r

Contributions in Petroleum Geology & Engineering ci) 5

(
( Contributions in Petroleum Geology and Engineering

( Series Editor: George V. Chilingar, University of Southern California

(
(
Volume 1: Geologic Analysis of Naturally Fractured Reservoirs Oilsand
Volume 2: Applied Open-Hole Log Analysis
Volume 3: Underground Storage of Natural Gas
( Volume 4: Gas Production Engineering
( Volume 5: Properties of Oils and Natural Gases
( Volume 6: Introduction to Petroleum Reservoir Analysis
r -- -----"'-""""-
' ,,'_. m~".~_
( Volume 7: Hydrocarbon Phase Behavior l I\.t;,:':nrm AiiEHHl'W !JEt p[;--' -a.-.
Volume 8: Gas Reservoir Engineering
( 81bUOTEGtl
(
( L~,N"':2:>",,=~JCt...~_.. . .J

'1

r
( r !APG-Bulgheroni
(
; l i:lf!ill/!IIIIIII 11/11111/
I/!
( 622338 C6275 0009329
!:)'-'-~fi,:c. -", . 1: (""n . -I ~~t('-~JI g3~-?S
(
e
(
(
(
(
(
(
(
(
~g~Golf Publshng
e Book Division
Company
~1'1", <:
p-tF'<~),\
(
Houston, London, Paris, Tokyo, Zurich ,,"r'JI -- \.~..
~o
\1 S' E~1I0iE{;.,. ~'i.
'"\--' i --1)
l ',\)\';::. ---- 1f
'rS, .,;,:..
( . A..
, ,("{iI./;7 ~

<. 1: //
( '."'~'"

ii
(
\. ,"
 \

l' I r-
. Iv~ -,"",- (
ti (
,1 ;:,
r r,\
Contents \i\\'\' i:.
~~:~ (

\ "
(

Nomenclature ............................................ ""~ir'~".-;./ (

(
Introduction 1
(

1. Petroleum Reservoir Fluids 4 (

Contributions m Petroleum Geology and Engineering Petroleum Reservoir Fluid Constituents, 4; Classification of Petroleum
(
Reservoir Fluids, 4; Hydrocarbon Fractions, 8; Residue, 9.

2. Compositional Determinations 10
Volume 5 Gas Chromatography and TBP Distillation, 10; Properties of Hydro- (
carbon Fractons, 22; Internal Consistency of Analytical Data, 30; Ex-
(
perimental Extension beyond C20+, 35.
Properties of Oils and Natural Gases (
3. Oil and Gas Property Measurements 39 (
Sampling, 39; Flash Separation and Compositional Analysis, 39; PVT
Measurements, 40; Black Oil Equipment, 40; Gas Condensate Equip- (
ment, 54; Volatile Fluid Equipment, 60; Measurement of Viscosity, 63;
Copyright 1989 by Gulf Publishing Company, Houston, (
Surface Tension, 66.
Texas. Al! rights reserved. Printed in the United States of
(
... Ainerica. This book, or parts thereof, may not be 4. Composition and Property Data 69
reproduced in any form without permission of the
(,
publisher.
5. Equations of State 79 (
Types of Equations of State, 79; Cubic Equations of State, 81; The
Library of Congress Cataloging-in-Publication Data (
SRK-EquatioiJ of State, 82; Phase Densities, 89.
Pedersen, K. S. (Karen Schou)
Properties of oils and natural gases/K. S. Pedersen, A. (
Fredenslund & P. Thomassen. 6. Flash Calculations 99
Two-Phase (P,T)-Flash, 100; Multi-Phase (P,T)-Flash, 104; (P,H) or (
p. cm.-(Contributions in petroleum geology &
engineering; v. 5) (P,S)-Flash Calculations, 105; Three-Phase (P,T)-Flash with a Liquid (,
Ineludes index. Water Phase, 108; Simplified Two-Phase (P,T)-Flash, 109; Phase Enve-
lope Calculations, 112. (
ISBN 0-87201-588-2
(
1. Petroleum-Analysis. 2. Gas, Natural-Analysis. 7. Characterization Procedures 114
1. Fredenslund, Aage, 1941- . n. Thomassen, P. (Per) , Classification of the Constituents of Oil and Gas Mxtures, 114; TBP G
1950- III. Title. IV. Series. Fractions, 114; The TBP Residue, 116; Characterization Procedures,
C.
TP691.P42 1989 120; Grouping of Pseudocomponents, 124; Calculation of the Ideal Gas
665.5-dcl9 88-24659 Heat Capacity, 126.
CIP
8. Simulation of PVT-Experiments 130
ISBN 0-87201-066-X (series) Constant Mass Expansion, 130; Differential Liberation, 131; Constant \,
Volume Depletion, 135; Separator Test, 140; Summary of Results of the
PVT-Simulations, 140.
 9. Comparison Between Experimental and Predicted
Thermodynamic Properties 141
Nomenclature
( Dew and Bubble Points (Phase Envelopes), 141; Flash Results, 144;
( Densities, 144; Enthalpy, 151; Summary of the Results of Thermody-
namic Properties, 154.
(
a: Equation of state parameter
( 10. Tuning of EOS-Parameters 155 a: Constant defined in Equation 15;18
The Needs for Tuning, 155; Parameters Available for Tuning, 155; Ex- a-a2: Constants defined in Equations 5-54 and 5-55
( amples on the Dangers of Tuning, 156; The Molecular Weight of the al-aS: Constants defined in Equation 11-6
Plus-Fraction as a Tuning Parameter, 166; Recommendations Regard-
(
ing Tuning, 169.
A: Aromatics (Chapter 1)
A: Constant defined in Equation 5-23
(
A: Constant defined in Equation 7-13
1l. Viscosity 172
( Definition of Viscosity, 172; Kinetic Gas Theory, 172; ViscosityCorrela-
A: Surface area (Chapter 13)
tions, 173; The ViscosityCorrelation of Lohrenz et al., 173; Calculation A: Constant defined in Equation 13-15
(
of the Viscosity Using Corresponding States Theory, 175; Comparison A: Constant defined in Equation 15-5
( with Experimental Results, 181. A,,: Constant defined in Equation 13-3
Ao-A2: Parameters defined in Equations 5-51 through 5-53
( Ao-A3:
12. Thermal Conductivity 187 Coefficients defined in Equation 5-86
( Kinetic Gas Theory, 187; Calculation of the Thermal Conductivity Us- API: API gravity (141.5/SG-131.5)
ing Corresponding States Theory, 188; Comparison with Experimental b: Equation of state parameter
( Observations, 194. b: Constant defined in Equation 15-18
( b-b4: Parameters defined in Equations 5-59 through 5-62
13. Surface Tension 196 B: Constantdefined in Equation 7-13
(
Estimation of the Surface Tension Using the Macleod-Sugden Correla- B: Constant defined in Equation 5-24
( tion, 196; Estimation of the Surface Tension from the Viscosity, 199; B: Constant defined in Equation 13-15
Comparison with Experimental Observations, 202. B: Constant defined in Equation 15-5
(
Ba: Shrinkage factor (defined in Equation 3-2)
( 14. Wax Formation and Inhibition 208 Ba-B: Parameters defined in Equations 5-57 and 5-58
Wax Composition, 208; Computation of Wax Formation, 208; Wax In- Coefficients defined in Equation 5-87
( Ba-B4:
hibitors, 214; Comparisons Between Calculated and Experimental Re-
sults, 217. B': Constant defined in Equation 13-5
( c: Peneloux parameter defined in Equation 5-39
C-C6: Parameters defined in Equations 5-68 through 5-73
15. Gas Hydrates 220
Types of Hydrates, 220; Hydtate Computation, 223; Hydrate Inhibi- C-Cg: Constants defined in Equations 13-16 and 13-17
tors, 228; Comparison with Experimental Observations, 230. C: Critical point (Chapter 1)
C: Constant defined in Equation 5-46
16. Simulation of Miscible Gas Injection 234 C: Density correlation factor defined in Equation 5-86
Process Description, 234; Calculation of the Minimum Miscibility Pres- C: Adsorption coefficient (Chapter 15)
sure (MMP), 238; Cell-to-Cell Simulation, 241; Comparison of Calcu- CF: Function defined in Equation 7-47
lated and Experimentally Determined MMP's, 246. CN: Carbon number
\.
e; Molar heat capacity at constant pressure
Index 249 e, Molar heat capacity at constant volume
C: Density correlation factor at standard conditions (Chapter 5)
C2: Density correlation factor at actual conditions (Chapter 5)

vi vii
 CO-C3: . Parameters defined in Equations 5-64 through 5-67 K: Watson Characterization factor (Chapter 7)
C-C4: Coefficients of ideal gas heat capacity polynomial Equation K: Constant defined in Equation 15-21
5-36 and Equations 7-43 through 7-46 1-17: Constants defined in Equation 12-18
C7+: Hydrocarbons with 7 and more carbon atoms L: Length of capillary tube Equation 3-10 (
D: Diameter (Chapter 3) L: Mean free pass between two molecules (
DL: Differential liberation m: Function of the acentric factor defined in Equation 5-16
e-e3: Functions defined in Equations 6-54 and 6-56 m: Constant defined in Equation 5-46
e-e6: Parameters defined in Equations 5-79 through 5-84 m: Number of pseudocomponents
EOS: Equation of State M: Molecular weight of inhibitor divided by molecular
Eo-E3: Parameters defined in Equations 5-75 through 5-78 weight of water
E: Fugacity MMP: Minimum miscibility pressure (
E: Function defined in Equation 12-12 Mr: r'th statistical moment defined in Equation 7-40
E: Function defined in Equation 12-36 MW: Molecular weight
Ej: Function defined in Equation 12-24 n: Refractive index (Chapter 7) (
fx,o: Function defined in Equation 12-22 n: Number of moles Equation 3-5 and Chapter 6
FA: Function defined in Equation 12-21 ( n: Constant defined in Equation 5-45 (
,
F: Distribution function (Chapter 7) n: Number of discrete components (Chapter 7) (
FCMP: First contact minimum miscibility pressure N: Naphthenes (Chapter 1)
F: Function defined in Equation 11-41 N: Stage number (Chapter 3) (
F2: Function defined in Equation 11-41 N: Number of components (Chapter 6)
g: G ravitational acceleration NCOMP: Number of components
g: Function defined in Equation 6-26 NSTEP: Total number of flash stages
g2: Function defined in Equation 6-28 ' P: Paraffins (Chapter 2)
G: Gibbs energy P: Probability function defined in Equation 7-17
GOR: Gas/oil ratio [PJ: Parachor (Chapter 13)
h.: Function defined in Equation 12-27 PVT: Pressure, volume, temperature
hj: Functon defined in Equation 12-25 q,,: Function defined in Equation 6-46
hx.o: Function defined in Equation 12-23 q,s: Function defined in Equation 6-47
H: Function defined in Equation 3-11 qo: Functian defined in Equation 6-45
H: Molar enthalpy q: Heat Ilow per unit area (Chapter 12)
H: Function defined in Equation 6-23 r: Crystal radius (Chapter 14)
HTAN: Function defined in Equation 11-41 r: Distance from lattice wall (Chapter 15)
1: Characterization Factor defined in Equations 7-21 and 7-22 r*: Critical radius (Chapter 14)
1: Variable used to define continuous distribution (Chapter 7) R: Gas constant
Ip: Quadrature point (Chapter 7) R: Parameter defined in Equation 7-27
i-h: Coefficients defined in Equation 11-38 Rs: Solution gas/oil ratio (defined in Equation 3-3)
J: Phase number (Chapter 6) s: Number of quadrature points (Chapter 7)
k: Constant defined in Equation 6-9 s: Mole fraction in salid phase
k: Boltzmann's constant (Chapter 14) Inhibitor concentration (Chapter 15)
. s:
kj: Binary interaction coefficient S: Standard Conditions (normally 15C and 1 atm)
k-k7: Canstants defined in Equation 11-40 S: Entropy
K: Equilibrium ratio (vapor mole fraction/liquid mole fraction) SG: Specific Gravity
K: Constant defined in Equation 5-45 SRK: Soave- Redlich -Kwong

viii ix
 'r
,/';.-.--.
.... ,
:1/

I
I l .

Viscosity ~
T:
T:
Temperature
Average temperature defined in Equation 15-8
17:
17: Parameter entering into Equation 7-17
Mole fraction of "continuous components" (Chapter 7)
\
;~'
.\;:;- ='
'.<'~ I
r ,

T*: Transition temperature defined in Chapter 6 17: . 15 " .

TBP: True boiling point 170: Function defined in Equation 11-37 "~. ,//
171: Function defined in Equation 11-37 ""~
v: Velocity
v: Parameter defined in Equation 7-23 1117': Function defined in Equation 11-37
v: Volume fraction 8: Shape factor (Chapter 11)
V: SRK volume defined in Equation 5-41 8: Parameter defined in Equation 11-39
w: Parameter defined in Equation 7-24
,,: Thermal conductivity
w: Weight fraction (or weight percent) "o: Function defined in Equation 12-16
w: Weighting function (Chapter 7) "1:
,,/:
Function defined in Equation 12-16
Function defined in Equation 12-20
w: Cell potential (Chapter 15)
,,1/ : Function defined in Equation 12-29
x: Liquid mole fraction
y: Vapor mole fraction p.: Chemical potential
( Number of cavities of type i (Chapter 15)
y: Variable defined in Equation 6-10 Vi:

( YKi: Probability that a cavity of type i is occupied by gas ~: Molar vapor fraction (Chapter 7)
molecules of type K ~: Viscosity reducing parameter (Chapter 11)
( z: Mole fraction p: Density
( Z: Compressibility factor p: Molar density
a: Surface tension
a: Molecular hard sphere diameter (Chapter 11)
( CREEK LETTERS T*: Parameter defined in Equation 6-32
T: Shear stress
( <1>: Fugacity coefficient
o': Parameter defined in Equation 6-43
Parameter entering into Equation 7-17 <1>: Volume fraction (Chapter 14)
\ a:
a: Parameter defined in Equation 11-34
,1
X: Binary interaction coefficient (Chapter 14)
( Acentric factor
a: Parameter defined in Equation 16-1 w:

l ac: Riedel parameter defined in Equation 13-4

( {3: Vapor mole fraction
,1 SUB- AND SUPERSCRIPTS
{3: Parameter entering into Equation 7-17
{3: Constant defined in Equation 13-5
')': Activity coefficient (Chapters 14 and 15) attr: Attractive
r: Camma function (Chapter 7) a: Pure water state (Chapter 15)
\
o: Solubility parameter (Chapter 14) B: Boiling point
~: Step size (Chapter 6) Br: Reduced boiling point
Function defined in Equation 12-16 {3: Empty hydrate lattice (Chapter 15)
11"c:
11" ': Function defined in Equation 12-16 c: Critical state
11" 1/: Function defined in Equation 12-17 e: Equatorial
1117: Function defined in Equation 13-18 F: Feed (Chapter 5)
1117': Function defined in Equation 11-38 F: Freezing point (Chapter 11)
1117 1/: Function defined in Equation 11-40 C: Caseous phase
1:: Number of moles in the second phase (Chapter 6) H: Hydrate phase (Chapter 15)
j: Parameter defined in Equation 6-27 i: Component index

x
xi
1: Phase number
11: Phase number
id: Ideal
int: Internal
j: Component index Introduction
k: Component index
L: Liquid
mix: Mixture
M: Methane Traditionally, chemical and petroleum engineers engaged in production
N: Component identification number (Chapter 16) of oil and natural gas from reservoirs have not devoted much interest to de-
o: Reference state (Chapter 6) tails regarding the properties and compositions of the hydrocarbon fluids
o: Reference component (Chapters 11 and 12) produced. Design of production schemes and equipment was mainly based
p: Separator conditions (Chapter 2) , on experience. This was, however, insufficient considering the enormous in-
pure: Pure component vestments required for oil and gas exploration and production. Today pro-
r: Reduced state (property at actual conditions/property at critical duction systems and equipment undergo detailed engineering analyses and
conditions) careful design 1:0 optimize the amount of hydrocarbons produced and mini-
R: Reference conditions (normally 15C and 1 atm) mize the operating costs and investment.
RA: Racket (Equation 5-43) Analysis and design of oil and gas production facilities and schemes re-
ref: Reference state quire thorough knowledge of the thermodynamic and physical properties of
rep: Repulsive hydrocarbon fluids. To know how much gas and how much oi~arator
res: Reservoir conditions produces under given conditions, it is necessary to know the vapor-liquid
res: Residual (Chapter 5) equilibrium relationships and the_d~~sities of the mixture. Understanding
S: Solid phase (Chapter 14) . the flow processes in the reservoir requires knowledge of viscosity and SUT-
spec: Specified value face tension. Heat exchanger design requires knowledge of the thermal con-
a: Surface tension ductivity of the flowing fluids. Other examples abound.
tr: Translational It would, naturally, be best if knowledge of the above mentioned proper-
V: Vapor ti ties was available from experimental observations. Those are, however, im-

