3519,8 2517.
26,0 31,9 65,5 12. 13,2 18,
0,0 2656,3 2436,6 0, 351,2 246,
312,5 275.7 346,9 1040, 767,1 589,
640,2 647,7 453,4 57, 60.3 47,
1090,7 1319,0 987.5 968, 1720,9 1354,
35,1 121,0 149,9 60, 626,6 847.
11867.4 7471.6 5976,5 7736, 5470,7 4616,
0,0 10589,9 9|45,7 0, 843,8 8ft.
103,4 176,2 169,1 160, 155,5 172.
» 5724.5 10211,210040.1 320, 3741.9 4425,
3; 26855,9 15998,6 9988,4 2594,0 4186,2 72(07.
25,1 54,6 103,4 48,2 49,2 67,
0,0 7296,7 6860,6 0,0 1269,3 1623,
\ 9596,0 6208,4 5619,8 3036,6 1502,6 1483,
d' 26. a 50,2 SB, 3 24,3 22, a 22.
46,6 e.B 899,5 752. a 0,0 787,3 563,
        a) These  points  ignored  in computing averages because they represent extreme deviations which
           are probably  due  to major error  in the measured values.  In the third and fifth runs, the
           entire run was  suspected, so the entire run was ignored.

-------
TABLE 25.
SYSTEM, COMPARISON OF CALCULATED AND MEASURED DATA OF BADGER AND SILVER
" -

Temp.

28,
20,
28,
28,
20,
20,
20,
28,
28,
20,
28.
28.
28,
28.
28.


Moles/Kq
ffis-
.189 ft
.191 0
,390 0
,B«5 |i
,380 0
,W97 VI
.192 0
.196 0
.192 0
,193 0
,«92 0
,088 0
.1)95 0
,088 0
.095 0


of Soln
iz_
,410
.'90
,195
.500
.638
,660
.670
,686
.700
,T45
,770
,790
,800
.815
.816


t
HgS
0.189
B.194
0,390
0.D«5
0.360
0,097
0,192
0,196
0,192
0.193
0,092
0,066
0.095
0.066
0,095


NH3
VK API.
3, 4,0
2. 3,0
?, 3.0
2, 2.8
1. 1.7
1, 1.5
1, l.«
I, 1.3
1, 1.2
0, 1,0
0, 0,8
8, 0.7
8, 0,7
8. 0,6
8, 0.6



MEAS
4.1
2,9

2j7

r,3

U3
0,9
0,9
0,7
0,0
0,0
0,5
0,b

Partial

VK
l.«
j »
42
3,6
13,2
I •, 1
16,0
18,6
20,6
30,2

• 3,4
47,6
55,1
56',9

Pressure.
CO,
API
I,T
3,*2
3,9
3.0
12,1
It,*
13,6
15,0
IT.9
25,7
31,3
36,1
00,3

46^4

no Hq

MEAS
1.5
3,5
37
3.7
13,1
12.1

19*0
20,5
29,2
35,2
42,4
45,1
0,0
0.0



VK
3.6

I3*,7
1,1
26,9
6.1
13.9
15.8
16,6
22,1
11.4
12,5

15,2
16,9


H,S
API
3,6
5,4
1«,1
1,1
26,7
5,7
13,1
15,0
15,7
20,'S
10,
11,
13,
13,
i«;



MEAS
3, .

ie'.
" 0.
27,
9,
II,
15,
15,
27,
12,
8,
0.
13,
16.

-------
        TABLE  26.
SYSTEM,  COMPARISON OF  CALCULATED AND MEASURED  DATA  OF BREITENBACH AND PERMAN16'23
                                                                       PtrtUl Prestur«. an Hg
oo
                                     Temp.
                                     21.
      20,
      2«,


      20'
      201
      20,

      20|
      20,
      20,
      20,
      29,
      40,

      40,'
                                     68,
                                     68,
                                     60,
                                     60,
                                     *»,

                                     "I
                                                Holei/Kq of Soln.
 1.174
 1,468
 1.762
 2.149
 2.916
 4,«04
 S.872
 a.a«a
11,744
14.680
IT.616

29|}40
15,212
41,104
46,976
 1.762
 2,916
 5,872
11,744
IT,616
29.160

 0!s87
 0,940
 1.468
 2,916
 4.404
 5,872
 8,801
11,744
.W08
,»Dfl

!e«0
,000
,000
,000
,000
,000
,000
.000
,000
.000
,f)R0

IHBB
.(100
                      tvnv
                      ,000
                      ,000
                      ,000

                      |ae0
                      ,000
                      ,000
                      .800
                      ,000
                      .001
,000

!o00
,000
,000

,000
.000
,000
,000
,000
,000
.000
.0U0
,000
.000

1080
.VH0
,0110
,000
,080
.000
,000
,000
,000

|000
,0tJ8
,000
,000
VK
12,
14,
l>,
22,
21,
17,
•5,
»».
65,
69,
79.
66,
5',
50,
42,
]4 ,
44,
69,
II',
167,
(79,
151,
108,
16,
51,
86.
159,
219.
268,
1«0.
181,
API
12.
H,
20,
28.
It.
*2,
91,
177,
100,
476,
724,
1551,
1117,
6011,
11274,
20101,
•T,

20?)
608,
1115,
4769,
14108,
Jl ,
51,
82,
177,
287,
411,
720,
1119,
MEAS
12,
15!
18,
24,
11,
58,

114 ,
166,
227,
298,
478,
6»6,
945,
1178,
1458,
fl 5 •
76,
167,
195,
612,9
1520,0
2760,0
10,2
«8,7
11,
I6S,
261.
161,
581.
114,