I
vap: Vapor I possible to measure for all hydrocarbon fluids in all relevant conditions. The
w: Water chief objective of this book is, therefore, to describe in detail how to use
wt%: Weight percent ,1 mathematical models to accurately predict transport properties (viscosity
x: X-direction \ and thermal conductivity), surface tensions, and thermodynamic properties
y: Y-direction 1 (density, enthalpy and phase equilibria) of naturally occurring oil and gas
.. Pseudo state (Chapter 5)
Dilute gas (Chapter 11)
mixtures. The book is intended for the design (chemical or petroleum) engi-
, neer in the oil and gas related industry.
The models described in this book are the results of a collaborative re-
, search and development project executed by three parties: Statoil a.s., the
Norwegian state oil company, Calsep A/S, a Danish engineering consulting
company, and the Department of Chemical Engineering at the Technical
, Universty of Denmark. The research project was carried out on three levels.
The basic level concerns the description of the oil or gas. Detailed compo-
sitional analyses of Statoil's North Sea oils were performed, including chro-
matography, TBP-distillation, and others. The analyses were checked for in-
ternal consistency by, for example, computing and measuring the average

xii 1
 2 Properties of Oils and Natural Gases Introduction 3

rnolecular weight of the residue. The end result of this basic level research
was a highly detailed and-cornpared with earlier work-rnuch irnproved
analytical description of the hydrocarbon rnixtures.
The second level of research concerns the developrnent of irnproved corre-
lations of Te, P and w (critical ternperature, pressure and acentric factor)
for the hydrocarbon fractions as a function of the rnolecular weight and
other properties. In addition, procedures for estirnating ("characterizing")
the cornposition of the C20+ residue were developed.
The third level concerns choice of equation of state, viscosity rnodel, etc.
by cornparing rneasured and predicted properties. This level of research in-
eludes further developrnent of the rnodels by, for exarnple, irnprovernent o
( the mixing rules. It also ineludes application of the rnodels to sirnulate rnac-
roscopic processes such as flash operations, wax forrnation, and reservoir oil!
gas miscibility.
There is a great de al of interplay between the three levels of research. The
availability of correct and detailed cornpositional analyses is a necessary
(
condition for developing meaningful and sound improvements in the corre-
lations of Te, Pe, and w in the equations of state. Sound models for Te, Pe,
and w are needed to furnish good predictions of densities and dew points
usng an equation of state. Improvements and adjustments continued on all
three levels throughout the research project.
Because the models presented in this book are based on the most detailed
and correct analytical information possible, the predictive capacity of the
models is very high. Although the models were developed almost exelusively
on the basis of data for North Sea reservoir fluids, they may also be applied
with confidence to petroleum reservoir fluids from other regions. It is also
important to realize that, although highly detailed analytical data were re-
tained with, for example, the Peng-Robinson equation of state. It should be
emphasized, however, that the characterization procedure of Pedersen et al.
described in Chapter 7 in detail, is tied to the chosen equation of state, al-
though the principIes may be applied with any equation of state. The flash
calculation procedure described in Chapter 6 can be used in conjunction
with any equation of state.
The third part of the book (Chapters 11-13) describes newly developed
correlations for predicting the transport properties (viscosity and thermal
conductivity) and surface tension of hydrocarbon mixtures. The transport
property models are based on the principIe of corresponding states, and may
be applied to both gaseous and liquid phases. One of the surface tension cor-
relations is elosely tied to the viscosity correlation.
The last part of the book (Chapters 14-16) describes some novel, impor-
tant applications of the models developed in this work: wax formation in
pipelines, gas hydrate formation, and studies of miscible gas injection pro-
cesses.
The intent has been as complete a presentation as possible. Thus, many
detailed experimental data are listed, and all details regarding the devel-
oped models are given so that they may be readily applied, based on the
material given in this book.
,j:
i:
I

quired to develop the models, it is notnecessary to have th same degree of 1

1
detail available in order to use the models. i
The book is divided into four parts. The first part (up to Chapter 4) intro-
duces and describes oils and gases. Different types of oils are described in
Chapter 1, and laboratory methods for compositional and property mea-
surements are given in Chapters 2 and 3. Chapter 4 lists compositional data
for several hydrocarbon mixtures, and extensive data ineluding densities,
dew points, viscosities, and others are given.
The second part (Chapters 5-10) describes the modeling of the thermody-
namic properties such as densities, dew points, and flash separations. From
the many different models available, the Soave-Redlch-Kwong equation of
state was chosen as the basic model for these properties. It is shown in Chap-
ters 8 and 9 that this model (when modified slightly using the volume trans-
lation concept) gives results that are in excellent agreement with experimen-
tal values. The choice of the Soave-Redlich-Kwong equation of state is,
however, somewhat arbitrary. Equally good results could have been ob-
 Petroleum Reservoir Fluids 5

Table 1-1
Physical Properties of Some Common Petroleum Reservoir Fluid
Constituents

1 Density (g/cm3)
Melting Point Normal Boiling at 1 atm and
Chapter 1 eomponent Formula (Oe) Point (Oe) 15e

Paraffins
Petroleum Reservoir Fluids Methane CH. - 184 - 161.5 -
(

PETROLEUM RESERVOIR FLUID CONSTITUENTS

l; Ethane

Propane

n-Butane
C2Ha

C3Ha

C.HlO
- 172

- 189.9

- 135
-

-
88.3

42.2

0.6
-
-
-

Petroleum reservoir fluids are multi-component mixtures, consisting pri-

f~ Iso-Butane

n-Pentane
C.HlO

CSH'2
- 145

- 131.5
- 10.2

36.2
-
0.626

marily of hydrocarbons belonging to one of the following three classes: .~ n-Hexane CaH'4 - 94.3 69.0 0.659

Iso-octane CaH,a - 107.4 99.3 0.692

Paraffns (or alkanes) consist of chains of hydrocarbon segments (e.g., -CH2
or -CH3). The molecules may be unbranched (normal paraffins), or n-Decane C,oH22 - 30 174.0 0.730
branched (iso-paraffins). The atoms are connected by single bonds. Meth- Naphthenes
ane (CH4) is the simplest paraffin and it is also, on molar basis, the most Cyclopentane - 93.3 49.5 0.745
CSHlO
common constituent of petroleum reservoir fluids.
Methyl cyclo-
Naphthenes (or cycloalkanes) are hydrocarbons similar to the paraffins, but - 142.4
pentane CaH'2 71.8 0.754
differing from these by containing one or more cyclic structures. The ele-
Cyclohexane CaH'2 6.5 81.4 0.779
ments of the cyclic structure (e.g., -CH2) are connected by single bonds,
,!,
and usually six carbon atoms are joined in a ring structure. Aromatics
Aromatcs are hydrocarbons containing one or more ring structures similar Benzene CaHa 5.51 80.1 0.885
to benzene (C6H6). The atoms are connected by aromatic double bonds.
Toiuene C7Ha - 95 110.6 0.867
Polycyclic aromatic compounds (or asphalthenes), like naphthalene
o-Xylene CaH'0 - 29 144.4 0.880
(CIOH8) and anthracene (CI4HIO), may also be presento
Naphthalene C,oHa 80.2 217.9 0.971
In addition to hydrocarbons, the non-hydrocarbons nitrogen (N2), carbon
dioxide (C02) , and hydrogen sulfide (H2S) are often found in petroleum
~.
mixtures. Finally, petroleum mixtures may contain helium, mercury, and
metal-organic compounds. These will not be covered in this work. Table 1-1 Table 1-2 shows typical molar compositions of mixtures belonging to each
shows some physical properties of common petroleum reservoir fluid con- t of the four classes. For each composition, the last hydrocarbon is marked
stituents. The molecular structures of some of these components are shown with a " +". It indicates that heavier components are contained in this frac-
in Figure 1-1. tion also. In addition to the composition, the state of a fluid is also deter-
mined by the temperature and pressure.
The differences in the molar compositions between each of the four
CLASSIFICATION OF PETROLEUM RESERVOIR FLUIDS classes of reservoir fluids are naturally reflected in differences in the fluid
.properties. Figure 1-2 shows schematically typical phase envelopes for each
~ class. Each phase envelope corresponds to a fixed overall composition. At
Petroleum reservoir fluids may be classified as gases, gas condensates, vol-
the temperatures and pressures inside the phase envelope, two phases (usu-
atile oils, or black oils.

"
 6 Properties 01 Oils and Natural Gases Petroleum Reservoir Fluids 7

PARAFFINS: H Table 1-2

Typical Molar Compositlons of Petroleum Reservolr Fluids
I
H-C-H
H
I
H H
, . I
I HH H H H H HH H H Component Gas
Gas
Condensate
Volatile
011
Black
Oil

H-C-H
I
H
H-C-C-C-H
III
H H H
"""""
H-C-C-C-C-C-C-C-C-C-C-H
I I I I I I I I I l'
H H H H H H H HH H
N2
e02
e, 90.0
0.3
1.1
0.71
8.65

70.86
1.67
2.18

60.51
0.67
2.11

34.93
Methane Iso-Butane n-Decane e2 4.9 8.53 7.52 7.00
e3 1.9 4.95 4.74 7.82

NAPHTHENES: e. (+n) 1.1 2.00 4.12 5.48

AROMATlCS: es (i+n) 0.4 0.81 2.97 3.80
H
H /H H H H C, (+n) 6+: 0.3 0.46 1.99 3.04
I 'c I e7 0.61 2.45 4.39
H-C-H H-./,~ I I
H"C C\j C ~C,_/C~ ea 0.71 2.41 4.71
H"t H--' I H H-C' ~C-H H-C '\: 'C-H e9 0.39 1.69 3.21
H, / \~ 11 I I 11 I elO 0.28 1.42 1.79
,C C'H H}' /(.H
C 'H H-C"'tC-H H-C~ /t ,t:C-H f~ ell
HJ-!-H
\ ./ \
H.......

I
~C
I
\:7'
I
e'2
c.,
0.20
0.15 12+:
1.02
5.31
1.72
1.74

H H H H H H ~ 0.11 1.74
c., 0.10 1.35
Methyl Cycl ohexane Benzene Naphthalene c., 0.07 1.34
Cyclopentane c., 0.05 1.06

Figure 1-1. Examples of hydrocarbon molecular structures. el? 17 +: 0.37 1.02

c., . 1.00
c., 0.90
e20 20+: 9.18
~

ally gas and liquid) are found. Outside the phase envelope only one phase
(usually gas or liquid) exists. Coming from the single phase region, the phase
envelope represents the conditions where the first trace of a second phase
may be observed. The phase envelope consists of a bubble-point branch and
a dew-point branch separated by the critical point, C. The bubble-pont

branch is the one at the lower temperature and higher pressure. At the bub-
ble-point branch the newly formed phase can be seen as a gas bubble. When
reaching the dew-point branch from the single-phase region the first sign of
't
a second phase will be the formation of a liquid drop. The bubble point
e branch lies between the two-phase region and the single-phase liquid re-
gion. Similarly, the dew-pont branch is bounded by the two-phase region
and the single-phase gas region. At the critical point, two identical phases
are in equilibrium. In the near critical two-phase region, the two phases in
equilibrium are almost identical. In the near critical single phase region it is
TEMPERATURE - hard to tell whether the phase is a gas or a liquid, and the expression su-
...
Figure 1-2. Phase envelopes for reservoir fluids. e is the critical point. percritical Huid is often used. The point in Figure 1-2 marked by Tres, Pros

~.
8 Properties of Oils and Natural Gases Petroleum Reservoir Fluids 9

represents typical reservoir conditions. At those conditions the gas is single

phase gas, and so is the gas condensate. The volatile oil and the black oil are
both single phase liquids. If at constant temperature the pressure is reduced
as indicated by a verticalline in Figure 1-2 the gas will remain single phase
gas. A liquid phase will at some stage precipitate from the gas phase, but the
liquid phase will disappear if the pressure is further reduced. Precipitation
of a liquid phase from a gaseous mixture as a resuIt of a reduced pressure is
called retrograde condensation. The pressure interval where two phases ex-
ist is called the retrograde region. When the pressure is reduced, a second
phase is also formed at some stage from a volatile oil, and from a black oil.
The second phase is a gas phase, and contains prmarily lighter components.
In a black oil some of the mixture components are so heavy that they will
remain in a liquid form even at very low pressures.
The highest temperature at the phase envelope of a given mixture is called
the cricondentherm, and the highest pressure is called the cricondenbar.
RESIDUE
The plus fractions of the petroleum mixtures of Table 1-2 consist of the
hydrocarbons with a normal boilng point above a certain temperature. The
usual technique for separating heavy hydrocarbon fractions is a TBP-distil-
lation as described in Chapter 2. The distillation process cannot always be
continued untillOO % of the petroleum mixture is distilled off. At some stage
in the process, the temperature may have risen to a leve! where a further
temperature rise will cause cracking of the remaining molecules. The tern-
perature where the distillation process is stopped depends on the distillation
pressure and the sample quantity. This is reflected in the examples on com-
positional data given in Table 1-2 where the last hydrocarbon number frac-
tion vares from C5 to C19 Components as heavy as Cso may be present in
the TBP-resid~e at a concentration that influences the overall behavior of
the reservoir fluido The average molecular weight and the density of the
TBP-residue may be measured, and this is important information used in the
characterization procedures described in Chapter 7.

HYDROCARBON FRACTIONS

While the lighter constituents of petroleum reservo ir fluids are well-de-

fined individual components, the heavier (C7 +) hydrocarbon fractions are
mixtures of different components. Each carbon number fraction represents
the hydrocarbon mixture having its normal boiling point within a given
temperature interval; the carbon number cIassification is just used for con-
venience. As is seen from Table 1-1, hydrocarbons of a different chemcal
structure but with the same number of carbon atoms may have ver y differ-
ent normal boiling points. A carbon number fraction in Table 1-1 may,
therefore, contain hydrocarbons with varying numbers of carbon atoms.
From molar compositions like those given in Table 1-2 it is not possible to tell
whether the components of a given C7+-fraction are mainly paraffinic (P), I
I
naphthenic (N), or aromatic (A) components. As described in Chapter 2, }
components < ClO may be quantitatively identified. A PNA-analysis may
give a rough idea of the dominant molecular structures within the heavier
hydrocarbon fractions. Aromatics have a lower molecular weight than the
paraffins and the naphthenes with the same number of carbon atoms. Paraf-
t'
fins and naphthenes on the other hand have a lower density than the aro-
i
matics with the same number of carbon atoms. Measurements of the aver-
age molecular weight and the density of each hydrocarbon fraction are,
therefore, of great significance in the characterization procedure (see Chap-
ter 7), which is performed in order to simulate the propertes of a given pe-
troleum reservoir fluido

1
 Compositional Determinations 11

Gas Phase Analysis

A typical natural gas chromatogram is shown in Figure 2-1, obtained us-

ing a Hewlett Packard 5880 gas chromatograph (Figure 2-2).
Chapter 2 6

Compositional Determinations 5 7 8

11
For oil and gas mixtures, the phase behavior and physical properties such
12
as densities, viscosities, and enthalpies are uniquely determined by the state 16
18
of the system, i.e., the temperature, pressure, and the composition. In sirnu-
lating the phase behavior and physical properties of complex hydrocarbon
mixtures accurately, it is necessary to have detai!ed and accurate cornposi- !I
o '1 22
tional information for each mixture. u
t--
o
;:
1 17 19 23

-, I
"WI.. ",LL~_
-; 1
i: 14 ~33 38
~ ~ l' I
"U
32 \

GAS CHROMATOGRAPHY AND TBP DISTILLATION o

~ ~ ~N MN

Traditionally, compositional data in the oil industry have only been re- ..J Z
" t--
Z t--
:t
U
t--
Z
t--
t--
Z
t--
t--
ported to C7 +, the compositional information being mainly based on low ""
;
V>
~ " "
o
temperature fractionate distillation data. This level is inadequate for accu-
rate modeling of the phase equilibrium and physical properties of the hydro-
U
t--
A 88 e Figure 2-1. Typical chromatogram
and column configuration. Peak
carbon mixtures.
In recent years, new methods have been developed for experimentally de-
A --1 PORAPAK R HMOLSIEVE13X~ numbers correspond to calibration
numbers given in Table 2-1. The an-
termining the composition of hydrocarbon mixtures. These methods yield a alytical conditions for the organic
B
far more accurate and detailed description of the hydrocarbon systems and and inorganic analyses are given in
are described in this chapter. Tables 2-2 and 2-3, respectively.
e -j SPLlT 1: 100 HWCOT CP SIL 51--8 From Osjord and Malthe-S0rens-
sen, 1983, reprinted from the Jour-
nal of Chromatography.
Gas Chromatography

The chromatographic separation of oxygen, nitrogen, carbon dioxide,

The appearance of capillary columns has enabled separation and quanti-
fication of many more individual components than prevously possible.
The compositional description of the reservoir fluid is done by analyzing
separately samples of the gas and liquid phases, which when recombined in
the correct gas to liquid ratio will yield the reservoir fluid composition. The
compositional description of the gas phase is carried out in one step: N2,
and approximately 60 hydrocarbons from C to CIO+ was achieved by using
a combination of packed and capillary columns. The sample was injected
va two time-programmed loops into packed and capillary columns with
thermal conductivity and flame-ionization detection. The packed columns
I were molecular sieve 13X, Porapak R, and the capillary columns were
I

CO2, CI-CIO by capillary column chromatography. The liquid phase com-
positional description is carried out in two steps:
1. To CIO+ using capillary columnchromatography.
2. From ClO to C20+ using a mini distillation apparatus.
I Chrompack Sil 5 (fused silica). The hydrocarbons were analyzed by split in-
jection using the capillary column, and temperature programming starting
at 30C, and the permanent gases were analyzed isothermally at 50C us-
t ing the packed column in a special compartment outside the main oyen.

10
12 Propertes 01 Ols and Natural Gases Compostonal Determnations 13

r:i~ CD<Or-LOtOOCO VCOC\JCOCDCOO(()VOf"-..C\JO'l

",,,,triC\itriu:icxi""':""':6tri,,,C\i
oi
v
cef- ll'iuiLriLLtricrj
CX:>CO "",,,,,,,"""'T""" N
ci- ~N

~
s:
CI
10101000VO
vNN<O<OOlOl
OQT"""QOOC\J
I,{)lOVNIDO>LOVC\JO'>O>O>I'-
<OC')OOC')C')NvNOlNNC')
10000vOOooooo~
coO
0000000 0000000000000 Cl
~ 6666666 6666666666666 O

Q) Q)
e e

I
<Il <Il
cc
Q) Q)
Q) Q) 0.0. Q)
e e
O o e
ltl >-
UI
CII
UI 'E
GI
O>
e
e--
Q) <Il <Il

!9~a5
Ca.a.
00
~~
>.>.
Q)
cCO
co~
~-
iii e <Il Q) O o ...c:..c ..c:-
e
..,;: o ma5- 0.--
.Q~g. Q5Q) .Q~
Q.