-------
       TABLE 27.  NH--C00-H0S-H,0 SYSTEM, COMPARISON OF CALCULATED AND MEASURED DATA OF VAN KREVELEN, ET AL
ON

Partial Pressure, mm Hq
Temp Moles/Kq of Soln. NH3 CO, H,S
*C NH3_ <%_ HjS VK API MEAS VK API. HEAS VK APJ HEAS
20, 1.140 0,4t0 a. 188 5,5 3.* 0
20. i,2«e 0,750 0.29t 0,7 B.B n
20, 2,160 0,910 0,160 3,T 4,0 0
20. 2.158 0,4M0 0,600 8,9 9.9 0
20* 2.250 i,4na d.210 1.7 1,9 0
20, 0,790 0,250 0,180 2,6 2,6 0
• 0, 1,178 0,1110 0,184 10, 6 |0,3 0
40, 1.140 0,410 0,180 IB, 0 9.7 0
40, 1,130 0,210 0,290 14,1 IJ,
»0, 2,160 0.940 B,3hB 11,
48, 2,1)0 0,400 0,bt)0 23,
40, 0,700 0,104 0.3S0 5,
40, 0,790 0,250 0,180 7,
40, 0.740 0,380 0,160 3,
40, 2,230 1,4100 8.210 5,
60. 1,170 0.4i<| 0,184 26.
60, 1,140 fl.4|0 0,180 24,
60, 1,130 0.210 0,290 33.
60, 2.160 0,950 0,360 28,
60, 2.1*0 0,400 0,6(90 15,
60, 2.250 1.340 0,200 18.
60, 0,700 0,104 0,3)0 13,
60, 0.790 0.250 0,160 17,
60. 0,740 0,360 0,150 10,
60, 1,020 0,620 0.140 9,
60, 1.2«i 0.6S6 0,124 19,
69. 1,2»9 0,643 0,214 IS.
Hi
24.
5,
6.
3.
6,
24,
zz.
30.
29.
55,
1«.
12.
l«,
*.
>.
1«,
"•
0
0
0
n
0
0
«
0
0
0
0
0
0
0
0
0
0
0
0
0 2.1 t.
0 44.7 38,
0 6.4 5,
0 8,5 0,
0 30,7 26,
0 1,7 I,
0 10,2 8,
0 11,0 9,
0 3.1 2.
0 30,
0 3,
0 3.
0 a.
0 42,
0 128,
0 43,
0 4fc.
0 14,
0 131.
»T,
S60,
*>.
36,
»»«,
265.
2^.
^ *
4,
7.
35,
109,
«J.
46,
!"•
137,
1"».
352,
22,
36,
114,
265.
0 121,3 116,
0 177,2 174,
0
0
0
0
0
0
e
0
0
0
0
0
0
0
0
0
0
0







0
0
0 3,7 3,7 4.2
0 48.2 44.0 49.0
0 13,4 12,
« 7,3 7,
0 19,7 16,
0 3,7 3,
0 11,3 10,
0 11,6 ID,
0 12,0 11,
0 33,7 35,
0 24,8 24,
0 33,1 30,
0 12,5 11,
0 29,3 26,
0 40,5 43,
0 29.3 32,
0 30,0 33,
0 34,1 35,
0 74,7 107,
0 67,2 79,
0 59,3 95,
96.6 94,
14,8
8.4
10.11
4,5
11.7
12,3
12,1
36,4
27,0
32,3
13.3
27,8
37,6
29,5
30,0
32,8
64,4
77,5
59,5
80,4
34,| 35,3 32.0
55, « 58,3 53,5
70,0 79,2 63.7
29.7 35,9 25.8
80,3 96,4 77,0

-------
      TABLE 28.   SUMMARY  OF REFERENCES  TO  EXPERIMENTAL DATA
Reference
    15
    16
    17
    18
    19
    20
    21
    22
    23
   24
   25
   26
 Type of Data
NH3-C02-H2S-H20
   50°C  to 120°C
NH3-C02-H2S-H20
   20°C
NH3-H20
   0°C to 60°C
                       97°C  to  147°C
NH3-C02-H2S-H20
  20°C
NH3-C02-H20
  20°C to 40°C
C02-H20
  0°C to 60°C
 80°C to 150°C
NH3-H2S-H20
 80°C to 120°C
NH3-C02-H20
  20°C to 100°C
NH3-H20
  0°C to 60°C
NH3-H20
  0°C to 60°C
NH3-C02-H20
  20°C to 40°C
H2S-H20
 (Buffered)
  80°C to 185°C
Used in Correlation
        yes

        yes

        yes

        yes

        no

        no

        yes

        yes

        yes

        yes

        yes

        yes

        no

        yes
                             70
                                                 (continued)

-------
                            TABLE 28.   (continued)
             27                NH3-C02-H20                    no
                                70°C to 120°C
             28                NH3-C02-H20                    no
                              Phase diagrams
                               60°C to 170°C
             29                NH3-H2S-H20                    yes
                                20°C to 60°C
              1                NH3-C02-H20                    yes
                              NH3-H2S-H20                    yes
                              NH3-C02-H2S-H20                yes
                                20°C to 60°C
             30                NH3-H20                        no
                                114°C to 317°C
             31                C09-H20                        no
                                270feC to 550°C
             32                H2S-H20                       no
                               160°C to 330°C
             33                NH3-H2S-H20                   no
                                70°C to 90°C
             34                H2S-H20                       no
                                25°C
             35                H2S-H20-Salt                   no
                               150°C to 330°C
             36                H2S-H20                        no
                              H2S-CH4-H20
                               71°C  to  140°C
             37                NH3-C02-H20                    no
                               60°C to  150°C
    More references are given  in I. Wichterle, J. Linek, and E.  Hala, Vapor-
Liquid Equilibrium Data Bibliography,  Elsevier (1973).  The references listed
above are ones for which copies of  the  data  have been obtained.
                                     71

-------
         TABLE  29.  SUMMARY UF DEVIATION ERRORS BETWEEN CALCULATED AND
                   MEASURED AMMO?!IA PARTIAL PRESSURES
      Partial
     Pressures
   in Table no.

        18
        19*)
        22
        23
        24
        25

        26 b)
               Temp
                °C
     Van Krevelen
    vs Meas.  Data
No.  pts    Ave.  Error
                80         6
               120         6
                60         6
               100         4
               140         4

                20         5
                40         4
                60         4

                20         4
                40         7
                60         9

                20         7
                40         3
                60         7
                80         5
               100         4

                50         3
                80        10
               110         5
               120         5

                20        13

                20        16
                40         7

         overall average
         ignoring data in
         Tables 23 & 30
               60
               29
              387
              130
               33

              184 J]

              107a
               60a'

               15
                8
                8

               45
               90
              139
               14
               15

               98
              126
              115
               73

                8

              2425!
              223.b)

               72
       API SWEQ
     vs Meas. Data
No. pts    Ave. Error
   10
    8
    6
    4
    4

    5
    4
    4

    4
    7
    9

    7
    3
    7
    5
    4

    3
   10
    5
    5

   13

   16
    7
 14
 17
  8
 18
  5

329 a!

15°a
 79a}

  3
  8
  8

 11
 10
 81
 10
  9

 43
 77
 35
 30

 12

128^1
Ji'
 24
a)

b)
These data appear to be of rather low quality and  should  be given very
little weight.