E
e a. u u u
~
E E
al 'C "C
u u
~ >.
'e"
Q)

CID ~.Q' ~~~

o
~g.Q)~Q5Q5 ctsc f7r;= <00)--5 <Il
U
O
'O .Q.~~-EEE
X~V('f) cC~Q) Q)
-O
CII >. >. x Q) .- .- ~ ~ .~ .~ ~ ~ Q5 .~ ~ ~

..
-O
el -S -5 Q) .S 9 9 el> .2o.U() Q)Q)EO<llQ)
------- -- ~-~~_'1'::,,;J:iit'*
ni Q)Q)J::OC')Ne
EE>-' ,f/.cu
g.-ciC}l~ ==o+g>.Q)~ Q)
;;::

CII oo-E~c:c:i5.
u >.(1) CI) Q) CI)-X
>'oc::c:cf;:-S,c:coQQlc
>.>."" eco ~Q)
Figure 2-2. Hewlett Packard 5880A chromatograph. M.ulti-column rigged for one- ...c:>.~ ~ Q) ID Q).~ E +>.0
step gas analysis. ~ I 1
(t)~:2:0_"""J:
Q)._~~Q)
Q.r:~-.:!:2:2()-oxz
~ -O
'c
a ~~I~~~I
C\I~(",)~~~C:: :2ii,.:,.:2~dJ":~-z E:6 C: ::J
Q.. s;
";" E
C\I ni
CIIcn
:a
e
:;o..o. Wf"-..oocnOT"""(\J C"?LOCO,....COO>O.,-NI.OI'-COO)
1
2
Through an elaborate calibration program, most of the individual com- !!!~
UI ,ez NNNNC')C')C') C')C')C')C')C')C')vvvvvvv

i ~
ponents have been identified. Table 2-1 shows a resulting list of components iii.. O B
and the quantity of each component (mole fraction) present in the sample
corresponding to Figure 2-1. Details about the analyses are given in Tables -
Z
:l
ni

- I~ ~ NNNOlVONOl
6000~NMM
vOOOC')lO COCOT"""lOLONl.O
triu:iu:iu:iu:iu:iu:iu:icxiu:icriu:itri~
a
(ij

- ci .!:l
00
C')
2-2 and 2-3. ni
ui ~ Ql

o ci -5
This information is usually too detailed for practical purposes, so the
compositional information is regrouped. In the present sample, the re-
t::
o -s. COWO><OOlVf'..T""" OllONvC')N VNT"""'COlO-c:tLO
,gE
grouped gas phase analysis can be either that shown in Table 2-4 or Table 2- io ~CI I VV
N O
100
T"""(\JT"""T"""(\J""''''''''OLOO>r-OOO
COC")CO(OT"""OW(O OC')Nv<OC')
OOC')OvONV
tDOOtDO<Dr-
0000000
U
Ql

a.
<OC')
~N
101OC')ONOI'-<O
6 triC\iC\i",666
OOOT"""O"-
666666 6666666 e
...
5. Both examples of regrouping will yield satisfactory descriptions of the gas ~ <O~~
phase composition. E ~
o M
o Q) Q) Q) Q) cc
al
e Q) Q) Q) e e e
<Il
a. e e c:: ..................
co ctI ctI cu -
'E o ~
::J
$Q)w
:::J C e eQ)a.c.
fi5 fi5 fi5 Q) ~
Liquid Phase Analysis GI
e
Q)
-o Ci :9w:92l$ a. Ccn

o 'x >. >-C:>-cc 0.>.>. >.Q)~ c:

Q.
o J:: Q) -Ecu-Eww --5-5 -E@Cl> ~
E Q) Q5cQ) Q)C:Q)0.0.Q) >.Q)wwwx~ (j)
c::a Q) Q) e Q) E <Il e
The liquid hydrocarbon phase, i.e., the oil, is a much more complex mix- o
O
Q) e c::
~ccu:5
Q) c: cu .-e $
@ oQ)C ~. ~~ ~ ~.. e ~ ~ d>
OJO
e-e q.2qQQ~ E99~9-Q5 :
ro
ture than is the gas phase. Many more components appear in a typical gas .~ cu
Q5~g.-g~Ng.~ NUC'):2:2I Q5.q.NCC"?~

chromatogram of an oil as shown in Figure 2-3 (Osjord et al., 1985). Consid- zo :2:[jO:~ :N-~ C: N'0N'NMC: ::2:35C"i6~ ~
e

erable overlapping of the peaks at the higher carbon numbers occurs, which e '"
u
o
:;..
.
o
.
makes good quantitative determination of the heavercomponents impossi- O'>O~N(,,).q.lO
,ezo
I.[)COf'..cx>O')O~N C')vlO<OI'-OO
ble from the gas chromatogram alone.
Nv
~~~ ~~~~~~ ...-NNNNNN O
E
i o
O u:
 ".
Compositional Determinations 15
14 Properties of Oils and Natural Gases

Table 2-4
Q)
e Regrouped Composition 01 Separator Gas
.o Q)
~ E
Q)
VI
Ol Qi Molecular
e Mole %
'c Q)
Component Weight % Weight
e Ol
Q)
.S e c: Nitrogen 0.779 28.01 0.659
~ .S:
.
E"'"
Q)-ijl ~
o VI_'
E'
e ~:::::::: E
I/I E
Carbn dioxide 10.499 44.01 5.652
O Cii
:E cS:"E ~:ti ~~ .2 'S; E E el) o 16.04 68.795
"t:J
e
oE
-.;;:::::ou
ro ro
>1ija..
LO
,.- ID ~
e
ULOo
:JNLO
~:o
roroLO
Methane
Ethane
46.577
16.326 30.07 12.863
8~EC'loC._ .. ():; 0'0 - ->-~
7.938
'c o 1: g 8 E ~ O :; ( O ~~~g.E~ Propane 14.775 44.10
,Qe: Uj ~
~Icu~c..woT"""Q)E
"~ ~ ~ ~ ~ CJ ....
(ij~~N-O(j
oo .. $ .. ~
oor::l ;-Butane 2.295
4.810
58.12
58.12
0.936
1.961
n-Butane
ro"
_cnQ.CU\U
Ol~ S~ 0)_
:i c.E.
Q) E
,"-;::;:::~c.~~
;-Pentane 1.050 72.15 0.345
u.cu1'Q;(/)~c'@$o ~~o~~'iiiE
1.283 72.15 0.421
Qj~Q) Clo; 0.'.:':= s ts 1-~]!~CfJ$2
::1 a. ..
....
n-Pentane
g; ~ ~ E g; ~ '5. '5. ~ .!!!.
~~'8E~E~
Q.
e, Paraffins 0.728 86.18 0.200
I-LL~~I-CfJCfJCfJ~E
~~~~~J5~ e6 Naphthenes 0.065 70.14 0.022

Ul
'
>-
iii

c: !I~
el)
E
o

Ol
e
o.
Ul
'
>-
iii

c:
!I~
el)
E
ea
~ ~
o

Q)
Ol
.S:
o.
E
ro
CfJ
e, Total
e7 Paraffins
e7 Naphthenes
C, Aromatics
0.793
0.159
0.359
0.055
84.59
100.21
85.95
78.11
0.222
0.038
0.099
0.017
f $ E 88.59 0.153
(J ea Q) ro (J e, Total 0.573
': Il. O CfJ ': 0.007
111
111
L:
VI
Q)
ea Paraffins 0.034 114.2

~ ~ E ea Naphthenes 0.127 99.5 0.030

O s, o O . ea Aromatics 0.049 92.1 0.013
~
N
..O .f: "& C?-=
N'" s:
~I E
~ ea Total 0.210 99.7 0.050
.!!-; ' .E ~ Q).E VI o '-
C, Paraffins 0.009 127.7 0.002
.c c: E 2 :ti ~ ~ ~ eg Naphthenes
. 0.000

- s rs o Ul 0.001 125.0
!!! ,2 -D :: E !!! 15 1/1 o X E:
e 0.001
~~ al N e ::; e .S: O

~ C, Aromatics 0.006 106.2
'5
c: o~ c.E; -<:
o '5
~"e I E
E
o .>
Q)

.
c:..t::
.- C, Total 0.016 118.5 0.003

o
o ()~ ~-g ~ e
o
._,....
e LO
-
CX)

ri
.~
'; ~ '; E ~
'O
Decanes plus 0.012 140.0 0.002
_ .2 E :Q ~ o Qi E :g o o
O
LO
_~ .::s:. -;:- ~ ~ 5 100.000
L()
E 100.000
~ ~ <0_ ~~ u)~ EE ~ iii
,- e o =c.ro-ro N:l....., u
oo ~ ~ ~ :ri ~ .f5
>.
iii
o o
g
o Eo .2 i: g "! ~ J!
~
o ~ o
0 ..0
LOQ)LO o
ea -Q)
Q)o
_Ola>
Q)I~ Gas molecular weight = 23.69
c:
O o
0'-
I c:w"Oc;;-o
:J'O'j[L=
Q) O -
0_
.. o
Q.;()
E
o iii
c:
.. :J ..
~ (ij s [LE~~E~i
,....C")UN.,...U~
E
,g
Gas gravity = 0.818
.- ~ Q) o Q)
Qi. .. E,?~~Ui

Q)
I E - VI -= ->-
ro 'O ro
:JQ)u=_rorooEe
ea o ~'c S >. s: en LO .~
Q)Q)"O

-tj
'"
.~
> e >
(ijo(ij
E:5~E~~a;
::l O> fA :J Q) (J)'C
~c:
>(0
_.
~ m$: ~ Q.""i::"'C:C
cn .. !Q"OQ) <l> ..
c. .- () c:
~ Q) .- o~"5oai'5~
O...JoO...Joo
".
~
Q) ECfJLL
~ ~ (ti
.-
g. ~ =a.
'S 'E
(ti .S L.... >. cu .S! ro Q) (1) =
E g~ M
~~
M In this work the oil phase analysis is divided into two parts:
Ea:LLl- u...J...JELL ~ VI ex>
e al
E
:J 1. To CJO+ using chromatography
'O
g'" u '"
al
al 2. from CJO to C20+ using distillation
E
o,
ro
c:
~
~Q,

Q)
u
'O
Q)
.>c:
.,
e;
~
(/)

ec. .s
el) ro
E c.
Q,
.s
Cii
Q)
e
-; ea
Q)-
VI
e
-; Liquid Phase Analysis to CLO+
el) ~ ::;; ::;;
E ~ E 'O ~ ~u E 'O
ea ro :J e :J e
'O 'O
~
Il.
Q3
c. u
lO

'O e u
lO

'O A gas chromatogram of the oil is obtained by use of a chromatograph fit-

E (ij o
.;-
:J
(ij o
.;-
2 u ~ u
.~
ted with a 50-m Chrompack Sil 5 (fused silica) capillary column. The analy-
.~ O Q)
c. O
e
Q) (ij E E (ij E sis is temperature programmed from 10 to 300C. The sample volume in-
b e
-c e
LL ~
e
<{
o
u:
 16 Ptoperties 01 Oils and Natural Gases Compostonal Determnatons 17

jected is 0.1 mI with a split ratio of 1:100. The effluent detector is a flame Table 2-5
ionization type (FID). Further details of the analytical conditions are given Regrouped Composition of Separator Gas
as follows: (Corresponds to the Analysis in Table 2-4)

Component Mole %

Analytical Conditions for Liquid Phase GC Analysis Nitrogen 0.66

Carbon dioxide 5.65
Column: Type: fused silica capillary Methane 68.80
Liquid phase: Chrompack CP Sil 5 Ethane 12.86
Propane 7.94
Carrier gas: He, = 22 cm/s
;-Butane 0.94
Length: 50 m n-Butane 1.96
Inside diameter: 0.23 mm ;-Pentane 0.34
Film thickness: 0.4 ,tm n-Pentane 0,42
Hexanes 0.22
Detector: Type: flame ionization
Heptanes 0.15
Fuel gas: H2, 30 ml!min Octanes 0.05
Make up gas: N2, 30 ml!min Nonanes 0.00
Temperature: 350C Decanes plus 0.00
1,
,1
Sampling: Syringe: Hamilton 7001 N r: 100.00
Sample size: 0:1-0.5 ,tI Gas Molecular Weight 23.69

Injector: Type: split injector

Ratio: 1: 100 In addition to the weight fractions of the collective groups such as the
Liner: packed Jennings tu be :1 C/s, one may from such a detailed starting point calculate average rnolecu-
Temperature: 300C lar, weights, average density, critical properties, and PNA distributions for
1,

each of the groups up to C10+' See Tables 2-8 and 2-9.

Oyen:
. Total time: - 2h
Source: Osjord et al., (1985).
Type: temperature programmed i~t
Identification of the components in the range CIO to C20 on the basis of the
Initial value: + 10C, 2 min
chromatogram in Figure 2-3 is seen to be impractical. Hence the liquid
Ratel: 3/min -> ll5C
phase analysis is extended beyond CIO+ by a distillation technique.
Rate 2: lOo/min -> 300C
Final value: 300C, 60 min
Liquid Phase Analysis from CIO to C20+

t- A mini-dstillation apparatus is shown in Figure 2-4. It is an adaptation of

To properly quantify the individual components, an internal standard of
known quantity is added to the sample. In this way one may compensate for
column losses that are known to occur from carbon numbers of 15 and
higher. The component used as the internal standard is iso-octane, which is
normally not present in naturally occurring oils.
As for the gas sample analysis, a calibration program enables thorough
identification and quantification of the components present between C2 and
Cg. Table 2-6 shows a typical example of such an analysis. This detailed an
analysis up to C10+ is normally not needed in practice. Therefore, a regroup-
ing of components may be carried out as shown in Table 2-7.
the commercially available Fischer HMS 500 automated rnn-dstllaton
;1 system. The required sample volume is about 100 mI. The distillation proce-
dure used is a modification of the ASTM D-2892 procedure for true boiling
111<
point (TBP) dstllton. The purpose of the mentioned modifications is to
,
l' narrow the distillation cut temperature ranges so that they correspond to
those of Katz and Firoozabadi (1978).
The distillation procedure results inan analysis as shown in Table 2-10,
where the analysis up to CIO is carried out at atmospheric pressure. From CIO
to C20+, the distillation is carried out at a vacuum of 20 mm Hg to reduce
the temperature level and avoid thermal cracking of the sample.
18 Properties o] Oils and Natural Gases Compositional Determinations 19

Table 2-6
Individual Component Report
u
---1 (Peak Numbers Correspond with Chromatogram in Figure 2-3) .,,~, t 1
~ "':~\I~'
,
CD
N
Molec- Liquid
< ;:;

j
Q)
" tl /
:a Weight Mole Volume ular Density ~
~ Peak Component % % % Weight (g/cm3)
.,
:J
! Q)
e
1 C2 0.007 0.058 0.017 30.070 0.3580
Ci 2 C3 0.072 0.412 0.122 44.097 0.5076
'
., '"
Q.
E
E
~
Cl
E
CO
(J)
3
4
te,
o-e,
0.051
0.189
0.222
0.816
0.078
0.276
58.124
58.124
0.5633
0.5847
., e
>-
u
~ '0
Q.
~~ 5 2,2-0M-C. 0.000 0.000 0.000 72.151 0.5967
:
E'\ ~~
~ ---,=---.3
"O
Q)

~
6
7
-c,
n-Cs
0.188
0.285
0.653
0.991
0.257
0.386
72.151
72.151
0.6246
0.6309
~=~~
Q)
O Light end total 0.792 3.152 1.137 63.092 0.5964
oS 8 2,2-0M-C, 0.012 0.034 0.015 86.178 0.6539
.8 9 cv-c, 0.052 0.185 0.059 70.135 0.7502

_31 "O
e
o
o.
10
11
2,3-DM-C,
2-M-Cs
0.028
0.165
0.081
0.480
0.036
0.214
86.178
86.178
0.6662
0.6577
CIJ 12 3-M-Cs 0.102 0.298 0.131 86.178 0.6688
~ ~ 13 n-C6 0.341 0.993 0.440 86.178 0.6638
~ oo O Hexanes total 0.699 2.071 0.895 84.745 0.6687
~ CIJ
14 M-CyCs 0.231 0.689 0.262 84.162 0.7534
ID
.o 15 2,4-0M-Cs 0.015 0.038 0.019 100.205 0.6771
~ E 16 Benzene 0.355 1.140 0.343 78.114 0.8842
~
e 17 Cy-C6 0.483 1.440 0.528 84.162 0.7831
~ ~co
-
f~ Q)
n,
Q)
18
66
19
2-M-C6
1, 1-0M-Cy-Cs
3-M-C6
0.000
0.116
0.122
0.000
0.298
0.307
0.000
0.131
0.152
100.205
98.189
100.205
0.6829
0.7590
0.6915
e ;~~ Ci 20 1, cis-3-DM-Cy-Cs 0.000 0.000 0.000 98.189 0.7493
:: :;----." .. E
co 21 1, trans-3-DM-Cy-Cs 0.052 0.133 0.059 98.189 0.7532
CIJ
22 1, trans-2-DM-Cy-Cs 0.048 0.122 0.054 98.189 0.7559
:,.., Q)
::;~
ro 25 n-C7 0.405 1.014 0.504 100.205 0.6880
=.g CIJ
e O Unspecified C7 0.171 0.427 0.215 100.205 0.6800
e., - ='" Q)
O Heptanes total 1.997 5.609 2.267 89.426 0.7542
"O
- -- --- ~ ---=-= e
".-=-
El--=:::::; O
o
ro
26
27
28
M-Cy-C6
1, 1, 3-TM-Cy-Cs
E-Cy-Cs
0.918
0.027
0.000
2.348
0.061
0.000
1.016
0.031
0.000
98.189
112.216
98.189
0.7737
0.7526
0.7708
~---..' g; '> o 29 2,2,3-TM-Cy-Cs 0.042 0.093 0.050 114.232 0.7200
E 30 2,5-DM-C6 0.018 0.039 0.022 114.232 0.6977
~ 31 2,4-0M-C6 0.000
o _ 0.000 0.000 114.232 0.7045
-E
Ol
: -----.;; 32 3,3-0M-C6 0.026 0.057 0.031 114.232 0.7141
ro '"'"
~ 33 1, trans-2, cis-3-TM- 0.025 0.056 0.028 112.216 0.7579
O -
~ ~ .c i cvc,
O Q 34 Toluene 0.958 2.610 0.941 92.143 0.8714
"'O
M ~ 35 1, 1, 2-TM-Cy-Cs 0.000 0.000 0.000 112.216 0.7769
N .~ 36 2,3-DM-C6 0.033 0.073 0.040 114.232 0.7163
~ ... ()u.,
QJ

::l
37 2-M-C7 0.137 0.300 0.167 114.232 0.7019
Cl :; 38 3-M-C7 0.094 0.206 0.113 114.232 0.7099
.- o 0.425
esuodsai al~ L1. CIJ 39 1, cis-3-DM-Cy-C6 0.190 0.211 112.216 0.7701
40 1, trans-4-DM-Cy-C6 0.072 0.162 0.081 112.216 0.7668

(table continued on next page)

 l
11
20 Prop'erties of Oils and Natural Gases Compositional Determinations 21
1

1
11
Table 2-6 Table 2-7
Continued Gas Components and Boiling Point Groups up to C10+ from Capillary
::1