These data are in disagreement with data  for  NH^-H^O  in Table 22 by
Clifford and Hunter.  From our evaluation we  believe  the  data of Clifford
and Hunter to be more nearly correct,  and more weight has been given to
their data.
                                     72

-------
       TABLE 30.  SUMMARY OF DEVIATION ERRORS BETWEEN CALCULATED AND
                  MEASURED CARBON DIOXIDE PARTIAL PRESSURES

Partial
Pressure
in Table no.
Van Krevelen
Temp.
°C
vs.
No. pts.
Meas. Data
Ave. Error%
vs.
No. pts
API SWEQ
Meas. Data
Ave. Error%
    22          20         4
                40         8
                60         9
                 0
                20
                40
                60
   23
      a)
 20         5
 40         3
 60         7
 80         5
100         4
   24           50         3
                80         9
               110         4
               120         3

   25           20        15
          overall  average
          ignoring data in
          Table 27
                          13
                           8
                           9
102
1288
 92

 63
 63
 98
148

  8
 35
 4
 8
 9
 1
 1
 1
 1

 7
 3
 7
 5
 4

 3
10
 5
 5

15
  5
 13
 10
 18
  4
  6
 15

117*
 66a\
                                                   23
                                                   29
                                                   20
                                                   25

                                                  J2
                                                   17
a)  These data appear to be of rather  low quality and should be given
    very little weight because deviations are large and have little
    apparent pattern.	
                                   73

-------
          TABLE 31.   SUMMARY  OF  DEVIATION ERRORS BETWEEN CALCULATED AND
                     MEASURED HYDROGEN SULFIDE PARTIAL PRESSURES

Partial
Pressures Temp.
in Table no. °C
18 80
120
150
80
120
19a) 2Q
40
60
20 20
40
60
21 80
100
120
140
160
x 185
24 50
80
no
120
25 20
27 20
40
60


No.

__
—
6
6
5
4
4
30
30
30
__
—
__
—
—
—
3
9
4
3
15
6
9
12
overall average
ignoring data
Table 23.
in

Van Krevelen
vs. Meas. Data
pts. Ave. Error%

__
—
22
44
787"!
225a
23a)
10
6
9
_ _
--
—
--
-_
—
211
139
188
46
10
12
6
8
24



vs.
No. pts
4
3
3
10
8
5
4
4
30
30
30
2
2
2
2
2
2
3
10
5
5
15
6
9
12



API SWEQ
Meas. Data
Ave. Error%
5
2
13
13
44
539!)
182a{
5a)
14
8
12
50
67
75
70
44
16
13
47
23
18
11
12
10
20
18


,
a)  These data appear to be of rather low quality.   Large and unreasonable
    adjustments would have to be made in the correlation model  to correlate
    these data; therefore they were ignored.
                                    74

-------
         Table                          Temp.         Ave.  Error.  %
          No-        System           Range. °C     NH^    CO: - Jg[
          20    NH3-H2S-H20        20°C to 60°C    ..... -      8

          22    NH3-C02-H20        20°C to 60°C      9      9

          25    NH3-C02-H2S-H20    20°C              8      8     10

          27    NH3-C02-H2S-H20    20°C to 60°C    ......      8

 From this  comparison, average errors are -about 10% or less in these tables-
 however only data from 20U to 60UC are compared,  llhen higher temperature
 data and other literature data are compared the agreement  is not  as good be-
 cause of extrapolation errors.  These comparisons are given in Tables  18,  19
 24,  and 25 where deviation errors can be summarized as follows.

                                        Temp.          Ave.  Error.  %
                     System           Range, °C      NH^
           18    NH3-H2S-H20        80°C to 120°C    139     —     58

           19    NH3-H2S-H20        20°C to 60°C     122b)   —    379b)
           23    NH3-C02-H20        20°C to 100°C     65b)   105b)  —

           24    NH3-C02-H2S-H20    50°C to 120°C    108     84    146


 This  comparison shows that deviation errors are baout 3  to  15 times higher
 than  for the systems from which the correlation was  derived.  This conclusion
 doesn't change significantly even when suspected data noted at the bottom of
 this  summary are ignored.  Thus, it is concluded that the Van Krevelen model
 does  well  at temperatures from 20 C to 60 C which is the region from which i£
 was derived, but its accuracy is much poorer at temperatures from 60  to 120_
 which is the range of commercial interest for sour water strippers.  Beychok
 has recently proposed an NH, Henry's constant published  by  Edwards et al
 which improves predicted MM, volatilities.  By this  method, the average error
 for NH3 at 80°C in Table 29Jis reduced from 60% to 23%.

 Evaluation of SWEQ Model

      The SWEQ model  has the advantage that both low- temperature and high-
 temperature data were used in developing the model,  thus it would be expected
 to give better results than the Van Krevelen model.   A comparison of the API
 SWEQ  model  with the Van Krevelen model  and with experimental data is given in
 Tables  18 to 27 under the heading API.   At some of the conditions in these
 tables,  direct comparison with the Van  Krevelen model  is not possible because
 the Van  Krevelen modes does not permit  the calculation of equilibrium data at
 NHo/H9S  ratios less  than 1.5 or for NFL/total  acid gas ratios less than unity.
 This  condition occurred in the following cases.

     b 'Measured  data  in  Tables  19  and 23 are  believed  by the author to be un-
reliable.  Data  in Table 19  deviate  radically.   Large  and unreasonable adjust-
ments  would  have  to be made  in  the model  to  correlate  these data.  Data in
Table  23 exhibit  large deviations  with  little  apparent pattern.

                                      75

-------
              Table                           Temp.
               No.        System            Range, °C

               ]8     H2S-H20             80°C to 150°C

                      H2S-NH3-H20         80°C and 120°C

               21     H2S-bufferrH20      80°C to 185°C

               22'    C02-H20             0°C to 60°C

               24     NH3-C02-H2S-H20     80°C to 120°C
                        (4 points)


       A comparison of the SWEO model  with the Van  Krevelen model  at low
temperatures where the Van Krevelen model  was derived gives the following
results.

lable
No.
20
22
25
27



System
NH -H S-H 0
WM _rn _H n
NH3-CO?-H2S-H20
NH-,-CO -H S-H 0

Temp.
Range
°C
?n fin
20-fiO
20
20-60

A\
NH^
VK SWEQ

Q 7
8 12

te. Error "/,
CO?
VK SWEQ

Q in
8 12

f
9
H'
VK
0

10
0


?$
SWEQ
n
1 1
n
i R
1 U
                           Overall Ave. Error     9   10    9    11     9    12

This comparison shows that the SWEq model does about as well as the Van
Krevelen model except that the Van Krevelen model appears to be slightly
better.  The overall average error from the Van Krevelen model is about
9% while the SWEQ model gives about 11%.

     The picture changes  considerably when high temperature data  are  compared
as follows.
                                      76

-------
                                Temp.	Ave. Error %
                                Range     ~lfi^       CO^	[O	
                                          ™-J™  _^_ __2
     18     NH3-H2S-H20        80-120     45   13  —  —   33   27

     19     NH3-H2S-H20        20-60     122a) 197a) ......  379a) 265a)

     23     NH3-C02-H20        20-199     65s }  29a) ^  9#> —  —

     24     NH3-C02-H2S-H20    50-120    108   53   84   24  146   31
               Overall  Ave. Error         77   36   84   24   90   29
            (ignoring Tables 23 & 27)

    a  Measured data in  Tables 23 and 27 are believed by the author to  be
      unreliable.   Data in Table 23 deviate radically.  Larqe and  unreasonable
      adjustments  would have to be made to correlate these data.   Data  in
      Table 27 have large deviations with little apparent pattern.

     This  comparison shows  the  SWEQ  model  to be superior to the Van Krevelen
model at high  temperatures  with deviations  averaging  about 84% for the Van
Krevelen model compared  to  about  29% for  the SWEQ model.  However, the pre-
dicition accuracy  is still  not  as  good  as  at lower temperatures.  This can be
partly explained by  the  fact  that  much  higher concentrations of the components
were studied  in Tables 18 and 24  at  high  temperatures compared with concen-
trations in Tables 20, 22,  25,  and 27 at  low temperatures.  Concentrations
up to 14 moles/Kg  of solution are  covered  in Tables 18 and 24 while concen-
trations to only about 3.5  moles/Kg  of  solution are covered in Tables  20,  22,
25, and 27.   These higher concentrations  place very high demands on the SWEQ
model and  makes actual correlation of the  data more difficult.  Besides this
problem there  is the normal  scatter  expected from the data due to measurement
errors.  It is not possible at  this  point  to say which errors are correlation
erros and  which errors are measurement  errors; however, it is possible to con-
clude that the sum of both  errors  is on the order of  about 29% for the SWEQ
model compared to  84% for the Van  Krevelen model.

     Another  test  for accuracy  of  the SWEQ model can  be made by comparing
mean deviations in p     /p   lc  where the mean value is calculated as follows.
             P
             meas

             calc/mean  value
i  = 1,N
         P
          meas
          calc

1/N
              (43)
            where N =  number of  points averaged
This comparison will show any bias errors that may exist between calculated
and measured data.  As in the case of average errors, the bias errors can
result from either correlation bias or from bias in the experimental data.
These comparisons for the SWEQ model are given in Figures 6, 7, and 8 for
                                     77

-------
NH-, C0«, and H9S respectively.  Figure 6 shows that mean ratios of Pmeas/Pcalc
fof NH. lie primarily above unity with only the data of Badger and Silver
and of Breitenbach and Perman lying below unity.  The amount of steam required
in a sour water stripper is primarily determined by the volatility of NH~.
For this reason, available NH3 volatility data from the literature from vari-
ous authors are compared in Figure 6 in an attempt to obtain a reliable
volatility correlation.  The author suspects that the points below unity are
in error and that the true NH3 partial pressures are slightly above unity.
The scatter between various authors primarily represents bias in their measured
data, but a line of unity which falls below all of these points probably
represents correlation bias.  This correlation bias in the case of ammonia can
be easily adjusted so that measured data will scatter symmetrically both above
and below unity; however, the author is hesitant to do this without further
justification from other data.

     Figure 7 shows that mean ratios of Pmeas/P aic for H^S appear to fall
nearly symmetrically both above and below OnTty, so againf the scatter between
the points probably represents bias in measured data between the various
authors.

     Figure 9, 10, and 11 show similar plots comparing mean ratios of
pcneas/pcalc from the Van Krevelen model for NH~, C0?, and H2S respectively.
These plots show wider scatter than it obtained from the SWEQ model.   The
difference has to be due to correlation bias.  This result would tend to infer
that there could still be correlation bias in the SWEQ model which has not
been identified.  If such bias exists, it has to be on the order of the
deviations appearing in Figures 6 to 8 or less.
                                     78

-------
     o


     re
     c
     re
     O)
       o
      re
      o
       re
       O)
10.0
9.0
8.0
7.0-
6.0
5.0
4.0
3.0
2.0

i.oJ
0.9
0.8
0.7
0.6
0.5
0.4*
0.3
0.2
0.1

_ 1 1 1 1 1 1 	 1
I O Miles & Wilson
- D Clifford
- A Van Krevelen NH3-C02
- Vcardon & Wilson
• Badger & Silver
Ofireitenback & Perman
-
V V
k A zn o
• -
-
: o
?
—


• i i i i _
—

-
_

D v X

_
—
-
-
1 1 1 1 1
               20  30   40   50   60   70  80   90   100 110  120 130  140 150


                                       Temperature,  °C




Figure €.  Ammonia mean ratio of measured over calculated partial  pressures

           based  on  SWEQ correlation.
                                      79

-------




o
fO

ro
O)
CM
O
0
ft
O
to
o
o_
CO
(T3
Ol
Q_







10.0
9.0
8.0
7.0
6.0
5.0
4.0

3.0

2.0

1:3
0.8
0.7
0.6
0.5
0.4
0.3
0.2

0.1

I I I I 1 1 1 1 1 1 1 I
_ -
_ -
m ***
_

ltm •„

-

b A $ v ^
t " v v i
^ «
-
-
- A Van Krevelen NH3-C02
OLange's Handbook
~ V Cardon and Wilson
Q Badger and Silver
i i i i i i i i i i i i
              20  30  40   50  60   70   80   90  100  110  120 130  140 150

                                    Temperature, °C
Figure  7.   Carbon dioxide mean  ratio of measured over calculated partial
            pressures based on SWEQ  correlation.
                                      80

-------

o
•r~
DC
c
03
0)
s:
CM
„
0
ro
u
Q.
10
(U
Q-E




10. n
.0
8.0
7.0
6.0
5.0
4.0
3.0
2.0

-

T-Ofl

0.8|
0.7
0.6
0.5
0.4

0.3

0.2

0.1
-
—
-
-

O
- A
V
- D
O
1
20 30



^^ V A o

O ^




Miles and Wilson
Van Krevelen NH3-H2S
Cardon and Wilson
Badger and Silver
Van Krevelen NH3-H2S-C02
i i i i i i
40 50 60 70 80 90

O

V *7

I
_
_


^
-

-

1 1 1 1 1
100 110 120 130 140 15
                                      Temperature, °C
Figure  8.  Hydrogen sulfide mean ratio of measured over calculated partial
            pressure based on SWEQ correlation.
                                     81

-------








o
•»™
4-*
c
Ol
i^1-
z
3E

•"• ""•
P.
LJ x/
O

v o a
A A ^
t~± 	 ~
0.8|- -1
0.7
0.6

0.5
0.4
0.3

0.2


0.1
_ —
_ _

- -
_
O Miles and Wilson
D Clifford
A Van Krevelen NH3-C02
V Cardon and Wilson
' Badger & Silver
O Breitenback & Perman
I i i J I i i i I i i i
               20   30   40  50   60   70   80   90  100  110  120  130  140  150

                                       Temperature, °C


Figure  9.  Ammonia mean ratio of measured over calculated  partial  pressures
           based on Van Krevelen correlation.
                                       82

-------
10.0
9.0
8.0
7.0
6.0
5.0
4.0
3.0
o
2 2-0
IO
s:
CM
8 i,rJ
0.9)
o 0.81
Ts 0.7
^° 0.6
£ O.J
0)
ex5 0.4
0.3
0.2

0.1

m
-
_
_
^,
-

_
o o
o
3 A ^ :
•B
•• _
A Van Krevelen NH3-C02
~ O Cardon & Wilson
D Badger and Silver
i i i i i i i i i i i i
20 30 40 50 60 70 80 90 100 110 120 130 140 15
                                      Temperature, °C
Figure  10.
Carbon dioxide mean ratio of measured over calculated partial
pressures based on Van Krevelen correlation.
                                       83

-------








o
to
OL
c
fO
2
CO

u
(O
u
o.
trt
(O
0)
o.