GC-Analysis .1-1
1I
Molec- Liquid
Weight Mole Volume ular Density Liquid
Peak Component % % % Weight (g/cm3) Molecular Density
Component Weight % Mole % Volume % Weight (g/cm3)
42 Unspecilied 0.028 0.062 0.031 112.216 0.7700
naphthene C, 0.000 0.000 0.000 16.0 0.260
42 Unspecilied 0.013 0.028 0.014 112.216 0.7700 C2 0.007 0.058 0.017 30.1 0.358
naphthene C3 0.072 0.412 0.122 44.1 0.508
42 Unspecilied 0.011 0.025 0.012 112.216 0.7700 ic, 0.051 0.222 0.078 58.1 0.563
naphthene n-C. 0.189 0.816 0.276 58.1 0.585
43 OM-Cy-C6 0.031 0.069 0.034 112.216 0.7700 2,2-0M-C3 0.000 0.000 0.000 72.2 0.597
44 1, trans-2-0M-Cy-C6 0.089 0.199 0.098 112.216 0.7799 te, 0.188 0.653 0.257 72.2 0.625
45 n-Ca 0.434 0.954 0.526 114.232 0.7065 e, 0.285 0.991 0.386 72.2 0.631
O Unspecilied C, 0.086 0.190 0.105 114.232 0.7000 Hexanes total 0.699 2.071 0.895 84.7 0.669
O Octanes total 3.231 7.957 3.551 101.978 0.7791 Heptanes total 1.j:l97 5.609 2.267 89.4 0.754
46 Unspecilied 0.047 0.094 0.051 126.243 0.7900 Octanes total 3.231 7.957 3.551 102.0 0.779
naphthene Nonanes total 2.427 5.241 2.598 116.3 0.799
47 2,2-0M-C7 0.009 0.018 0.011 128.259 0.7144 Oecanes & 90.853 75.971 89.553 300.3 0.868
48 2,4-0M-C7 0.017 0.033 0.020 128.259 0.7192 heavier
49 1, cis-2-0M-Cy-C6 0.024 0.054 0.026 112.216 0.8003 Source: Osjord el al. (1985). '11 I
50 E-Cy-C6 + 1, 1, 0.281 0.599 0.305 118.000 0.7900 11

3-TM-Cy-C6 Table 2-8 I

51 Unspecilied 0.047 0.093 0.051 126.243 0.7900 Distribution of Weight, Mole, and Volume % in a Crude Oil Sample,
naphthene
from Capillary GC-Analysis
52 3,5-0M-C7 0.017 0.034 0.020 128.259 0.7262
53 2,5-0M-C7 0.003 0.006 0.004 128.259 0.7208
54 Ethylbenzene 0.114 0.270 0.112 106.168 0.8714 Liquid
68 Unspecilied 0.027 0.054 0.029 126.243 0.7900 Molecular Density
naphthene Componen! Weight % Mole % Volume % Weight (g/cm3)
55 m- + p-xylene 0.697 1.649 0.687 106.168 0.86.83 Gas components 0.792 3.152 1.137 63.1 0.596
56 4-M-Ca 0.020 0.039 0.024 128.259 0.7242 Hexane group 0.699 2.071 0.895 84.7 0.669
57 2-M-Ca 0.054 0.106 0.064 128.259 0.7173 Heptane group 1.998 5.612 2.263 89.4 0.756
58 Unspecilied 0.009 0.018 0.010 126.243 0.7900 Octane group 3.231 7.957 3.551 102.0 0.779
naphthene Nonane group 2.427 5.241 2.598 116.3 0.799
58 Unspecilied 0.082 0.163 0.089 126.243 0.7900 Oecane group 2.483 4.666 2.654 133.6 0.801
naphthene Undecane group 2.239 3.799 2.388 148.0 0.803
58 Unspecilied 0.007 0.014 0.008 126.243 0.7900 Oodecane group 2.569 3.994 2.709 161.5 0.812
naphthene Tridecane group 3.127 4.479 3.238 175.3 0.827
59 Ortho-xylene 0.230 0.545 0.223 106.168 0.8844 Tetradecane group 3.153 4.172 3.214 189.8 0.840
60 3-M-Ca 0.023 0.045 0.027 128.259 0.7242 Pentadecane group 3.988 4.890 4.041 204.8 0.845
61 1-M,3-E-Cy-C6 0.078 0.155 0.083 126.243 0.8000 Hexadecane group 3.417 3.937 3.438 217.9 0.851
62 1-M,4-E-Cy-C6 0.034 0.068 0.037 126.243 0.7900 Heptadecane group 4.286 4.577 4.355 235.1 0.842
63 Unspecilied 0.006 0.013 0.007 126.243 0.7900 Octadecane group 3.060 3.076 3.099 249.8 0.845
naphthene Nonadecane group 3.749 3.604 3.756 261.2 0.854
63 Unspecilied 0.004 0.007 0.004 126.243 0.7900 Eicosane plus 58.781 34.772 56.664 424.5 0.888
naphthene
64 n-Cg 0.471 0.923 0.559 128.259 0.7214 Apparent average molecular weight: 254.0 Source: Osjord el al. (1985).

O Unspecilied Cg 0.124 0.243 0.148 128.259 0.7200 Moiecular weight corrected lor benzene: 251.1
O Nonanes total 2.427 5.241 2.598 116.277 0.7995 Measured density (g/cm3) 15C, 1 atm: 0.856

Source: Osjord el al. (1985).

22 Properties 01 Oils and Natural Gases Compositional Determinations 23

Table 2-9 vs. carbon number profiles, corresponding to different PNA distributions.
PNA Report This is in agreement with observations by Yarborough et al. (1978).
Paraffinic, Naphthenic, and Aromatic Contents of the Boiling Point Similarly, Figure 2-7 shows the molecular weight vs. carbon number for
Groups up to n-Cg several different oil samples. As observed by Katz and Firoozabadi (1978),
the average molecular weights of the hydrocarbon fractions are not signifi-
Molecular Density
cantly different for the different oils. This can be explained by the fact that
Component Weight % Mole % Volume % Weight (g/cm3)
even though the individual molecular structures within each hydrocarbon
Hexane group 0.647 1.886 0.836 86.2 0.663 fraction may vary considerably, from n-paraffins to isomers to cyclic struc-
paraffins
Hexane qroup 0.052 0.185 0.059 70.1 0.750
tures, their molecular weights wiJI not be very much different. Different
naphthenes structures within a given hydrocarbon fraction have ver y similar ratios of
Heptane group 0.713 1.787 0.889 100.2 0.686 hydrogen atoms to carbon atoms, and hence the molecular weights are also
paraffins
similar.
Heptane group 0.930 2.682 1.034 87.1 0.769 lt is thus fe asible to use general, average molecular weights for the frac-
naphthenes
Heptane group 0.355 1.140 0.343 78.1 0.884
tions from Cg to C19. The average values of the molecular weights of Figure
aromatics 2-7 and the average densities of Figure 2-5 are shown in Table 2-12.
Octane group 0.870 1.912 1.054 114.2 0.707 Unlike Katz and Firoozabadi, general, average density values for each
paraffins ,. fraction are not used. As shown in Figure 2-6, the density vares considera-
Octane group 1.404 3.435 1.556 102.6 0.772 I
bly with the structure (PNA distribution) of the hydrocarbon fraction.
naphthenes
Octane group 0.958 2.610 0.941 92.1 0.871
aromatics
Nonane group 0.739 1.446 0.877 128.3 0.721
paraffins
Nonane group 0.646 1.331 0.699 122.0 0.792
naphthenes
Nonane group 1.042 2.464 1.022 106.2 0.872
aromatics
11,
Source: Osjord el al. (1985).

Table 2-11 compares compositnal data obtained by gas chromatography

and by distillation. There is generally satisfactory agreement. One advan-
tage of the analysis by distillation is, however, that it produces a physical
sample of each hydrocarbon fraction present in the oil sample.

PROPERTIES OF HYDROCARBON FRACTIONS

For the hydrocarbon fractions (such as CIl), and the residue (e.g., Czo+),
it is important not only to know their weight fraction, but also their physical
properties such as densities, molecular weights, and PNA distributions.
These properties will be different for each oil and may thus help to distin-
guish an oil of one reservoir from that of another reservoir.
Figures 2-5 and 2-6 show plots of the density vs. carbon number for sev- Figure 2-4. Fischer HMS 500 mini-distillation equipment modified for continuous
eral North Sea oil samples. There are significant differences in the density distillation of crude oils.
24 Properties 01 Oils and Natural Gases Compositional Determinations 25

Table 2-11
.~ ~ Comparison Between Results from a TBP Distillation and a Capillary
; o~mw~oororoWNN~~Wroro-
Q)
W~ID~mvOvmomOvO~ONO GC-Analysis. (Average of Six)
3 E ~M~OO~~O~W~~OOONM~WO
E-=
8~
~NMmIDo~~rooooommmmmo
. - Oistillation GC-analysis

MW Oensity Weight MW Oensity RSO'

~ Welght
(%)
Q)ION~~_C')O~orow_~ro_C')oC') Fraction (%) (g/mol) (g/cm3) (%) (g/mol) (g/cm3)
E ",,,,:~'<I;,,,'<I;,,,~~~,,,,,,:'<I;,,,~~,,!~
:J lOCOCOLOCOcnt{)f'..LOV("')C')(\J .....
,....,.....,-C"')
T""" (") 0.886 65 0.621 0.792 63.1 0.597 0.28
_ '1"'"""
C5
~ C6 0.737 82 0.695 0.699 84.8 0.669 0.26
2.371 91 0.751 2.000 89.4 0.754 0.14
>~ C7
~MElvCOCOCO("')("')lOcocomoo~colOlOV,....
103 0.778 3.237 102.0 0.779 0.21
~
e o
WNwC')wroNmmONC')C')vvv~m
~ww~~~ro~~rororororororororo C8 2.825
0.39
~S cicici6cici6cicici6ci6ci66ci6 C9 2.539 116 0.793 2.429 116.3 0.799
90.642 306 0.869 90.846 300.3 0.868 0.002
ClO+ 0.49
~ C'0 2.479 132 0.798 2.437 133.6 0.801
Q)
i~ NCO("')CJ)COCOOVOOVMf'..,....l!)Vl!)CJ)
1.916 147 0.803 2.191 148.0 0.803 0.40
~
c::
3 Q)
lOCJ)vCOONOlOO,....COcnvv(")oco,....
oooom~w~ci~wm~M~w~ooooci e"
C 2.352 163 0.817 2.523 161.5 0.812 0.33
EO ,....C\Jvco,......ocoooCOCJ)CJ)CJ)CJ)CJ)CJ)CJ)O '2
Q)
;;s::E 2.09l 175 0.836 3.106 175.3 0.827 1.06
"C U C'3
c:: 0.840 1.13
U
o ..
~IN~wwmrovrowovmvv~m_v
C14
C'5
3.667
3.722
190
205
0.843
0.849
3.124
3.984
189.8
204.8 0.845 1.60
"C ~~"":~~r:":~'"":~":~~~~~~t.t1 0.851 6.05
Q) o::IV COOOCOCOOlV<OV(,,)C\lNT"""OOOQ'I"'""" C'6 2.034 215 0.853 3.383 217.9
235.1 0.842 3.82
C\J 0.844 4.244
~ C17 4.135 237
'1"'""""- T""" T"""

:cea ..
.!!! .
C'8
C'9
3.772
3.407
251
262
0.846
0.857
3.201
3.523
249.8
261.2
0.845
0.854
12.12
5.07
o ;;sJ: roNro~C')O
59.130 421.6 0.888 1.31
'";' ea ld .21 ~N~~~mlOC')lOco,....(",)"""o<O'I"'"""om
lOf'..cocnO ..... f'..C")vlOf'..COm,....C\JvlO.,...- C20+ 61.057 426 0.885
N-
o
o~ "1"':"'""""""''1'''""" '1"'"""'1"'"""'I"'""""-T'"'N NNC\JM
Relative standard deviation on weight % from GCanalysis.
Q) ::E
- O
.c c:: So~rce: Osjord el al. (1985).
!!l-.: ~~ mf'..VN,......mOf'..cnf'..co"",V,....MNCOO
~
-.:.!!l El: ~~~~~~~~~~~~~~":~~~
e
a..
ID
;;s
E 'j
83:
CI vorov_oorovroNwmoNv~O
-
'I"'"""'I"'"""("')lOCOOCO"""""cocoCOCJ)CJ)CJ)CJ)O
- Measurement of Density

1- ~
..
lmro~ro~ __ roNm_C')C')roNmwN
E s: T"""f'..O>T"""OOOOvCOMC\Jv"""'f"...mvMv
Density measurements of the hydrocarbon fractions and residues are

--
~
ea
ea
~
'
3:
~w~~~mm~~~~MN~~~~~
,.... '1"'""" M readily carried out using either a pycnometer or a Mettler/Paar densitome-
ter. The latter method [Kratky et al. (1969), and Leopold (1970) 1 is based on
e -g6 LOC\JO)..-C") f'..W(")LO"",.,.....mCX)("')C\lN
density-dependent frequency changes of an oscillating glass U-tube. A Met-
Q)
J:~

c.rimcciwr= ocriui"":w"":M~6NN tler/Paar instrument is shown in Figure 2-8.
IC')<DO'>~~ ~rooNC')~wr--moo
(ijci. T"""T"""T"""T"""'I"'"""T"""T"""C\lC\I The instrument is calibrated against air and water at each temperature of
.ao E
Q)
1\
interest. With the two calibration points, the unknown density is readily ob-
-
tained from the measured frequency of the U-tube .

~~I
~~
.F

~
"'''!~''':~~''''<I;N~~''':~~~~~
comcoco.,...-.,...-veof'..lOM
I("')ID~N~~~~~M~~OOO~MM
~~~~~NNNNNMMMM
.....
f'..Mf'.......
,....

Measurement of Molecular Weight

; - 1\ 1\
UU
~
e Several techniques have been proposed for determining average molecu-
~ lar weights of hydrocarbon fractions and residues. The techniques are based
u U) + +
t! ctj U') W ,... a) 01 ~ ~ :: ~ ~ :! ~ ~ ~ -~ ~ ~ on boiling point elevation, freezing point depression, vapor phase osmosis,
u. Cl
 26 Properties o] Oils and Natural Gases eom positional Determinations 27

0.90
0.89 0.98
0.88
0.96
0.87
0.94
0.86 <.-;:; iflli

---
t<) 0.92
f'I) 0.85 ~
2 0.84 ~ 0.90
U el ~...---.> --
0- 0.83
~ 0.88 f-;.;;:r.,r-------
<, 0.82 /G// '-""-/'---
>- Vi 0.86
r- 0.81 Z ,;7 ".,'
W .s-:: Aromatic Oils '
(/) 0.84
O Medium Paraffinic Oils
Z
w 0.80 /,./ //
O
0.79
0.82
.~/ ------------- Paraffmic Oils
Aromatics

0.78 0.80 _._"'//''''' -

.. ........ ~~e.h-"~I:':'
Katz & Firoozabadi
_
/ ---------
0.77 0.78
--/ /'

0.76 0.76 I I I I I I I I I I
9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
I I I I I I I I I I I I

0.75
CARBON NUMBER
0.74 .
8 9 10 11 12 13 14 15 16 17 18 19 Figure 2-6. Density profiles of North Sea oils and condensates.
CARBON NUMBER ~.,. 400 ~//::~.
Figure 2-5. Density 01 distillation Iractions Irom paraffinic North Sea crude oils //'/ /./ .. '

380 / / / .
and condensates.
,/"<./
//~<,/

.:
360 /. .....
~ 340 ~// /.'
O
~
el 320
Table 2-12 'S...
Average Densities and Molecular Weights of 11 Paraftinic el 300
North Sea Crude Oils and Condensates ':'--.
~ 280
el /i~'~""
/jz....
/
W 260
Molecular Weight Density /~ .....
Fraction (g/mol) (g/cm3)
3:
240 Aromatic Oils
o::: 0

Medium Paraffinic Oils

e9 118.6 0.778 j 220'
elO 133.6 0.790 ::> Ea!!'t~i.<:.QiI~. __ o __

148.0 0.792
o 200 Aromatic Fraction
e" ~ ~~e.h_a!i:
~~~~~'!. _
e, 161.5 0.801 O lBO
c., 175.3 0.816 ~ ~~~~-
160
c., 189.8 0.829
c., 204.8 0.834 140
c., 217.9 0.840
120~ ( I I I I I I I I I I I I I I I I I I I I
c., 235.1 0.837 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
c., 249.8 0.840 CARBON NUMBER
c., 261.2 0.849
Source: Osjord el al. (1985).
Figure 2-7. Molecular weight profiles 01 North Sea oils and condensates.
 r
I
t

28 Properties 01 Oils and Natural Gases Compositional Determinations 29

......
91220

~k~.J.';
tUl 8Y"T"MTII<
&tl.~, _
_

I< .
;

,
yield the average molecular weight of the oil sample. The method is very
sensitive and, therefore, difficult to use, and must be carefully calibrated.
A molecular weight apparatus based on freezing point depression, and of
type Cryette A is shown in Figure 2-10.

Freezing Curve of Sample

Temperature

--- ---
~
1 __
liT
--=-- Freezing Temp.
Benzene

Time

Figure 2-9. Freezing point depression diagram.