10.0
9.0
8.0
7.0
6.0
5.0
4.0
3.0

2.0




H
0.8
0.7
0.6

0.5
0.4


0.3
0.2


0.1
i i i i I I I I I i i i
. _
-
-
-
V
V



o


_ v :
-
— _

mm «
w •


- O Miles and Wilson
A Van Krevelen NH3-H2$
- V Cardon and Wilson
O Badger and Silver
O Van Krevelen NH3-H2S-C02
iii i i i i i i i i i
              20  30  40   50   60   70  80   90   100  110  120 130  140 150

                                       Temperature,  °C
Figure  17.   Hydrogen sulfide mean  ratio of measured over calculated partial
             pressures based on  Van Krevelen correlation.
                                     84

-------
Evaluation of New BYU Data

     An evaluation of new NH.-l-LS-hLO and NH.-CO,-H9S-H90 data measured
at Brigham Young University Can^proDably be BestSiafle b§ comparison of
measured data with predicted data from the SWEQ model.  These comparisons
are made in Table 18 for the hLS-H90 and NH~-H9S-H90 systems and in Table
24 for the NH--CO -H2S-H20 system/ The SWEQ m&del^predicts low temperature
data on which the Van Krivelen model is based with about the same accuracy
as the Van Krevelen model, and at the higher temperatures from 50°C to 120°C
in Tables 18 and 24, the accuracy is much better than the Van Krevelen model.
As discussed above, it  is not possible to separate correlation errors from
measurement errors; so  margins of error have to include both effects.  The
following gives a summary of the average errors between predicted and measured
data.
Table   Temp.  No. of
 No.     *C   Points
                                         Ave. Error %
                   18
                   19
         80
        120

         50
         80
        110
        120
10
 8

 3
10
 5
 5
14
17

43
77
35
30
                    Overall  Ave.  Error %  36
23
29
20
25

24
13
44

13
47
23
18

29
It is concluded from this comparison that the new BYU data in Tables  18  and
24 are consistent with literature data correlated by Van Krevelen et  al.,
with average scatter between measured and correlated data being on the order
of 36%, 24%, and 29% respectively for NH3, C02, and H2S.  Two experimental
runs given in Table 24 were ignored in computing these averages and ammonia
analyses on two additional runs were ignored.  The points ignored are noted
at the bottom of Table 24.  The reason for ignoring these points is that the
deviations are so large that the experimental points appear to be unreasonable
and probably in serious error.

     Mean ratios of Pmp,,/Pr,lr plotted in Figures 6, 7, and 8 for NH-,  CO-,
and H2S respectively snow some bias between data in Tables 18 and 24  as
follows.          Table
                   M«Y                   Comments
                   24
             H2
                                  appears okay
                                  appears okay
             NH3  at  50°C and 80°C  appears about
               40% too  high
             CO?  appears okay
             H2§  appears okay

                     85

-------
Based on this comparison, it is concluded that any bias in the measured
data is small except for NH, at 50°C and 8QOC in Table 24.  If these data
points were ignored in ccmpdting the average error above for NH-, then the
overall average error would be reduced from 36% to 24% which is comparable
to deviation erros for CC0 and H.-.S based on the SWEQ model.
                         C.      (L
Ammonia Fixation by Acids and Release by Caustic Addition
                                        /
     Little direct data appear in the literature on the volatilities of NH-,
C0?, and H9S9f-irom aqueous solutions as a function of pH»  One study made by
Shfh et air,  ' is given in Table 21 for the volatility of H2$ from buffered
solutions.  In this table, the predicted H2S pressures are consistently lower
than measured values by a factor of about 6.7.  This prediction error could
be the result of the salt concentration in the buffer solution which is not
accounted for by the SWEQ model.

     In addition to the data by Shin et a!., new measurements of pH versus
caustic addition have been made at BYL).  These results are shown as the
plotted curves in Figures 12 and 13.  These comparisons show that predicted
free NH- concentrations are also lower than measured values as occurs in the
case of HpS.  These data tend to indicate that the SWEQ model might be pre-
dicting both too low H2S and too low NHL partial pressures, but we doubt
this based on the measured volatility data of NFL and hLS examined in this
report.  With these discrepancies, calculated pH levels coul:d be in error by
+_ 0.5 unit; this is a rather large error so more work should be done to re-
solve this question.

     Ammonia fixation effects due to carboxylic acids and the release of
NH- by caustic addition are predicted by the SWEQ model as given in Table
32.  This table gives a comparison of calculated tray to tray NH_, C0?, and
HpS volatilities going down a separation column at total reflux at 20 psia
column pressure.  The initial vapor phase concentrations of NH,, C0?, and
HpS were 100 ppm on a weight basis for each component.  The first set gives
calculated vapor and liquid compositions for three trays under conditions of
no carboyxlic acid or caustic present.  In this example, the liquid concen-
tration of all three components drops to 0.1 ppm or less on the third tray.
When 500 ppm by weight or carboxylic acid is added to the liquid on each tray,
then the ammonia concentration goes up to 142.4 ppm on the third tray indi-
cating that the ammonia is fixed and is unstrippable.  If caustic is then
added to a level of 172.5 ppm in the liquid on each tray, then the ammonia is
released from the carboxylic acid and concentrations less than 0.1 ppm are
predicted for NH,, C02, or H^S in the liquid of the third tray.  If too much
caustic is added; then HpS wfll be fixed in the liquid phase; thus 500 ppm of
caustic produces an HpS concentration in the liquid phase of the third tray
of 310.1 ppm H2S.  From this table, it appears that the optimum pH for equal
volatility of NH, and H?S is about 8.5.  This pH corresponds to the hydrogen
ion concentration in the liquid phase at the temperature of the tray in the
separation column.  In actual practice, samples of liquid would probably be
taken for pH determination at room temperature.  The effect of temperature
on pH can be calculated from the SWEQ model.  Figure 14 gives a plot of pH
at 25°C and at 120°C for the addition of caustic to the mixture shown in
Figure 1.  The effect of temperature will be different depending on the
mixture, but this plot can give some idea of the effect.

                                     86

-------
00
—I
                                                                                           11
                Figure 12.  Free ammonia versus pH adjustment by caustic addition at  25°C.