Figure 2-8. Mettler/Paar Densitometers.

gel permeation chromatography, and other factors. All of these methods

have limited ranges of accuracy and should be applied to specifc ranges of
molecular weights.
Most oils, oil fractions, and residues have molecular weights in the range
of 100-500. This covers the hydrocarbons from C7 to C35. The method most
suitable for determining molecular weights within this range is that based
on freezing point depression.
The principle, of this method is that the freezing point of apure solvent
(e.g., benzene or nitrobenzene) is lowered when the solvent is contaminated
with a small amount of oil. Figure 2-9 shows a typical temperature vs. time
curve whea a solvent is cooled. The horizontalline corresponds to the melt-
ing point of the pure solvent, Tm. The temperature-line curve for the oil-
contaminated solvent will after a certain interval reach a constant value <1
lower than Tm. The temperature difference, .6. T, is a measure of the molecu-
lar weight per mole of a sample of the oil. If the apparatus is calibrated us- Figure 2-10. Apparatus tor measurement 01 molecular weight by freezing point
ing substances of known rnolecular weight, the observed value of .6. T may depression in benzene.
 30 Properties o] Oils and Natural Gases Compositional Determinations 31

INTERNAL CONSISTENCY OF ANALYTICAL DATA if the two values of oil sample average molecular weights agree to within 2-
3 molecular weight units.
In addition, one may back-calculate the values of CI+ and C20+ molecu-
To this point, most of the analytical data from the gas chromatographic
lar weights, which also must agree closely with the experimentally observed
measurernents and all of the data from the distillation are available in terms
values for Co+ and C20+.
of weight fractions. The molecular weights, therefore, play an important
The average density of the oil sample, p, can also be calculated from the
role in converting the weight fractons to mole fractions. The molecular
data for the fraction given in Table 2-14 as follows:
weight of the residue is especially important.
The previo LIS section indicates that the molecular weight measurement is
dffeul to perform. Often relatively large errors are associated with experi-
p = E w/E w/p (2-2)
mentally determined molecular weights. Fortunately, when a distillation
analysis has been performed, it is possible to check the reported weight frac-
Table 2-14
tions and associated molecular weights of components, hydrocarbon frac- Data for Computing Table 2-13
tions, and residue, for consistency. Only an additional measurement of the
111'-, ~
molecular weight of the whole oil sample is required. . 1 2 3 4 5 6
For example, an analysis by distillation of an oil sample results in the in- Density
formation shown in Tables 2-13 and 2-14. In order to check the consistency Component Weight % MW Mole % g/cm3
of the reported C20 + molecular weight, the average molecular weight of the Gas 4.19 54.8 8.52 4.7 0.~64 7.4
whole sample is calculated as follows: es 6.78 72.2 10.45 7.5 0.628 10.8
e6 7.97 84.8 10.46 8.8 0.668 11.9
e7 15.18 91.5 18.46 16.8 0.738 20.6
MW ~ (~ w, I MW,J-' (2-1) es 17.05 104.3 18.19 19.0 0.763 22.3
1'1: e9 9.81 119.0 9.18 10.9 0.783 12.5
l'
),
(elO+)
elO
(39.01)
7.48
. (175)
133
(24.74)
6.28
(43.3)
8.4
(0.825)
0.796
(47.3)
9.4
where w, is the weight fraction and MW the molecular weight of fraction i. el1 5.82 145 4.46 6.5 0.796 7.3
This value may be compared with a measured mean molecular weight for 1,
e, 4.39 158 3.10 4.9 0.809 5.4
the whole oil sample. The analysis data are considered internally consistent
!I c., 4.21 171 2.74 4.6 0.820 5.1
I c., 3.43 183 2.09 3.8 0.830 4.1
II c., 2.73 197 1.54 3.0 0.837 3.3
Table 2-13 c., 1.78 210 0.94 2.0 0.848 2.1
Measured and Calculated Molecular Weights and Densities from c., 1.92 226 0.95 2.1 0.845 2.3
North Sea Gas Condensate c., 1.49 241 0.69 1.7 0.845 1.8
c., 1.36 250 0.61 1.5 0.854 1.6
e20+ 4.42 319 1.54 4.9 0.891 5.0
Condensate c-. c-.
Measured molecular E 100.00 100.00 111.1 132.9
weight (MW) 111 175 319 Colurnn 3 is obtained by: (column l/column 2)/E(column l/column 2)
ealculated MW
Colurnn 4 is obtained by: column 2 x column 3/100
using elO+ MW 111 Colurnn 6 is obtained by: column l/column 5
ealculated MW
ealculated MW using elO+ MW = (4.7 + 7.5 + 8.8 + 16.8 + 19.0 + 10.9 + 43.3) = 111.0
using e20+ MW 111 175 Calculated MW using e20+ MW = 111.1 (see table)
Measured density (g/cm3) 0.756 0.825 ealculated MW 01 the elO+ Iraction using e20+ MW = 100(8.4 + 6.5 + 4.9 + 4.6 + 3.8
0.891
ealculated density + 3.0 + 2.0 + 2.1 + 1.7 + 1.5
using elO+ density (g/cm3) 0.752 + 4.9)/24.74 = 175
ealculated density ealculated density using elO+ density = 100/(7.4 + 10.8 + 11.9 + 20.6 + 22.3 + 12.5
using e20+ density (g/cm3) + 47.3) = 0.752 g/cm3
0.752 0.826
ealculated density using e20+ density = 100/132.9 = 0.752 g/cm3
32 Properties 01 Oils and Natural Gases Compostonal Determinatons 33

Table 2-15
where w is the weight percent of component i and p the density of compo-
Composition Analyses, Gas/Oil Ratios, and Formation Volume Factors of
nent or fraction i.
Separator Fluids
Again, the calculated and measured average oil densities should agree
closely. Stock Tank Stock Tank Separator
Oil Gas Gas
Component Mole % Mole % Mole %
Determination of Reservoir Fluid Composition Nitrogen 0.00 0.20 0.66

Carbon dioxide 0.00 3.96 5.65

The reservoir fluid arriving from the well to the platform is separated into Methane 0.00 24.86 68.80
.dli
a "separator lquid" and a "separator gas" at platform separator tempera-
Ethane 0.20 20.40 12.86
ture Tp and pressure Pp' The separator liquid is in turn flashed at ambient
conditions (T; and Po) in the laboratory. The principle of the two stage flash Propane 2.14 28.42 7.94
separation is shown in Figure 2-11. The pracedure is described in more de- ;-Butane 1.10 4.78 0.94
tail in Chapter 3.
n-Butane 4.25 10.97 1.96
The reservoir fluid composition is obtained by combining the analyses of
;-Pentane 2.68 2.21 0.34
the laboratory gas and the stock tank oil, from which the composition of the
separator liquid is determined. The analysis of the separator liquid is in turn n-Pentane 4.32 2.53 0.42
combined with the analysis of the separator gas to yield the composition of Hexanes 6.66 1.05 0.22
the reservoir fluid (or well stream).
Heptanes 11.90 0.54 0.15
To recombine the previously mentioned analyses correctly, it is necessary
. Octanes 13.14 0.10 0.05
to know the gas-to-liquid ratio, the so-called gas/oil ratio (GOR), both at
platform and at ambient conditions. In Chapter 3 it is shown how to mea- . Nonanes 7.73 0.00 0.00
sure gas/oil ratios in the laboratory. The princple of the flash experiment is Decanes plus 45.89 0.00 0.00
shown in Figure 2-11. An example of such data is given in Table 2-15.
E 100.00 100.00 100.00

Gas/oil ratio at separator, GORo = 442 Sm3/m3 separator oil

Gas/oil ratio at ambient flash, GORo = 36.9 Sm3/m3 stock tan k oil
Flash lormation volume lactor 01 separator liquid = 1.165 m3/m3 stock tan k oil
Figure 2-11. Principie of two-stage
I Gas
flash of well stream. Average molecular weight 01 the stock tan k oil = 160.4 kg/mole

Basis for the recombination can be, e.g., 1 std m'' of stock tank oi!. The
Oil total number of moles of each component (or fraction or residue) in this std
r, Pp
rn'' of stock tank oil is calculated. The molar volume of an ideal gas at atmo-
spheric pressure and OC equals 22.4 m'. The total individual number of
First Stage
(separator
I . Gas moles of gas sample corresponding to 1 std m3 of liquid sample may be ob-
conditions)
tained fram:
~
Moles gas = GOR/ (22.4 x T0/273) (2-3)
r, Po f-------Oil
where GOR is either GORp or GORo [platform (field), or ambient condi-
Second Stage tions, respectively]. From the moles of gas plus the composition of the gas
(ambient samples, the amounts of individual gas-phase components in both gas
conditions) streams can be determined.
Feed
 34 Properties 01 Oils and Natural Gases
Compositional Determinations 35

Finally, the total number of moles of each component from the test sepa- Table 2-16
rator phases are added (see Figure 2-11) and normalized to yield the compo- Number of Moles of Each Component in Each Separator Stream
sition of the recombined fluid, i.e., the fluid in the reservoir.
Stock Tank Stock Tank Separator
Component Oil Gas Gas
Example: Recombination of Stock Tank Oil, Stock Tank Gas, and Separator
Nitrogen - 0.0031 0.1438
Gas. A reservoir fluid is flashed at the field separator, and the separator liq-
uid is flashed to ambient conditions in the laboratory, yielding a stock tank Garbo n dioxide - 0.0618 1.2311

liquid and an evolved gas. The composition analysis and other experimental Methane - 0.3878 14.9915
information are given in Table 2-15. Ethane 0.0107 0.3182 2.8022
From that information, first compute the number of moles in each of the 0.4434
Propane 0.1147 1.7301
streams. As a basis, 1 m3 of stock tank oil is chosen.
i-Butane 0.0590 0.0746 0.2048

Moles stock tank oil = 860 kg/160.4 kg/mole = 5.36 moles ir; n-Butane 0.2278 0.1711 0.4271

i-Pentane 0.1436 0.0345 0.0741

Moles of evolved gas = 36.9 m3/(22.4 moles/m- x 288K/273K)
n-Pentane 0.2316 0.0395 0.0915
= 1.56 moles
Hexanes 0.3570 0.0164 0.0479

The volume of gas (Sm3) per 1 rrr' stock tank oil is 442 x 1.165 = 515 Sm3 Heptanes 0.6378 0.0084 0.0327

Octanes 0.7043 0.0016 0.0109

The number of moles of separator gas is thus:
Nonanes 0.4143 - -

515/ (22.4 x 288/273) = 21. 79 moles

',.. Decanes plus 2.4597 - -
f"
t:; ; 5.36 1.56 21.79
i-ti
The total number of moles of the reservoir fluid per m3 stock tank oil is :610
thus: ''

The recombination computations are normally performed for composi-

tions to CJO +. However, for equation of state calculations computations to
5.36 + 1.56 + 21. 79 = 28.71 moles
C20+ are preferred. The CJO to C20+ compositions are normally obtained by
extending the recombination result using the distillation data. Using these
The number of moles of each component in each stream may now be com- .~jl data, the recombined CJO + -fraction is divided into its appropriate ratios up
puted by multiplying the mole % given in the compositional analysis by the f',y, to C20+.
total number of moles in each stream and dividing by 100. The results are ~ The result at this stage is a well checked, consistent composition analysis
shown in Table 2-16. of the reservoir fluido
Finally, the number of moles of each component in the reservoir fluid is
obtained by adding the number of moles in each stream:
EXPERIMENTAL EXTENSION BEYOND C20+
Moles of n-pentane in the reservoir fluid = 0.2316 + 0.0395 + 0.0915
= 0.3626 moles In some cases (e.g., dew-point calculations) quantitative extension be-
yond C20+ is advantageous.
The mole % of n-pentane in the reservoir fluid = 0.3626 x 100/28.71 Breakdown of the C20+ fraction may be obtained by distillation, either at
= 1.26 mole % high temperatures (up to 550C) or at low pressures (- 2 mm Hg). This is a
ver y time consuming and difficult analysis, because the process conditions
The final molar composition of the reservoir fluid is shown in Table 2-17. are not easly controlled.
 -ae

36 Properties of Oils and Natural Gases Compositional Determinations 37

Table 2-17 Table 2-18

Reservoir Fluid Composition Extended Compositional Data for a North Sea Gas Condensate

Moles in Density PNA Distribution, Mole %

Reservoir (g/cm3)
Component Fluid Mole % MW P N A
Component Mole % 1 atm, 15C
Nitrogen 0.1469 0.51 N2 0.64
Carbon dioxide 1.2920 4.50 CO2 9.16
Methane 15.3793 53.57 f C, 68.8

Ethane C2 8.43
3.1311 10.91
C3 5.11
Propane 2.2882 7.97 ,1
i-Butane
ic, 0.81
0.3384 1.18
n-C. 1.45
n-Butane 0.8260 2.88
te, 0.52
i-Pentane 0.2522 0.88 n-C, 0.53
n-Pentane 0.3626 1.26 C6 0.63
Hexanes 0.4213 1.47 C7 0.83 0.741 96 0.50 0.42 0.08

Heptanes 0.6789 . 2.36

Cs 0.95 0.780 107 0.45 0.38 0.17
C9 0.52 0.807 121 0.48 0.27 0.25
Octanes 0.7168 2.50
ClO 0.26 0.819 134 0.47 0.30 0.23
Nonanes 0.4143 1.44
Cll 0.20 0.810 147 0.56 0.27 0.17
Decanes plus 2.4597 8.57
C'2 0.17 0.828 161 0.55 0.24 0.21
r; 28.71 100.00 0.16 0.849 175 0.54 0.22 0.24
C'J
C,. 0.15 0.857 190 0.49 0.27 0.24

C'S 0.11 0.868 206 0.52 0.20 0.28

An example of such analysis is given in Table 2-18. The apparatus used is 0.086 0.872 222 0.55 0.19 0.26
C'6
a modified Fischer HMS 500 distillation apparatus (see Figure 2-4), operat-
C17 0.078 0.859 237 0.57 0.20 0.23
ing at 2 mm Hg. The analysis time is 2-3 days. By this distillation it is possi-
C,s 0.068 0.854 251 0.70 c. 0.11 0.19
ble to extend the experimental composition analysis to C30+.
C'9 0.050 0.866 263
There is, however, an empirical method of extrapolating the experimental
C20 0.046 0.873 339 C20+
C20 + analysis, described in Chapter 7. This method performs well for most
purposes and thus saves the time-consuming C20 to C30 + analysis. C2, 0.035 0.876
C22 0.025 0.876
C23 0.034 0.875
C2 0.023 0.877
C2S 0.017 0.876
C26 0.018 0.878
C27 0.014 0.882
C28 0.012 0.886
C29 0.013 0.889
CJO+ 0.047 0.908

38 Properties oj Oils and Natural Gases
REFERENCES
Katz, D. L., and Firoozabadi, A., "Predicting Phase Behavior of Conden-
sate/Crude-Oil Systems Using Methane Interaction Coefficients," lournal
01 Peto Techn., Nov. 1978, pp. 1649-1656. 111'
Kratky, O., Leopold, H., and Stabinger, H., "Density Determination of Liq-
uids and Gases to an Accuracy of 10 - 6 g/cm'', with a Sample Volume of
only 0.6 cm"," Z. angew. Physik, 4, 1969, pp. 273-277.
Leopold, H., "Die Digitale Messung der Dichte von Flssigkeiten,"
Elektronik, 19, 1970, p. 297.
Osjord, E. H., and Malthe-Serenssen, D., "Quantitative Analysis of Natural
Gas in a Single Run by the Use of Packed and Capillary Columns,' ]. 01
Chromatography, 297, 1983, pp. 219-224.
Osjord, E. H., Renningsen, H. P., and Tau, L., "Distribution of Weight,
Density, and Molecular Weight in Crude Oil Derived from Computerzed
Capillary GC Analyss,' Journal 01 High Res. Chrom. & Chrom. Comm.,
8, 1985, pp. 683-690.
Yarborough, L., "Application of a Generalized Equation of State to Petro-
leum Reservoir Fluds,' Paper presented at the 176th National Meeting of
the American Chemical Society, Miami Beach, FL, 1978.
l"
,.'
\.
"
Chapter 3
Oil and Gas Property Measurements
This chapter describes some of the most commonly used methods for mea-
suring densities, viscosities, saturation points, and other properties of reser-
voir fluids. It is not possible within one chapter to describe all the relevant
methods in detail. An attempt is made, however, to give the reader an un-
derstanding of the methods used to collect the data shown in this book.
SAMPLING
Hydrocarbon samples to be analyzed for the previously mentioned prop-
erties are normally collected at the first stage (field) separator. Representa-
tive samples of both the separator liquid and the separator gas streams are
collected. Sarnples may 'also be obtained from the bottom of the well, so-
called bottom hole samples. These are, however, not very common as they
are only recommended when taken from undersaturated reservoirs.
Before starting an extensive analysis it is important to evaluate the aval-
able samples and to choose the sample thought to be the most representative
for the reservoir hydrocarbon fluids. Samples obtained during the most sta-
ble separator conditions are normally chosen. In the laboratory, the satura-
tion pressure (bubble point) of the separator liquid is determined at separa-
tor temperature. The sample is considered to be representative if the
saturation pressure is equal to the test separator pressure at the sampling
time.
The opening pressure of the corresponding gas phase sample may also be
checked. It should be equal to the separator pressure at sampling time.
If both of these pressures agree with the value of the separator pressure,
further processing of the samples may take place.
FLASH SEPARATION AND COMPOSITIONAL ANALYSIS
After quality control, the separator gas sample may be subjected to corn-
positional analysis as described in Chapter 2. The gas sample flask is heated
39
40 Propertes of Oils and Natural Gases Oil and Gas Property Measurements 41

to separator temperature and left to equilibrate. Figure 3-1 shows a typical

apparatus for this purpose. A sample of the gas is bled through a gas sam-
pling valve and injected into the gas chromatograph. From the cornposi-
tional analysis it is possible to detect whether the sample has been contami-
nated with, for example, air.
In order to analyze the separator liquid sample, which has been collected
under pressure, it must be stabilized to atmospheric conditions by a single
flash process. The high-pressure sample container is, therefore, connected to
a single-stage flash separator as shown in Figure 3-2. The separator liquid
sample is thus allowed to separate into two phases. Figure 3-1. Separator gas
11
The volume of the gas phase is determined. The liquid is allowed to settle sample f1ask rigged for
in the bottom of the flash chamber, after which the liquid is drained and chromatographic analysis.
weighed, and its density is determined.
The gas/oil ratio of the separator liquid sample, CORo, may now be cal-
culated:

CORo = volume gas (Sm3)/volume oil (Sm3) (3-1) t

where Sm3 means the volume of fluids at standard conditions, normally
15C and one atmosphere.
The composition of the gas and liquid phases may be determined as de-
scribed in Chapter 2.

;l\

PVT MEASUREMENTS
f
/l./
:ti
This section describes the types of PVT measurements (often called PVT ~.;

analyses) most commonly performed on reservo ir fluids. In order to study

;~'
the properties of the reservoir fluid, it is necessary to recombine the separa-
tor gas and the separator liquid samples. This is done by mixing the two i:
Figure 3-2. Single stage
samples proportionally to their produced rates (CORp).
Ruska model 2353-803
There are two main types of PVT equipment available: PVT systems for flash separator.
black oils and PVT systems for gas condensates and volatile fluids. These are
discussed separately. t;

BLACK OIL EQUIPMENT

A typical schematic diagram of a black oil PVT.system is shown in Figure

3-3. The main parts of the system are a mercury pump, a calibrated steel
cell, a heating bath with silicone oil, and a precision manometer or dead
weight gauge. The principle of operation is as follows:
;1'1
"
'----'
 42 Properties of Oils and Natural Gases Oil and Gas Property Measurements 43

1. Heating bath with silicone oil The mercury pumps can be either hand or machine operated. In both
2. PVT cell (volume normally :f cases the pumps must be graduated and calibrated to a volumetric scale. By
retracting the cell pump and advancing the charge purnp, a known volume

f
0.6 - 1.0 1)
3. Valve of sample can be transferred to the cell.
Q) 4. Graduated mercury pump The cell is immersed in a thermostat and heated to the reservoir tempera-
ture, or any other required ternperature, before the cell is charged. This is
5. Precision manometer
important because heating after charging could, without proper monitoring
6. Cell entry valve . ,\
of the oil pressure, lead to excess cell pressures due to thermal expansion of
o
)
:. ..~

~
t :1,
j "<
... :~.
the mercury and the sample. If the sample has been brought into the two-
*"
.~
r phase regin, heating after charging is normally safe but should be moni-
i ffl' tored.
';,

CD Figure 3-3. Schematic diagram of

black oil PVT apparatus.
Bubble-Point Determination

Once the system has been equilibrated at the desired (usually reservoir)
temperature, the PVT analyses can start.
Normally, the first analysis performed is the determination of the bubble
point. The bubble point of a system is defined as the pressure where, at con-
stant ternperature, the oil system is in equilibrium with an infinitesimally
small gas bubble.
The PVT cell is accessed by opening valve 6 shown in Figure 3-3, and
mercury is injected into the cell by means of the mercury pump. The
amount of mercury is recorded after having closed valve 3, and the system is
equilibrated at the resulting pressure. Based on the amount of mercury in-
jected into the cell, and the amount of sample charged to the cell, one can
calculate the sample volume at this pressure. This process is repeated several
(0 times, and the resulting volume pressure data is plotted as shown in Figure
1. Heating bath - with silicone oil 3-5.
2. PVT cell Normally, a clear discontinuity is found. This point is the bubble point of
3. lsolating valve the mixture at the given temperature.
4. Graduated mercury pumps (a) cell (b) charging Several factors influence the accuracy by which the saturation pressure
5. Precision m"~ometers can be determined by the graphical technique. The change of slope in Fig-
6. Cell entry valve ure 3-5 depends on the compressibility difference between the two-phase
7. Hydrocarbon sampling bottle or recombination cell system and the single-phase system. At high pressures and temperatures, and
when the system approaches the critical region, this difference is not very
Figure 3-4. Black oil PVT apparatus rigged for sample transfer.
large.

A ready, recombined sample is charged to the cell at constant pressure

through valve 6 at the top of the cell. The cell may be equipped with one or
more windows for monitoring the charging and subsequent operations.
The charging process is performed at the pressure in the sample cylinder.
Constant pressure is maintained using the mercury pumps shown in Figure
3-4.
Differential Depletion
Other important properties of the oil system can be derived through a d-
ferential depletion or differential liberation experimento These properties
are:
44 Properties oi Oils and Natural Gases Oil and Gas Property Measurements 45

Gas off AII gas

displaced
... . . Gas.
.
t'i
r :o
: Gas:
r >"'1
v
N
1 1: Gas.

(1J
E
>"' > Oil
:J 11 Oil 11

1 I
~ -> Oil >C'l
t
11
V -;;;
s >~
>"'
ro v
One Pha se >"' Oil >~I Oil
v v
>N >LD
Ps Pressure

Figure 3-5. Black oil bubble point determination.

L 1
1. Shrinkage factor, Bo. This is a measure of the ratio of the volume of
P1 - Psat P2 < Psat P2 < Psat P2 < Psat P3 < P2< Psat
the hydrocarbon system at reservoir conditions to the volume at stock
tan k conditions: Figure 3-6. Schematic representation 01 a differentialliberation experiment.

Bo = res. m ' ol/Sm ' stock tank oil (3-2)

2. Solution gas/oil ratio, Rs. This is a measure of the volume of gas in

solution at given pressures (see also Equation 3-8):

Rs = Sm3 gas/Sm ' stock tank oil (3-3)

3. Density of the reservo ir fluid at different pressures, including the satu-

ration pressure:

p = kg of Fluid/rn'' (3-4)

4. Real gas deviation factor Z, i.e., compressibility factor, for the gas
phase at given pressures and temperatures:

Z = PV/nRT (3-5)

where V is the volume and n the number of moles.

A schematic diagram of the differential depletion experiment is shown in
Figure 3-6. The equipment is shown in Figure 3-7. Figure 3-7. A PVT system used tor differentialliberation analyses. One sees
The reservo ir fluid is, at constant temperature, brought to a pressure from left to right: the motorized pump, PVT cel!, and bath gas col!ection sys-
above its saturation pressure. After this, the pressure is brought to a value tem with condensate trapo
46 Properties 01 Oils and Natural Gases Oil and Gas Property Measurements 47

below the saturation pressure. The system will then be in the two-phase re-
gion. After the system has been equilibrated, the gas phase is removed from
the cell at constant pressure. The gas volume is measured with a gasometer
and recorded.
The new volume of the oil is calculated from the readings on the mercury
pump before and after the discharge of the gas. These readings will also give i
~';ftt
the gas phase volume at the given pressure and temperature. M

The hydrocarbon system is again brought into the two-phase region, this ~ Ps
time at a lower pressure, and the gas discharge procedure is repeated. '" Vs.
~
The pressure step intervals are chosen so that 8-10 stages will occur be- ~
tween the saturation pressure and the atmospheric pressure.
At the last stage, care must be taken to discharge al! of the remaining gas,
whereas the residual oil is left in the cell. After the last part of the gas is
discharged, the cell is closed, its internal pressure is increased in steps, and
the resulting ol volumes are recorded. Normally, one records volumes at
100 200 300
each 50 bar up to 300 bar. The plot of these volumes can be extrapolated
Pressure [Bar J
back to atmospheric pressure, so that a measure of the residual oil volume at
the given temperature, which is often that of the reservoir, can be deter- Figure 3-8. Determination of saturation pressure and volume from a differential
mined. depletion experiment.
Then the cell temperature is reduced to 15C, and the oil volume at this
temperature is determined as previously explained. This volume is the refer-
ence residual volume.
The oil is then discharged from the cell, and its density s determined as Bo .s
explained in Chapter 2.
The data recorded in this analysis can be represented as shown in Figure
3-8. Ps and Vs are the saturation pressure and volume, respectively; VRT the ~
o

liquid volume at one atmosphere and temperature, Ti and V R is the refer- ~

ence liquid volume at 15C. From these data the shrinkage factor of the sys- o
tem at the saturation pressurernay be calculated: '"

Bo,s = VSIVR (3-6)

The whole Bo-curve (see Figure 3-9) may be calculated as follows: .,
100 200 300
Pressuro [Bar)
Bo = VIVR (3-7)
The solution gas/oil ratio of the reservoir fluid, Rs, at pressure stage N can Figure 3-9. The differential liberation Bo-factor as function of pressure.
be obtained from
NSTEP
Rs =
n~N
L+ 1
V gas,nlV R (3-8) The density of the reservo ir fluid at each pressure reduction step can be
determined from the known density at 15C, the liquid volume, and the
volume and molecular weight of the gas. The density versus pressure data
where NSTEP is the total number of flash stages and V gas,n is the volume of may be presented as shown in Figure 3-11.
gas (Sm3) liberated at flash stage n. These data can be represented as shown Deviation from ideal gas behavior may be expressed via the compressibil-
in Figure 3-10. ity factor, Z, as defined in Equation 3-5. Employing Equation 3-5 twice,
48 Properties of Oils and "fatural Gases Oil and Gas Property Measurements 49

Figure 3-10. The solution GAS( 1) GAS(2) GAS (ST)

gas/oil ratio as function of
1-

Ps pressures. 1st STAGE 2nd STAGE

o, 3rd STAGE
p l'
T1 P2,T2
cc
V1 Pst' \t

OIL to stock tank

Figure 3-12. Three-stage flash separation train.

Pressure The two first stages are accomplished in the same PVT cell as was used in
the differential liberation experiment, and the last stage is carried out by
Figure 3-11. Oensity 01 reser-
p transferring the second stage Huid to a single flash apparatus (see "Flash
voir fluid as function 01 pres-
sure.
Separation and Compositional Analysis," discussed earlier). From this ex-
periment the overall COR, COR at individual stages, compositions of the
gas and liquid phases, properties of the stock tank Huid such as density and
molecular weight (see Chapter 2), and also the Bo factor of the oil can be
obtained. Tables 3-1 and 3-2 show a typicallaboratory result of a separator
test experimento

Gas Reinjection Related Analysis

Ps Pressure
In arder to understand the reservoir Huid behavior under gas injection
).~
processes, two different analyses have been designed:
i.e., at two different pressures and temperatures and for the same number of
moles, will result in: l. Swelling test
2. Slim tube minimum miscibility pressure determination
Z PVT2Z2/P2V2T\ (3-9)
=
.. Swelling Test. When gas is injected into a reservoir containing an under-
saturated ol, the gas can go into solution. This has the effect of swel!ing the
Equation 3-9 relates a compressibility factor at condition 1, where it is
oil, i.e., the volume of oil becomes larger.
unknown, to the compressibility factor at condition 2, where it is assumed to
Simulation of this effect can be performed in an ordinary PVT cell, start-
be known. Z2 is normal!y the compressibility factor at ambient conditions,
ing with the original reservoir Huid in question. Injection gas has been com-
and it is usually clase to unity.
pressed into a separate steel bottle.
A srnall, known volume of injection gas is transferred into the PVT cell. A
Process Simulation by Flash Separation
new saturation pressure is determined using the techniques already de-
scribed, and a new saturation volume is recorded. This process is repeated
A three-stage flash can be performed in arder to furnish support data for until the saturation pressure of the fluid is equal to the estimated injection
process simulation. pressure of the system.
Such a flash sequence is shown in Figure 3-12. The data from this test can be presented as shown in Figure 3-13.
Stage three is the stock tank stage, where the pressure is one atm and the
temperature is 15C. For the other two stages, temperature and pressure can Slim Tube Tests. One method of improving oil recovery involves injection of
be chosen as desired. gas into the oil reservoir. The gas may be nitrogen, carbon dioxde, Hue gas,
 Oil and Gas Property Measurements 51
50 Properties 01 Oils and Natural Gases

Table 3-2
Analyses of Separator Gases from Separator Tests
-
"'O(!ll:
'"
ea
!5 .~i :
io I:
o Component Stage 1 Stage 2 Stage 3
Q)> J:
~
<el [O ~ I~
(/)a. ea '"
. ea o o o Hydrogen sulfide O O O
(!IiL
Carbon dioxide 1.56 2.05 2.60

Nitrogen 1.65 0.55 0.23

..o Q,l .
Methane 88.49 85.44 61.49
iE1 o
. j
ea _
..,
~ Ltll"<t
~ N~ ~ <Xl o
g.
(/)
e af Cl
Ethane 5.79 7.72 17.38

Propane 1.01 1.95 7.70

iso-Butane 0.35 0.57 2.75

e
o el) M n-Butane 0.47 0.66 3.47
'':; E"L o
ea j o CD
E-o ~ iso-Pentane 0.23 0.25 1.44
.. >0 ea
af LL
n-Pentane 0.16 .0.30 0.94
(5
..11: Hexanes 0.12 0.15 0.82
. .>< .><
U
ea -"'e ea .. ~
:o
e
'"
(.1
O Heptanes plus 0.17 0.36 1.18
al N tU ~ i
~ >0 N ID . (.1 "O
ea ~ o co n.O.9 (") E 100.00 100.00 100.00
~ -"''''Ltl
.., e ... co

";"
M!::
-o
U)
o
000
el) o ~~
"O
e -
1
M
E
~....::.:.a.
~ Q;
::
Q)
a.

_~
_CII
.c
::J
U)

m~~ ~~ Pi - Injection pressure

-
CII
~a:
U)
5EO
_
en;
ea ea
"!I'
Ltl
o
~
" "<t
N
<el
~ ci ~ TI ~
uS!"'"
Q)rJlu
!'
"'O
Ps - Saturation pressure

~ (!la: cu - e
ea e
m Vs - Saturation volume

-o
(.1 O

I I I I I I~ME~ 5 Vi - LiQuid volume at injection

e~a tU ~~ ~ pressure
ea
Q. 5C:8 C u:rICD Ol . ~ m ~ [
CII
CJ)
1
ea ea
(!la:
Oi
ci
N ~ oleD
ME~=
-e-- o
Ci>~~
a. ~..
* ~
"O

u
:o
.!:
e
en
&.
E
A V - Swelling volume

I ~~2 ~ 8
S~ ~~ ~ g
~Q)~o~~~ ~~o 'O-e
[~~o
Q,lQ)
;; ~ ~ ~ ~,...~
"'O~!' ~
~ s
e
QJ
E
:::l
(/)F ~rJ~ ~ ~ o
Cl o (.1 s: >
~o:g :g .E
..o ..
Q)
(f)
~ t~
C/) ID....
.~~*
... >-~
~a."'1II
~ :a Ol
co
"<t
N
": T"""

~ I el tl
tUEO
-.~ ea
o
t5
ea
-;jctS
el::
cu ctI
Q) f! CD N M
O,D-
~ ID
-
Q) "'O
...
o
(/)a. EUE Eci~-;;;
c: .S.2 .2 ~ U tU
I III I l;o~~~~C> Ps p Pressure
el)
'~~~oca~~
Cl
ea
~o~~~o8g
00
N
'" Ci5CEi~x(/)';
cu cu 0- Q) c: ea (1)
Figure 3-13. Graphical representation of swelling test data.
o o ce e s o s
! ISfiff ~ ::
52 Properties of Oils and Natural Gases Oil and Gas Property Measurements 53

or natural gas. It is advantageous to have a miscible displacement. This system is set, and gas is allowed to displace oil through the sand pack. The
means that at the gas/oil front only one phase is achieved no matter in what amount of produced fluid is monitored as time proceeds.
proportion the oil and the gas are mixed. The injected gas and the reservoir The sequence is repeated for several pressures, and the recovery, (pro-
oil may be miscible by first contact or miscibility may be achieved as a result duced oil)/(initial oil), is recorded for each pressure. The resulting data may
of a multiple contact process. If the gas and oil become miscible, complete be plotted as shown in Figure 3-16.
displacement of the reservoir fluid will be accomplished, and the recovery
may be larger than 90 %. It is, therefore, of great practical and economical GAS
interest to find out if the gas plus the oil form one phase, i.e., are miscible. If
so, it becomes important to know what the lowest pressure at the reservoir f
temperature is where complete miscibility occurs in the displacement pro-
cess. This is called the minimum miscibility pressure and it is determined in
a slim tube test.
The minimum miscibility pressure may be determined from visual obser-
vations of the displaced fluido Usually it is determined as the breakover
point in the recovery curve from a series of displacement experiments. An
example of a displacement curve is illustrated in Figure 3-14, showing oil
(])
o
recovery as a function of pressure at a fixed temperature. The concept of
minimum miscibility pressure is further discussed in Chapter 16. 1,

A schematic diagram of a slim tu be apparatus is shown in Figure 3-15. 1. - Mercury pump

The main part of the equipment is a piece of l(dn. steel tubing, approxi- 2. - Oil reservoir
mately 10 m long and packed with sand. Two containers, one for the injec- 3. - Gas reservoir
tion gas and one for the reservoir oil, are connected to mercury pumps. Us- 4. - Sllm tube - steel tublng 1/4" approx. 10 m
ing the mercury pumps, the reservoir fluid and the gas may be forced to 5. - Slght glass - visual flow cell
flow through the steel pipe. A sight glass to observe the produced Huids and 6. - Back pressure valve
a flash separator to determine the amount of fluids produced are provided. 7. - 011 and gas separator
The sand pack is first saturated with the oil sample to be studied, and the Figure 3-15. Schematic diagram of slim tube apparatus.
initial volume of fluid in the coiled tubing is recorded. The pressure of the

Breakover
POin\
100 MMP .... :.::::::::::::::::::::::::::::::/'

Figure 3-16. Pressure as a

//
90 Figure 3-14. Example of oil
recovery as a function of function of recovery from a
>-
Q; pressure in a slim tube experi- slim tu be analysis.
> 80 P
o ment.
(J

I
Q)
a::
;f.
70 o Miscible

60

1
O Irnrniscible
//
100 200 300 50% 100%
Test Pressure (bar) Recovery
54 Properties oj Oils and Natural Gases
Oil and Gas Property Measurements 55

The pressure at which 90-100% of the initial fluid in the sand pack is
recovered is defined as the minimum miscibility pressure of the system.
Several modifications of this apparatus can be made. Online density de-
termination of the fluids using a high-pressure densitometer, and viscosity
measurements using a capillary tube viscosimeter, are examples of such
modifications.
'1~"
J
.~
.:

GAS CONDENSATE EQUIPMENT Figure 3-18. Sloane type PVT

cell in its thermostatic enclosure.

At reservoir conditions, gas condensates usually exist in one single gaseous

phase. Therefore, gas condensates must be studied using types of PVT cells
different from those used for black oils.
There are two main types of gas condensate PVT cells: the long win-
dowed cell (or three windowed cell), and the Sloane type of cell. Figures
3-17 to 3-20 show the different features of the cells.
The windows in the cells are located so that visual observation of the
whole or the bottom part of the hydrocarbon sample can take place.
The dew point is defined as the pressure at which an infinitesimally small
amount of Iiquid is in equilibrium with a gas phase. At constant tempera-
ture, this can be observed as the pressure at which a small amount of liquid 1 Orilice lor introduc-
/ tion 01 mercury
appears in the system. For gas condensates, normally the upper dew point is through the upper
cap, lor operation
01 the pisto n

2 Piston with joist

Figure 3-19. A cutaway diagram packing
of a Sloane type gas condensate
cell.

Hola and val ve tor

introduction 01 gas
Hole and valve lor
oline
4 introduction and
sampling 01 gas
Spiral lor stirring
gas. The liquid is
stirred by a spiral 5
6 Sapphire viewing
coil wound around
window
the axis
~
Lighting window 8 7 Viewing window

Liquid voluma 9
measuremant
chamber 1 O Magnetic agitator
Figure 3-17. A Ruska three-windowed gas condensate/volatile oil cell in its ther-
mostatic enclosure.
56 Propertes of Oils and Natural Gases Oil and Gas Property Measurements 57

Below the dew point, the liquid volume is determined at each pressure.
These readings result in the liquid dropout curve. Data from the constant
mass study can be plotted as shown in Figure 3-22. It should be noted that
the dew point is obtained by extrapolating the liquid dropout curve.
The amount of liquid formed is usually relatively small, and therefore
special attention must be paid to the liquid volume determinations. The
available cell designs oHer diHerent methods by which the liquid volumes
can be measured. Best results are obtained in cells where charge volumes are
up to 400 cm'' at the dew point. Significant liquid volumes will then be
formed, allowing relatively accurate liquid volume measurements. The
amount of liquid is best determined in a cell where the window configura-
tion allows light to shine through the system.
For the previous reasons, a Sloane type cell is recommended. It can be
used to study very lean gas condensate systems.
However, a Sloane type cell is not recomrnended for studies of volatile or
near critical fluids.