-------
  125
  TOO
   75
 (D
 fD
    50
   25
                                           8
                                      PH
10
Figure 13.   Free ammonia versus pH adjustment by caustic  addition  at  80°C,

-------
                     TABLE 32.  COMPARISON  OF CALCULATED NH,, C09,  and  H«S VOLATILITES VERSUS
                                EFFECTS  FROM CARBOXYLIC ACID AND^CAUSTKTADDITION, TRAY TO  TRAY
                                FROM THE COLUMN AT TOTAL REFLUX AND 20  PSIA COLUMN PRESSURE
                   RCOOH
oo
                                           Initial Basis for Each Set
or OH"
1n Liquid
ppm wt
Tray
1
2
3
1
2
3
1
2
3
1
2
3
RCOOH

0
0
0
500
500
500
500
500
500
500
500
500
OH

0
0
0
0
0
0
172.5*
172.5
172.5
too
500
500
1s 100 ppm
ppm by wt.
in Vapor
NH,

100
5.8
0.6
100
138.8
142.2
100
4.5
0.2
100
3.9
0.2
co2
100
0.3
0.0
100
0.0
0.0
100
0.9
0.0
100
60.0
29.5
of NH,, COo, & HoS 1n Vapor

H,S
L
100
1.3
0.0
100
0.1
0.0
100
3.7
0.2
100
171.2
250.0
"" bpm by wt.
in Liquid
NH,
J
5.8
0.6
0.1
138.8
142.2
142.4
4.5
0.2
0.0
3.9
0.2
0.0
COg
3.0
0.0
0.0
9.0
0.0
0.0
0.9
0.0
0.0
60,0
29,5
11.8
H2S
1.3
0.0
0.0
0.1
0.0
0.0
3.7
0.2
0.0
171.2
250.0
310.1
Vapor/Liquid
wt Basis
NH.
•*
17
9.7
4.2
.72
.98
1.00
22*
23
23
26
26
26
C0?
370
1100
2590
5600
5100
5090
116
83
82
1.7
2.0
2.5
Ratio
H2S
78
241
620
1750
1570
1550
27*
20
19
0.6
0.7
0.8
pH
at col .
Temp.
in Liquid

8.035
7.515
7.026
6.195
6.332
6.341
8.500
8.646
8.652
10.165
10.097
10.026
             Appears close to optimum caustic addition for best NH3 and H2S volatility.

-------
                  12
                  11  -
                                              Initial  Concentrations


                                              NH0
ppm, wt
  100
                                              co2

                                              H2S


                                              RCCOH
                                              	I
  100

  100


  500
                              100      200      300       400

                                     Caustic  Added,  ppm  by wt
          500
Figure-  14.   Sample  plot of  pH versus caustic addition showing variation of
             pH at 25°C and  at column temperature.
                                    90

-------
                                  SECTION 7

                           ACCURACY OF CORRELATION


    The overall accuracy of the SWEQ model can be assessed by examination
of the error summaries in Tables 29 to 31 for NH-, C0?, and H?S respectively.
From these tables, the overall average error between measured and predicted
partial pressures can be summarized as follows.

                               Temperature Overall Ave. Error %
              Compound          Range, °C      VK    SWEQ

          Ammonia              20 to 140°C     72     24
          Carbon dioxide       20 to 120°C     35     17
          Hydrogen sulfide     20 to 185°C     24     18


This comparison shows that SWEQ module is superiorio the Van Krevelen
model.9)

     Data at low temperatures  are represented better by both models than data
at high temperatures as shown  in the following comparison taken from the
previous section of this report.

                                         Ave. Error %
                                     up  to  60*^   above 60°C
                    Compound         VK    SWEPT   VK     SWEQ

                Ammonia              9     10    77      36
                Carbon dioxide       9     11    84      24
                Hydrogen sulfide     9     12    90      29

This comparison shows  that  both models  predict the low temperature data
quite well; but at high temperature, the Van Krevelen model deviates consider-
ably from measured data, and errors  between  the SWEQ model and measured data
increase from about 11% to  about  24%.

     Users of theSWEQ  model must  be  aware  that the errors  summarized above
are average errors and that there might be regions where the correlation is
less accurate.  More experimental data  is  required before  a better assessement
can be made.

   a^This is the model published  by  Van Krevelen without any modifications.
                                      91

-------
                                   SECTION 8

                                    SUMMARY


     A new correlation model  has been developed for calculating sour water
equilibrium data at temperatures from 20°C to 140°C.   The correlating equations
in this new SWEQ have been used to obtain a computer program capable of handling
the various chemical and physical  equilibria of NHV  CCL, and HUS in sour water
systems including the effects of carboxylic acids on  NH~ fixation and release
by caustic addition.

     This new SWEQ correlation model  has been used to evaluate published and
new vapor-liquid equilibrium data and comparisons are made with the Van
Krevelen prediction equations as published by Van Krevelen.  Average errors
between calculated and measured partial  pressure data can be summarized as
follows.

                                           Ave.  Error %
                                     U£
                                     V*
                                        to 60°C    above 60°C
                     Compound        VK    SWEQ    V_K    SWEQ

                 Ammonia              9     10     77     36
                 Carbon dioxide       9     11     84     24
                 Hydrogen sulfide     9     12     90     29


     This comparison shows that both models predict low temperature data
quite well; but at high temperatures, the Van Krevelen model deviates con-
siderably from measured data, and errors between the SWEQ model and measured
data increase from about 11% to about 29%.  Comparisons with variations of
the Van Krevelen model as published by other authors have not been made.

     Vapor-liquid equilibrium measurements made at Brigham Young University
are predicted by the SWEQ model with the following average errors.


                      Compound    Ave. Error %

                  Ammonia              36
                  Carbon dioxide       24
                  Hydrogen sulfide     29

Data on measured NH~ partial pressures from NHg-CO^^S^O mixtures appear
too high by about 40% at 50 C and 80 C.  If these points are ignored, then
the average ammonia error is reduced from 36% to 24%.


                                      92

-------
     Details of the SWEQ correlation model, correlating equations, the
computer program, and evaluations of experimental data are given in this
report.
                                       93

-------
                                  REFERENCES


 1.  D. W. Van Krevelen, P. J. Hoftijzerv and F. J. Huntjens, Recueil
     Des Travaux Chimiques Des Pays-Bas, 68, 191-216 (1949).

 2.  M. R. Beychok, -Aqueous Wastes from Petroleum and Petrochemical
     Plants. John Wiley & Sons, New York, N.Y., (1967).

 3.  T. J. Edwards, J. Newman, and J. M. Prausnitz, AIChE Journal. 2]_, 248
     (1975).

 4.  R. L. Kent and B. Eisenberg, Hydrocarbon Processing. Feb., 87-90
     (1976).

 5.  D. L. Cardon and R. M. Wilson, "Ammonia-Carbon Dioxide-Hydrogen Sulfide-
     Water Vapor-Liquid Study", Final Project Report in progress for the API.

 6.  D. C. Bomberger and J. H. Smith, Report on "Evaluation of Ammonia
     'Fixation1 Components in Actual Refinery Sour Waters", Stanford Re-
     search Institute, December 10, 1977.