Figure 3-20. Pump and valve arrangement for ~Ioane type cel!.
INCIPIENT FORMATION
, OF usuiu
y I

of interest, i.e., the liquid is formed ("drops out") by pressure reduction.

I I
.... .... .

. . .
The separator samples, gas and liquid, are prepared as for black oil analy- ....

1, ,,,'

siso The PVT cell is charged, first with the gas, by allowing the gas to ex- .... ....
pand into the cell and displace the mercury. The volume of mercury dis- .... . . . .

placed can be measured, and thus the volume of gas in the cell is known. l vt,
Gas
v
Ga s
..
'"'"
Gas I l. Gas
.... t2 . . . . >
.... .... " ',' ...
. . . . . . . . .. ....
1
M

....

1
....
1 ..... . . . .
I v

f
>
Constant Mass Expansion .... . . . .
ts
....
In order to achieve a representative reservoir fluid, separator liquid is
added to the cell in the gas/oil ratio measured during the sampling. Once
this is done, the system in the PVT cell is at reservoir temperature, corn-
pressed to above the recorded reservoir pressure. At this pressure, the cell is
left (usually overnight) for equilibration. The system is checked to see if it is
all in single phase, and the total system volume is then recorded. A constant
mass expansion can start. Figure 3-21 shows a schematic diagram of this
process.
The pressure is reduced in a stepwise manner. The system is observed visu-
ally, and the volume recorded at each pressure level by reading the mercury p,' Psat
P2 > Psat P3 = Psat P4 < Psat PS<P4<Psat
pump. "
The dew point pressure can be somewhat difficult to observe, and there- Figure 3-21. Schematic representation of constant mass expansion experiment
fore the stepwise pressure reduction is continued below the dew point. on a gas condensate.
 58 Properties o] Oils and Natural Gases Oil and Gas Property Measurements 59

Constant Volume Depletion

TR:::. ~>"..
l..
:. :::: r:"
G."II
In order to simulate production behavior of the reservoir condensate
fluid, a constant volume depletion experiment is performed. A schematic di-
agram of the process is shown in Figure 3-23.
~: G"
>"" : : : :
.
v G".
>'
',1 G"
11
>'"'"
_ . Gas. >'" . .. > A", 1: Gas:
The system is brought to just below its dew point. The dew-point volume >~ .... ' V > ...
of the system is known from the previous constant mass expansion study.
A .:: ..
When the system is equilibrated at pressure P, the excess volume above the 1
....

1
dew-point volume is displaced into a flash separation system. Such a system
is shown in Figure 3-24.
The flash system must be designed in such a way that it is possible to col- I Gil

lect and measure the small amounts of liquid that are condensed out of the

P, - Psat P2 < Psat P2 < Psat P2 < Psat P3<P2<Psat

Relative volume
Q)
E \ -,-,- Const ant mas s
Figure 3-23. Schematic representation 01constant volume depletion experiment
_0_0_ o Q)
11:
"o E on gas condensate.
> "
o
)11
;)1

e, >
o
4-
.:
.e. Figure 3-24. Left side:
o JIi\ Flash separation system
Q)

~ E
lor gas condensate mix-
+'
o"
CL
o
<-
-o
-o

"cr
--'
""
~\ 'O"
-
>

'O
>
Q)
>
+-'
<O

Qj
o::
tures. Right side: Gaso-
meter tor gas volume
measurements.

4 r ~~

'~'-- -I,
.8

I ! I l:a ' -te I

100 200 300 f 400 r 500
DP 361.2 Pressure,[barl
Reservoir Pressure

Figure 3-22. P-V and liquid dropout curves ter a recombined sample at 121 oc.~
;;1.,
60 Properties 01 Oils and Natural Gases Gil and Gas Property Measurements 61

produced gas stream. The total gas volume can be measured using a gasom- Table 3-3
eter. Constant Volume Depletion Data at 121C Hydrocarbon Analysis of
Produced Wellstream
From compositional analyses of the gas and the liquid phases, and by
knowing the volume of each phase, a produced well stream composition can
Component Reservoir Pressure (bar)
be determined.
362.2 307.7 274.4 222.4 171.5 120.2 71.8
The pressure is further reduced, and more liquid is formed. The amount
of liquid in this phase is measured, and the excess volume above the dew- Carbon dioxide 9.16 9.14 9.29 9.25 9.39 9.31 9.20

point volume is displaced out of the cell at constant pressure. The amount of Nitrogen 0.64 0.72 0.74 0.69 0.74 0.63 0.64
produced gas is again recorded. Methane 68.80 70.05 70.27 70.68 71.30 71.42 69.12
This stepwise pracedure is repeated until atmospheric pressure is reached.
Ethane 8.43 8.54 8.59 8.44 8.34 8.51 8.78
At the end of the analyses a molar balance is performed to check if some
fluid has been lost during the experimento The balance may be written as: Propane 5.11 5.13 5.18 5.06 4.92 5.02 5.70

iso-Butane 0.81 0.80 0.81 0.79 0.75 0.76 0.97

No moles charged to the cell = No moles produced from the cell. n-Butane 1.45 1.42 1.45 1.39 1.33 1.35 1.79

iso-Pentane 0.52 0.49 0.50 0.49 0.46 0.46 0.68

The agreement should be within 1 % in order to avoid errors in the subse-
quent calculations from the anaIysis. n-Pentane 0.53 0.50 0.51 0.49 0.46 0.46 0.72

TypicaI data from a constant voIume depletion can be seen in TabIe 3-3. Hexanes 0.63 0.54 0.55 0.55 0.51 0.53 0.80
Figure 3-25 shows the difference in the two liquid. dropout curves from Heptanes 0.83 0.62 0.64 0.66 0.62 0.68 0.81
constant mass expansion and constant volume depIetion, respectiveIy.
Octanes 0.95 0.55 0.52 0.59 0.59 0.59 0.54

Nonanes 0.52 0.27 0.21 0.26 0.24 0.19 0.16

Decanes + 1.62 1.23 0.74 0.66 0.35 0.09 0.09

VOLATILE FLUm EQUIPMENT L 100.00 100.00 100.00 100.00 100.00 100.00 100.00

Mol. wt. 01 C1O+ 200 180 168 157 151 157 147

Some of the naturally occuring hydrocarbon mixtures can be classified as Density 01 C1O+ 0.856 0.828 0.827 0.821 0.819 0.817 0.817
voIatile fluids (see Chapter 1) or near-critical fluids. Gas phase compressi-
The PVT behavior of these systems is difficuIt to analyze, because the gas bility lactor, Z 1.003 0.946 0.922 0.882 0.882 0.903 0.934
phase and liquid phase are, at reservoir conditions, virtualIy impossible to Mole % produced O 9.6 7.6 13.4 14.8 16.3 15.5
distinguish. 45.4
Cum. mole % produced O 9.6 17.2 30.6 61.7 77.2
These fluids are normally anaIyzed as for gas condensates (see "Gas Con-
Molecular weight 28.31 26.47 25.60 25.42 24.81 24.49 25.44
densate Equpment").
A long windowed cell, as shown in Figure 3-26, is often used. This cell Two-phase Z-Iactor 1.003 0.943 0.918 0.887 0.870 0.869 0.872
allows viewing of the whoIe sample at any stage of the anaIysis. Viscosity (cP) 0.0332 0.0271 0.0246 0.0215 0.0187 0.0164 0.0147
The cell is charged, and the pressure in the cell is raised by means of a
mercury pump. The sample is Ieft at reservoir conditions to equilibrate.
The pressure is then reduced in the cell, and observations are made to de-
termine the saturation pressure. In the case of a volatile fluid it may be diffi- changes. Figure 3-27 shows the typicaI, Iarge, liquid dropout curve changes
cuIt to see whether a phase transition is a dew or a bubble point, beca use the over a temperature range of only 20C.
two phases are nearly identical. Once the saturation pressure of the system has been found, and the con-
Phase changes are rapid, and a near criticaI system exhibits Iarge changes stant mass study has been performed, the work program of the anaIysis fol-
in its PVT and phase behavior for relatively small temperature and pressure lows that of a gas condensa te system.
 62 Properties o] Oils and Natural Gases Oil and Gas Property Measurements 63

100
o -- o Constant Mass ~
~~
Constant Vol ume ~
~
~M
~
..
o
~
:ro
~
8.
e
5% ..
o

:~
!::v.----~
-o
~
-o
cr

:~ O_O_O_O--a-O"tb
.-e_~===-t:;:::::e~A-tr-~tB
e-e
~
~
ew ~
~
~
w
Constant mass expansion at 150 'C
2~ I ~ Constant mass expansion at 140 "C

Constant mass expans!on at 130 'C

~
1%
10

I I I I ~.

100 200 300 t 400 t 500 w M 120 160 200 240 280 320 360
D.P. Res. pressure P [bar)
373.5
Pressure. [bar) Figure 3-27. Liquid dropout curves tor a volatile fluid. Note that a dew point is
obtained at 130C, whereas bubble points are obtained at 140 and 150C.
Figure 3-25. Liquid dropout curves tor a gas condensate mixture trom a constant
mass expansion and a constant volume depletion experiment.

MEASUREMENT OF VISCOSITY

An important property of a hydrocarbon system is its viscosity at relevant

pressures and temperatures.
A few methods have been developed for these measurements at reservoir
pressure and temperature. The most commonly used methods are:
l. Rolling ball viscosimeter
2. Capillary tube viscosimeter

Rolling Ball Viscosimeter

Figure 3-28 shows a schematic diagram of a rolling ball viscosimeter, and

Figure 3-26. Long-windowed cell used tor volatile tluids, coupled to a capillary
tube viscosimeter (measurement ot viscosity section).
Figure 3-29 shows a photograph of a typical unit. It consists of the fall tube,
an expansion chamber, and a steel ball of known dimensions.
A hydrocarbon Huid is transferred into the evacuated viscosimeter at con-
stant pressure. The system must be equilibrated at above reservoir pressure
for several hours before the analysis can start.
64 Properties 01 Oils and Natural Gases Gil and Gas Property Measurements 65

Figure 3-28. Schematic dia- Using a switch, the magnet is deactivated, and the ball starts to fal!. The
f1
SOLENOID gram of a rolling ball visco- v- time it takes to reach the bottom of the tu be is recorded.
~OUTlET simeter. With this time and the known density of the fluid, its viscosity can be
BARREL SEAL
determined Irom the calibration curves.
GAS CHAMBER
This method is generally found to be reliable; it is, however, rather cum-
bersome to operate.

Capillary Tube Viscosimeter

INLET

With the capillary tube viscosimeter, the viscosity of flowing fluid is cal-
culated using the flow equation:

r = D2 LlP/32 Lv (3-10)

where D = diameter
L = length of the capillary tu be
LlP = pressure drop
v = velocity of the fluid flowing through the capillary tu be

The equipment used is shown schematically in Figure 3-30. In addition, a

typical capillary tu be viscosimeter is shown in Figure 3-26.
Reservoir fluid is charged to the adjoining PVT cell and equilibrated at
reservoir conditions or above.

Figure 3-29. Photograph of a rolling ball viscosimeter.

Prior to charging, the viscosimeter must be calibrated with known cali-

bration fluids. The ball size should be chosen so that the calibration curve is
linear in the range of operations, normally 0.2-3 cP. Calibrations are made G)
at each of the three angular positions in which the viscosimeter can be
placed.
To start the viscosity determination, the top valve is closed in order to seal 1. Mercury pump 4. OP transducer
the upper end of the ball tube. The apparatus is turned upside down, allow- 2. PVT ce ll, vi sual 5. Back pressure valve
ing the steel ball to roll to the topo The valve stem of the top valve is made 3. Coi led capi llary tube 6. Dil and gas separator
magnetic by a coil, so that the ball will stay at the top when the apparatus is
turned back to measurement position. Figure 3-30. Schematic diagram of a capillary tube visccsirneter.
66 Properties o] Oils and Natural Gases Oil and Gas Property Measurements 67

To measure viscosity, the fluid is charged to the coiled capillary tube, and
equilibrated. Flow of fluid through the capillary is established using a mo-
torized pump. The pressure drop across the coil is monitored at constant
flow rateo The temperature and pressure are kept constant.
The process can be repeated at different pressures above the saturation
pressure, and the viscosity calculated from Equation 3-10.
Since the coiled capillary tube does not have a perfectly homogenous in-
ner-diarneter, it must be calibrated with calibration fluids of known vscos-
ity.

SURFACE TENSION

The surface tension between two fluids, as defined in Chapter 13, is an

important fluid property in the evaluation of reservoir Huid flow. Surface
tension exists between a water phase and an oil phase, between a water
phase and a gas phase, and between an oil phase and a gas phase. The sur-
Figure 3-31. Pendant drop
face tension can, with some difficulty, be measured at reservoir conditions
apparatus.
using either a pendant drop or a spinning drop apparatus. Both methods are
optical, in the sense that the measurements are made on images of bubbles of
one fluid suspended in the other.

Pendant Drop Tensiometer

Figure 3-32. Pendant
drop.
The basic apparatus, consisting of a high-pressure sight glass or PVT-cell,
is shown in Figure 3-31. This can be charged with brine at reservoir condi-
tions. Through a capillary tube extending into the cell, an oil droplet is seen
suspended frorn the tip of the capillary tube. The shape of the droplet (see
Figure 3-32) will depend on the interfacial tension between the fluds, and
the interfacial tension can be calculated from the droplet's geornetry (Ci-
rault et al., 1984 and Adamson, 1976) using the following expression: 1\1
!
~
a = /::;.pgDelH (3-11)

where /::;.p = difference in the density of the two fluids

g = gravitational acceleration
De = equatorial diameter (see Figure 3-32)
D, = distance shown in Figure 3-32
lIH = tabulated function of D,lDe (Adamson, 1976).
,~
{{
 ~.
"~,/

68 Properties of Oils and Natural Gases

Spinning Drop Tensiometer

This method can only be used for liquid-liquid measurements. A glass

tu be is filled with formation brine and an oil droplet is introduced into the
tube. The tube is spun at a high speed. The oil bubble can be viewed using a
strobe light. The shape of the bubble will depend on the interfacial tension Chapter 4
between the fluids. The interfacial tension is calculated as described by
Chatenay et al. (1982).
Composition and Property Data

REFERENCES
Adamson, A. W., "Physical Chemistry of Surfaces," John Wiley and Sons,
lnc., 1976.
Chatenay, D. et al., "Measurernent of Low lnterfacial Tension, Comparison
Between a Light Scattering Technique and the Spinning Drop Tech-
nique," I. Dspersion Science and Technology, 3, 1982, pp. 245-260.
Girault, H. H. J., Schiffrin, D. J., and Smith, D. D. v., "The Measurement
of lnterfacial Tension of Pendant Drops Using a Video Image Profile Dig-
Ten petroleum reservoir fluid compositions have been selected to test the
flash calculation procedures of Michelsen described in Chapter 6 coupled
with the characterization procedure of Pedersen et al. described in Chapter
7. The compositions-comprse five gas condensate mixtures, one volatile oil
and four black oils (see Chapter 1 for a description of these kinds of mix-
tures). Six of the mixtures are from the North Sea, two from the U.S., and
two from the Middle East. A survey of the mixtures is given in Table 4-1,
and the mixture compositions are shown in Tables 4-2 through 4-12. The
composition data for mixtures one through four comprise both weight and ,

itizer," [ournal of Colloid and Interface Science, 101, 1984, pp. 257-266. molar compositions. For the remaining mixtures only molar compositions i I

are given. 11
Si~ulation results for the mixtures are shown in Chapters 8 and 9. I

Table 4-1
Survey of the Mixtures tor which Molar Compositions Are Given
t
Mixture Composition
I
No. Type of Mixture Origin in Table

1 Gas Condensate North Sea 4-2

2 Black Oil North Sea 4-3

3 Black Oil North Sea 4-4

4 Volatile Oil North Sea 4-5

5 Gas Condensate North Sea 4-6

6 Black Oil North Sea 4-7 and 4-8

7 Black Oil Texas 4-9

8 Gas Condensate Texas 4-10

9 Gas Condensate Iran 4-11

10 Gas Condensate Bahrain 4-12

''0
70 Properties oj Oils and Natural Gases Composition and Property Data 71

Table 4-2 Table 4-3

Composition of Mixture 1 (North Sea Gas Condensate) Composition of Mixture 2 (North Sea Black 011)

Molecular Density (g/cm3)

Molecular Density (g/cm3)
Component Weight % Component Weight % Mole % Weight at 15C, 1 atm
Mole % Weight at 15C, 1 atm
N2 0.571 N2 0.145 0.56
0.60
CO2 1.450 3.55
CO2 5.031 3.34

C, 40.667 C, 6.757 45.34

74.16
C2 1.531 5.48
C2 8.126 7.90

C3 C3 1.516 3.70
6.254 4.15
ic, 1.401 0.71
ic, 0.378 0.70

n-C. 1.44 n-C. 0.891 1.65

2.855
ic, 1.306 0.53
ic, 0.489 0.73

n-C, 1.637 n-Cs 0.580 0.87

0.66
C6 C6 1.043 1.33
2.355 0.81
C, C, 2.276 2.73 89.9 0.757
3.749 1.20 91.2 0.746
Ca . Ca 3.125 3.26 103.2 0.777
4.100 1.15 104.0 0.770
C. 2.342 2.14 117.7 0.796
C. 2.577 0.63 119.0 0.788
ClO ClO 2.379 1.94 133.0 0.796
2.329 0.50 133.0 0.795
Cl1 1.466 Cl1 2.205 1.62 147.0 0.800
0.29 144.0 0.790
C'2 1.458 C'2 2.179 1.47 160.0 0.815
0.27 155.0 0.802
C'3 2.693 1.69 172.0 0.833
C'3 1.624 0.28 168.0 0.814
C,. 1.413 0.22 181.0 . 0.824
C, 2.789 1.62 186.0 0.843

c., C,S 2.937 1.59 200.0 0.849

1.165 0.17 195.0 0.833
C'6 2.553 1.30 213.0 0.858
C'6 1.057 0.15 204.0 0.836
C17 2.388 1.11 233.0 0.851
C17 1.096 0.14 224.0 0.837
c., 0.729 0.09 234.0 0.839
c., 2.885 1.26 247.0 0.856