 7.  H. S. Harned and S. R. Scholes, J. of Amer. Chem. Soc., 63, 1706
     (1941).                                                 ~~

 8.  Cuta and Stratfelda, Chem. Li sty, 48_, 1308 (1954).

 9.  B. N. Ryzhenko, Geokhemiya. 137-138 (1963).

10.  Handbook of Chemistry and Physics. 51st ed., The Chemical Rubber Co.,
     D-122.

11.  Handbook of Physical Constants, Revised ed., The Geological Society
     of America, Memoir 97, Sec. 18 (1966).

12.  A. L. Ellis and N. B. Milestone, Geochim. Cosmochim. Acta, 31, 615
     (1967).

13.  Handbook of Chemistry and Physics, 51st ed., The Chemical Rubber Co.,
     D-143-D-144 (1970-71).

14.  American Petroleum Institute, Publication No. 946, "Sour Water Strip-
     ping Project Committee on Refinery Environmental Control, American
     Petroleum Institute", June 1975.

15.  E. H.  M.  Badger and L. Silver, J_. Soc. Chem. Jjid_., 5_7_, 110-12  (1938).

                                      94

-------
16.  Breitenbach,  Bull.  Univ.  Wis.  Eng. Exp. Sta. Ser. 68 (as given in
     Perry's Chemical  Engineers'  Handbook. Fouth ed., 1963).
17.  I. L. Clifford  and  E.  Hunter,  J_. Phys. Chem. , 37,101 (1933).
18.  I. 6. C. Dryden,  J_.  Soc.  Chen. Ind. , 66_, 59 (1947).
19.  S. Ikenko,  Kogyci  Kagaku Zasshi ,  64,  627 (1961).
20-  Lunge's Handbook  of Chemistry, Eighth ed.,  1952  (original  source  not
     given).
21.  D. H. Miles and G.  M.  Wilson,  "Vapor-Liquid Equilibrium Data  for  De-
     sign of Sour  Water  Strippers", Annual Report to  API  for 1974,  October
     1975.
22.  E. Otsaka,  S. Yoshimura,  M.  Yokabe,  S. Inque,  Kogyo  Kagaku  Zasshi.
     63_, 1214-1218 (1960).                             -- " --
23.  E. P. Perman, J_.  Chem.  Soc.  London,  83, 1168 (1903).
24.  Sherwood,  ing.  Eng.  Chem..  1_7, 745 (1925).  (as given  in  Perry's Chemi-
     cal Engineers'  Handbook,  Fouth ed.,  1963).
25.  S. Pexton  and E.  H.  M.  Badger, J_.  Soc. Chem.  Ind. . 57^,  106  (1938).
                                                         i
26.  T. T. C. Shih,  B. F.  Hrutfiord,  K. V. Sarkanen,  and  L.  N. Johansen,
     TAPPI, Technical  Assoc. of_ the Pulp  and Paper  Industry,  50, (No. 12)
27.  T. Takahashi,  Kogyo Kagaku Zasshi.  65_,  837-843  (1962).
28.  T. Takahashi,  S.  Yoshimura,  K.  Fukii, and  E.  Otsaka,  Kogyo Kagaku
     Zasshi.  65_,  743-745 (1962).
29.  V. E. Terres,  W.  Attig,  and  F.  Tscherter,  Gas,  u.. Wasserfach.- 98,
     512-516  (1957).
30.  M. E. Jones, J_. Phys.  Chem.  67,  1113-1115  (1963).
31.  S. D. Malinin, Geochem. Int.. 11 ,  1060-75 (1975).
32.  T. N. Kozintseva.x Geochem.  Int..  No. 4,  750-756 (1964).
33.  V. I. Oratovskii,  A. M.  Gamolskii,  and  N.  N.  Klimenko, J_. Appl.Chem.
     USSR*. 37,, 2363-2367  (1964).
34.  H. Gamsjager and  P.  Schindler,  Helv. Chim. Acta.  52,  1395-1402  (1969).
35.  T. N. Kozintseva,  Geokhemiya.  121-134 (1965).
                                      95

-------
36.   J. L.  Vogel, M.S.  Thesis,  The University of Tulsa,  1971.

37.   R. J.  Frohlich, Ph.D.  Thesis, Polytechnic Institute of Brooklyn Uni
     versity, Microfilms No.  60-3497,  1957.

38.   M. R.  Beychok, Hydrocarbon Processing.  Sept.,  261-263 (1976).
                                     96

-------
                                  APPENDIX

           COMPUTER PROGRAM  FOR  CALCULATING  SOUR WATER EQUILIBRIA
                     BASED ON  THE VAN  KREVELEN  EQUATIONS


    Table 33 gives a listing of  the  computer program used for calculating
NH,, C02, HpS and.kLO partial  pressure data  for comparing the SWEQ model with
the Van Krevelen  " equations.   The  input and output of this program is very
similar to the SWEQ model.

     The main calculations are done  starting with the following statement.

                DO 2030  I =  1,100

This is the start of an  iteration loop which extends to statement 2030.  This
iteration loop calculates the  amount of C03= (BT) and H?NCOO" (EPS) in
solution for various assumed concentrations  of  HCO~~ (Ac).  The concentration
of HCO?~ is adjusted in  each iteration so that  the total of HCO-" + C02~ +
HpNCOO* concentrations add to  the COp  content of the mixture.  The folTowing
Fortran symbols are used for the chemical equilibrium constants.


               EK1     C02(g) +  NH3 +  H20 +  NH4+ + HC03"             (A-l)
               EK2     NH4+  + HC03- -»  H2NCOO~ + H20                  (A-2)

               EK3     NH3 + HC03- -> NH4+ +  C03=                     (A-3)

               EK4     H2S(g) +  NH3 "*  HS~ +  NH4+                     ^A"4^

Other symbols have the same  meaning  as symbols  in the SWEQ computer program.
After correct values of  HCOo", C0~~, and H?NCOQ- concentrations are found,
the program proceeds to  calculate NH-,  COp,  H-S and H^O partial pressures and
vapor concentrations.  The results are then  printed out.

     This Van Krevelen computer  program only computes vapor composition and
pressure from a specified liquid composition and temperature.  No other
options were programmed.  The  equations for  the chemical equilibrium constants
and ammonia Henry's constant were obtained by fitting tabular values given
by Van Krevelen.  The Henry's  constant of ammonia above 90 C is based on
Beychok's2) graphical extrapolation.   Because of the tabular and graphical
nature of Van Krevelen's correlation and the graphical nature of the Henry's
constant for ammonia given by  Beychok,  there is some arbitrariness in the
computer program because another person using different equations to fit  the
tabular data and graphs  would obtain slightly different results.


                                       97

-------
     BeycholoRQ longer recommends the ammonia Henry's constant published
in his book,  ' but comparison is made with the book because it represents
a basis for comparing any changes or variations.

     Further discussion of the Van Krevelen-correlation can be obtained by
referring to either Beychok ' or Van Krevelen» '
                                    98

-------
TABLE A-l.  COMPUTER PROGRAM USED FOR CALCULATING VAPOR-LIQUID
          EQUILIBRIUM DATA FROM THE VAN KREVELEN CORRELATION
  QPEN(UNIT«20,DEVICE>'DSK', ACCESS- 'SE9IN',FIUE«'SWSD')
  DATA WA, WC, WS,WW/17t03,44. 81,34.08,18. 02/
  RHQ • \
  REAO(20,1000)
250
1994



I99S

1996


2000

2010
2020

2030

2040
3000

3010

3020
3030
       TC*273,15
       TK*i,8
       XA+1E-19
       XCMC-19
       X5+1E-19
       XW+JEM9
      1000*RHO/(XA+XC+X3*XW)
       XA*F/WA
       XC'F/WC
       CC
TK
TR
XA
XC
XS
XW
f «
CA
CC
C5
AC
EK2
EK3
CCS
6A i
CSS
SO i
00 2030  X»l,100
CAS * CA*AL« 1.5
IF(CAS)  1994,1994,1996
PA « 0
PC • 0
PS « 0
60 TO 1501
FORHATC  M2S  AND C02 JN EXCESS')
60 TO 1501
EPS • E*2*CAS*AC

IF(Dfc) 2000,2000,2010
BT « 0
60 TO 2020
BT f EK3*CAS*AWDE
CCE * AU+EPS+BT
        0
       CS
        0
       0
                                     This is a  test to see if the ratio
                                     of NHj/acid gas is greater than 1.5
                                     If not so, then the computation is
                                     skipped.
                                i
                                r
Iteration  loop to calculate amounts
of HCO", COj,
and HgNCOtt"
  At « AL*(.5*,5*CC/'CCE)
                               3000,3000,2030
  FORHATC  ITERATION 010 NOT CONVERGE IN 100 CYCUES'J
  HW • EXP(14tfl66*6996,6XtTR.77t67))
  ZFCTC-90)  3020,3010,3010
  HAD • EXP(^3,17..022**C)
  60 TO 3030
  HAD « EXPC*17.03+4315/TK)
  HA * ,92*HAO*EXPCi0576*CAS)
  HA » I/(HAt51,7l}
  PA « S1,71*HA*CAS
  TNW • C1000*RHC
  TNM « CA+CC*BT*OE+6A+SO*TNW*CS
  PM « HW*TNW/TNM
                                                             (continued)
                             99

-------
                     TABLE A-1, (continued)
1000
1861
1010
1011
1020
1030

1040
1050
1060
1070
1089
        EM • -»EXP(»l,764*ffc*7*AUOG(TC))*.0e9*CS«U,929..539/TK}*CC
             5lf7i*DE*C3/(EK46CA8*51.7t)
             PA/P
             PC/P
             PS/P
             PW/P
        PC
        PS
        P
        YA
        YC
        YS
        YW
1501    XA
        XC
        XS
        xw
        YT
        YA
        YC
        YS
        YW
        HRITE(5,1030)
        WRITE(5,1080)
        gK « YA/XA
        WRITEC5,|040)  XA(YA,CK
        EK * YC/XC
        EK « YS/X3
        EK n YW/XW
                      XS|YS,EK
        GO TO Z
        fORMATtSl)
        FORMAT(•  PH DID NOT  CONVERGt  IN  100  CYCLES*)
        PCRMATC  TEMPERATURE DID  NOT  CONVERSE IW 100  CYCLES')
        FORMAT(»  COMPONENT          LIQUID      VAPOR    K-VALUE*)
        FORMAT(//'  TEMPERATURE, C»,F6f2,/« PRESSURE,  PSIA«,F8,2,/
        FORHATC*  AMMONIA          *,3F10t5)
        FORMAT(»  CARBON DIOXIDE   *,3Fl0,5)
        FORMATC  HYDROGEN  suLFiDE»,3Fi0,5)
        FORMAT(»  WATER           »,3F10,5)
        ffND
                                100

-------
                                    TECHNICAL REPORT DATA
                            (Please read Instructions on the reverse before completing}
 REPORT NO

 EPA-600/2-80-067
  riTLE AND SUBTITLE
     2.
 A New Correlation of NH3,  C02, and H2S Volatility Data
 from Aqueous Sour Water  Systems
                                   3. RECIPIENT'S ACCESSION-NO.
                                   5. REPORT DATE
                                    April  1980 issuing date
                                   6. PERFORMING ORGANIZATION CODE
 . AUTHOH(S)                                	

 Grant M.  Wilson:  Brigham Young University
                                   a. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
 American  Petroleum  Institute
 2101 L  Street Northwest
 Washington, DC  20037
                                   10. PROGRAM ELEMENT NO.

                                    C33B1B
                                   11. CONTRACT/GRANT NO.

                                     R804364010
12. SPONSORING AGENCY NAME AND ADDRESS
 Robert  S.  Kerr Environmental Research Laboratory
 Office of Research and Development
 U.S.  Environmental Protection Agency
 Ada.  Oklahoma  74ft?ft      	;	
                                                             13. TYPE OF REPORT AND PERIOD COVERED
                                                               Final 1976-1977
                                   14. SPONSORING AGENCY CODE
                                      EPA/600/15
15. SUPPLEMENTARY NOTES

 American Petroleum  Institute project officer:   Ron Gantz
16. ABSTRACT
         A new correlation model has been developed for calculating sour water
         equilibrium  data at temperatures from 20°C to 140°C.  The correlating
         equations  in this new sour water equilibrium model have been used to
         obtain a computer program capable of  handling various chemical and
         physical equilibria of NHs, CO?, and  H2S in sour water systems in-
         cluding the  effects of carboxylic acids  on ammonia fixation and re-
         lease by caustic addition.  A bibliography of related literature data
         is included  in  the report.
17.
                                 KEY WORDS AND DOCUMENT ANALYSIS
a.
                  DESCRIPTORS
                                               b.lDENTIFIERS/OPEN ENDED TERMS
                                                c.  COSATI Field/Group
  Ammonia
  Hydrogen sulfide
  Carbon dioxide
  Volatility
  Aqueous systems
  Carboxylic acids
  Caustic
Henry's Constants
Chemical equilibrium
   constants
Sour water equilibria
Sour water stripping
Correlation of volatility
   data
Computer program
Ammonia fixation
 18. DISTRIBUTION STATEMENT

  RELEASE TO PUBLIC
                     19. SECURITY CLASS (This Report)

                        Unclassified	
                                                                                    ;ES
                                109
                                               20. SECURITY CLASS (This page)
                                                  Unclassified
                                                                           22. PRICE
                                           101
                                                                     4 U.S. GOVERNMENT PRINTING OFFICE. 1980-657 -146/5655

-------