C,. 1.137 C, 2.571 1.07 258.0 0.868

0.13 248.0 0.844
0.47 C20+ 51.898 13.32 421.0 0.914
C20+ 5.896 362.0 0.877
72 Properties of Oils and Natural Gases Composition and Property Data 73

Table 4-4 Table 4-5

Composition of Mixture 3 (North Sea Black 011)
Continued

Molecular Density (g/cm3)

Molecular Density (g/cm3)
Component Weight % Mole % Weight at 15C, 1 atm
Component Weight % Mole % Weight at 15C, 1 atm
N2 0.091 0.403 f:
~ e7 3.489 2.38 91.9 0.742
CO2 0.353 1.000 ~ ea 4.331 2.59 104.7 0.765
C, 5.845 45.396
Cg 3.329 1.75 119.2 0.788
C2 1.014 4.202 ~,t ClO 3.173 1.50 131.0 0.791
C3 0.314 0.887
3.666 1.55 147.0 0.796
C"
te, 0.262 0.561
C'2 2.408 0.93 161.0 0.811
-q
c, 0.242 0.518
1.13 171.0 0.826
C'3 3.125
ic, 0.375 0.647
2.952 1.01 182.0 0.836
~~" C"
n-Cs 0.170 0.294
c., 2.521 0.80 195.0 0.843
Ca 0.678 1.011
c., 2.878 0.86 208.0 0.848
C7 2.118 2.878 91.68 0.745
C17 2.211 0.60 228.0 0.844
Ca 3.467 4.114 104.99 0.767
c., 2.701 0.68 247.0 0.848
Cg 3.135 3.322 117.60 0.798
c., 2.184 0.54 252.0 0.859
ClO+ 81.936 34.767 293.53 0.902
C20+ 28.773 4.34 411.0 0.903

Table 4-5 Table 4-6

Composition of Mixture 4 (North Sea Volatile 011) Composition of Mixture 5 (North Sea Gas Condensate)

Molecular Density (g/cm3) Molecular Density (g/cm3) PNA Distribution Mole %

Component Weight % Mole % Weight at 15C, 1 atm Component Mole % Weight at 15C, 1 atm P N A
N2 0.258 0.58 N2 0.12
CO2 2.297 3.27 CO2 2.49
e, 13.780 53.89 C, 76.43
C2 4.108 8.57 C2 7.46
C3 4.254 6.05 e3 3.12
-c, 0.969 1.05 ic, 0.59
n-C, 2.263 2.44
n-C, 1.21
ic, 1.013 0.88 ic, 0.50
n-es 1.348 1.17 n-Cs 0.59
Cs 1.970 1.45
Cs 0.79
(table continued on next page) (table continued 011 next page)
 ,,-
1<v,
(:1/
. -1:: I
74 Properties oj Oils and Natural Gases Composition and Propertq Data 75" I
') (
'--1

Table 4-6 Table 4-7 "

Continued Continued

PNA Distribution Mol.e % Molecular Density (g/cm3)

Molecular Density (g/cm')
Component Mole % Weight at 15C, 1 atm
Component Mole % Weight at 15C, 1 atm P N A
C, 3.33 99 0.7395
C, 0.95 95 0.726 0.564 0.361 0.076
"1
Cs 4.06 106 0.7518
Cs 10~ 106 0.747 0.113 0.611 0.277
C9 2.76 120 0.7756
C9 0.78 116 0.769 0.483 0.311 0.206
ClO 1.33 139 0.7930
C'O 0.592 133 0.781 0.530 0.275 0.195
C11 1.79 146 0.7902
C11 0.467 152 0.778 0.681 0.193 0.'26
C'2 1.70 160 0.8060
C'2 0.345 164 0.785 0.757 0.123 0.120
C'3 1.81 174 0.8203
C'3 0.375 179 0.802 0.709 0.183 0.108
C, 1.46 194 0.8311
C, 0.304 193 0.815 0.635 0.209 0.156
C'5 1.49 205 0.8446
C,S 0.237 209 0.817 0.729 0.168 0.103
C'6 1.08 218 0.8515
C'6 0.208 218 0.824 0.624 0.232 0.H4
C'7 1.13 234 0.8542
C'7 0.220 239 0.825 0.668 0.185 0.147
C,S 0.99 248 0.8561
C,S 0.169 250 0.831 0.675 0.192 0.133
C'9 0.88 265 0.8663
C'9 0.140 264 0.841 0.652 0.190 0.158
C20+ 7.64 465 0.9350
C20+ 0.833 377 0.873 0.519 0.320 0.161 .
Source: Pedersen et al. (1985).
Source: Pedersen et al. (1985).

Table 4-7
Composition of Mixture 6 (North Sea Bla<:kOil) Table 4-8
TBP Data for Mixture 6
Molecular Density (g/cm3)
Component Mole % Weight at 15C, 1 atm Temp, Cum. Wt % Temp, Cum. Wt % Temp, Cum. Wt %
C Distilled C Distilled C Distilled
N2 0.34
100 10.54 218 28.43 288 43.10
CO2 0.84 127 14.54 237 32.48 304 46.52
C, 49.23 152 19.17 255 36.13 318 49.67
176 21.55 272 40.06 332 52.67
C2 6.32 197 24.92
C3 4.46 Densily 01 TBP sample al 15C, 1 atrn = 0.8499 (g/cm3)
Source: Pedersen et al. (1985).
ic, 0.86

n-C, 2.18

i-C5 0.93

c, 1.33 ,
C6 2.06

(table continued on next page)

76 Properiies of Oils and Natural Gases Compositiol1 and Propertu Data 77

Table 4-9 Table 4-11

Molar Composition of Mixture 7 (Black Oil from Southwest Texas) Molar Composition of Mixture 9 (Gas Condensate from Iran)

Molecular Density (g/cm') Molecular Density (g/cm')

Component Mole % Weight at 15C, 1 atm Component Mole % Weight at 15C, 1 atm

C, 52.00 ,N2 0.08

C2 3.81 CO2 2.44

C, 2.37 C, 82.10

ic, 0.76 .C2 5.78

n-C, 0.96 e, 2.87

ic, 0.69 -C, 0.56

n-Cs 0.51 -c, 1.23

C6 2.06 ic, 0.52

C7 2.63 99 0.749 e, 0.60

Cs 2.34 110 0.758 C6 0.72

Cg 2.35 121 0.779 C7+ 3.10 132 0.774

ClO+ 29.52 221 - 0.852 Source: Firoozabadi el al. (1978).

Source: Hoffman el al. (1953).

Table 4-12
Molar Composition of Mixture 10 (Gas Condensate from Bahrain)

Table 4-10 Molecular Density (g/cm')

Molar Composition of Mixture 8 (Gas Condensate from Southwest Texas) Component Mole % Weight at 15C, 1 atm

N2 11.71
Molecular Density (g/cm')
Component Mole % Weight at 15C, 1 atm CO2 6.50

91.35
1;12 5 0.05
C,
C2 4.03 C, 79.06

C3 1.53 C2 1.62

ic, 0.39 C3 0.35

n-C, 0.43 -C, 0.08

ic, 0.15 c, 0.10

e, 0.19 ic, 0.04

C6 0.39 n-es 0.04

C7 0.361 100 0.745 e6 0.06

Cs 0.285 114 0.753 C7 0.06 97 0.7243

Cg 0.222 128 0.773 Cs 0.05 111 0.7413

ClO+ 0.672 179 0.814 Cg 0.04 124 0.7611

Source: Hoffman el al. (1953). ClO+ 0.24 195 0.8238

Source: AI-M::.hrnnc: ::Inri Ti,,!:! (1 QCl7\
78 Properties o} Oils and Natural Gases-

REFERENCES

Al-Mahroos, F. M. and Tjoa, G. H., "Analysis of the Phase Behavior of

Khuff Gas System in Bahrain Field," SPE Paper 15766, 1987.
Firoozabadi, A., Hekim, Y, and Katz, D. L., "Reservor Depletion Calcula-
tions for Gas Condensates Using Extended Analyses in the Peng-Robinson
Equation of State," The Canadian lournal o] Chemcal Engneerng, 56,
1978, pp. 610-615. Chapter 5
Hoffmann, A. E., Crump, J. S., and Hocott, C. R., "Equilibrium Con-
stants for a Gas-Condensate System," Petroleum Transactons AIME,198, Equations of State
1953, pp. 1-10.
Pedersen, K. S., Thomassen, P, and Fredenslund, Aa., "Thermodynamics
of Petroleum Mixtures Containing Heavy Hydrocarbons. 3. Efficient
Flash Calculation Procedures Using the SRK Equation of State," Ind.
Eng. Chem. Process Des. Dev., 24, 1985, pp. 948-954. Equations of state (EOS) are, for pure substances, mathematical relations
between volume, pressure, and temperature. For mixtures, EOS in addition
include composition. EOS are ver y versatile tools for engineering applica-
tions. They can be used for all states of matter (mostly gas, vapor, and liq-
uid), and they can describe transitions between states.
The development of EOS for the representation of real systems started in
1873 with the publication of the van der Waals (vdW) EOS. Since then hun-
dreds of different EOS have been proposed. It is beyond the scope of this
book to review in detail all these developments. In fact, all of the oil/gas
pressure-volume-temperature and phase equilibria computations are based
on one equation of state. For broad reviews of EOS, the reader is referred to
Abbott (1979), Leland (1980) and Tsonopoulos and Heidman (1986). In ad-
dition, Tzouvaras (1986) gives a particularly clear and detailed account of
the different types of EOS.

TYPES OF EQUATIONS OF STATE

Leland (1980) distinguished between four families of EOS:

l. van der Waals family
2. Benedict-Webb-Rubin family
3. Reference-fluid equations
4. Augmented-rigid-body equations
The vdW family encompasses simple, mostly cubic EOS. Their main
characteristic is the separation between the repulsive and attractive effects.

,':Z)'

79
80 Properties oj Oils and Natural Gases Equations o] State 81

Despite their simplicity, these EOS display quantitatively correct perfor-

CUBIC EQUATIONS OF STATE
mance, even being able to describe multiphase equilibria, tricritical points,
and other complicated phenomena (Michelsen and Hedemann, 1988). Cu-
bic EOS are discussed further in the next section. Cubic EOS are explicit regarding pressure and may be written in terms of
The Benedict-Webb-Rubin (BWR) family includes complicated EOS and a repulsive and an attractive contribution to the pressure P:
are empirical extensions of the virial EOS. Besides the Benedict-Webb-Ru-
P = prep _ pattr
bin (1940) EOS itself, the most significant members of this family for oil/gas (5-1)
application are the Starling (1973) EOS, and the Lee and Kesler (1975)
EOS. Classical examples of cubic EOS are those of van der Waals (1873):
The reference-fluid equations aim at accurate representation of a large
amount of PVT data for pure substances. The equations contain many pa- P = RT/(V - b) - a/V2 (5-2)
rameters. An important application of reference-fluid equations is their use
as reference fluids in corresponding state theories, see e.g., MolIerup and and Redlich and Kwong (1940) (R-K):
Rowlinson (1974).
The augmented-rigid-body family combines description of repulsive P = RT/(V - b) -;- a/[T05 V(V + b)] (5-3)
forces between hard molecules of different shapes with expressions for the
molecular attractions. The sound theoretical basis of the repulsive term is where R = gas constant
the most important and promising characteristic of these EOS. The Cama- T = temperature
han and Starling (1972) express ion for the repulsion between hard spheres V = molar volume
has been repeatedly used. Approximations to th~ augmented-rigid-body
EOS include simplfcatons of the expression for the repulsive term, such as There are two parameters; b represents the "real," or hard-sphere volume of
in the Cubic-Chain-of-Rotators EOS (Kim et al., 1986). the molecules, and a represents the intermolecular attraction.
Today alI four families of EOS receive considerable attention. Recent lit- The parameters a and b in Equations5-2 and 5-3 are normalIy deter-
erature, however, mainly contains developments in EOS of the vdW family, mined by imposing the critical conditions:
because of their simplicity and practicality, and of the augrnented-rigid-
body EOS, because of their theoretical foundation and potential for accu- [(aPlaVh = (a2Plav2h = O]crit. point (5-4)
rate extensions.
The oil/gas PVT- and phase equilibrium computations in this book are The vdW and R-K EOS may be extended by altering their ternperature-
based on a cubic equation of the vdW family. The reasons for this choice dependence and by including parameters in addition to a and b. Jensen
are: (1987) defines two generalized cubic EOS, from which many of the com-
monly used equations may be derived:

1. Cubic EOS of the vdW family yield relatively simple expressions for
Type 1: The repulsive-terrn modified R-K EOS
the thermodynamic properties and phase equilibrium relationships of
interest.
P = [RT(2V + b(T) )]/[V(2V - b2(T))] - a(T)/[V(V + b3(T))] (5-5)
2. The EOS of the other three families, though more complicated in na-
ture, do not give quantitatively better descriptions of mixture phase Type 2: The attractive-term modified vdW EOS
transitions than do cubic EOS (Tsonopoulos and Heidman, 1986).
P = RT/[V - b(T)] - a(T)/[(V + b2(T))(V + b3(T))] (5-6)

It is welI known that the more complicated EOS yield better predictions Equations 5-5 and 5-6 each contain four temperature-dependent pararne-
of liquid densities than the cubic EOS. This disadvantage of the cubic EOS ters: a, b, b2, and b3
is taken into account. Some of the EOS that may be derived from Equations 5-5 and 5-6 are:
 ~
,~ !
82 Properties oj Oils and Natural Gases Equations o] State 83

Soave-Redlich-Kwong (SRK), (Soave, 1972) the critical point expressed in Equation 5-4. By inserting the expression for
the pressure given in Equation 5-7 into Equation 5-4 the following relations
P = RT/(V - b) - a(T)/[V(V + b)] (5-7) may be derived for the pure component a-parameter at the critical point
and for the pure component b-parameter:
Type 2: a = a(T), b, = b, b2 = O, b, = b
llc = 0.42747 . R2 . T2e/Pe (5-12)
The Peng-Robinson EOS (PR), (Peng and Robinson, 1976)
b, = 0.08664 . R . Te/Pci (5-13)
P = RT/(V - b) - a(T)/[V(V + b) + b(V - b)] (5-8)
Values for Te and P, of some common petroleum mixture constituents are
Type 2: a = a(T), b = b, b2 = (1 + 205)b, b, = (1 - 205)b given in Table 5-1. The temperature dependence of the a-parameter enter-

The Adachi-Lu-Sugie EOS (ALS), (Adachi et al., 1983) Table 5-1

Critical Temperature (Tc), Critical Pressure (p c) and Acentric Factor (w) of
P = RT/(V - b) - a(T)/[(V - b2)(V + b3)] (5-9) Some Common Petroleum Mixture Constituents

Peneloux's modified SRK EOS, (Peneloux et al., 1982) Te (Kelvin) Pe (atm) w

N2 126.2 33.5 0.040

P = RT/(V - b) - a(T)/[(V + c)(V + b + 2c)] (5-10) CO2 304.2 72.8 0.225

H2S 373.2 88.2 0.100

Type 2: a = a(T), b = b, b2 = e, b3 = b + 2c
C, 190.6 45.4 0.008

The Ishikawa-Chung-Lu EOS, (Ishikawa et al., 1980) C2 ' 305.4 48.2 0.098
,
C3 369.8 41.9 0.152
i P = [RT(2V + b(T) )]/[V(2V - b(T))]
te, 408.1 36.0 0.176
- a(T)/[T05V(V + b(T))] (5-11)
n-C. 425.2 :f7.5 0.193

Type 1: a = a(T), b, = b(T),b2 = b(T), b, = b(T) ic, 460.4 33.4 0.227

n-Cs 469.5 33.3 0.251

Many more examples exist (Jensen, 1987). The presence of parameters in 507.4
n-C6 29.3 0.296
addition to a and b as in the R-K and vdW EOS permits inclusion of con-
Source: Reid el al. (1977).
straints other than those given by Equation 5-4. Careful evaluations in-
formed the choice to base the oil/gas computations on the SRK EOS. No
other cubic EOS tested yielded better results for naturally occurring oil and ing Equation 5-7 is expressed in the form of a term ct(T) which multiplied
gas mixtures. The choice of the SRK EOS (and modifications thereof) as the with aci gives the final expression for the a-parameter of the SRK-equation:
basis for the oil/gas property and phase equilibrium calculations is justified
by the many results shown in subsequent chapters. a(T) = acct(T) (5-14)

The a-parameter is obtained from the following expression:

THE SRK-EQUATION OF STATE
ct(T) = (1 + m(l - T~i5))2 (5-15)
The analytical form of the SRK-equation of state is given in Equation 5-7.
where m = 0.480 + 1.574w - 0.176w2 (5-16)
The a and b parameters are found from the pressure-volume relationship at
84 Propertes of Oils and Natural Gases Equatons of State 85

It is seen from Equation 5-15 that O'(T) = 1 at the critical temperature,

- where z, and z are mole fractions of components i and j, respectively and:
where a therefore becomes equal to 8.c.In Equation 5-15 Wis the acentric
factor which is defined as follows (Pitzer, 1955): a = (aa)1I2( 1 - k) (5-20)

W= - IOglOPriap (at Tr = 0.7) - 1 (5-17) kij is a binary interaction coefficient, which is usually consdered equal to
zero for hydrocarbon-hydrocarbon interactions, and different from zero for
where Priap is the reduced vapor pressure (P/Pci) of component i. nteractions between a hydrocarbon and a non-hydrocarbon, and between
unlike pairs of non-hydrocarbons. Table 5-2 shows the non-zero binar y in-
teraction coefficients to be recommended for use with the SRK-equation
The acentric factor may be interpreted as the non-sphericity of molecules
(Reid et al., 1977) for some lighter petroleum mixture constituents.
of type i. Values for Wof a number of different components are given in
Equation 5-7 can be rewritten in terms of the compressibility factor, Z:
Table 5-1. The dependence of m on Wgiven in Equation 5-16 is found from
experimental vapor pressure data for aliphatic hydrocarbons ranging from Z3 - Z2 + (A - B + B2)Z - AB O
= (5-21)
CitO C 10 (Soave, 1972).
where Z is defined ;ts:
Table 5-2
Non-Zero Binary Interaction Coefficients for Use with the SRK-Equation Z PV/RT
= (5-22)
of State Between Some Common Petroleum Mixture Constituents

and A and B are given by the following expressions:

Component
Pair N,
-
CO, H,S
A = a(T)P/R2T2 (5-23)
N2 0.00 0.00 0.00

CO2 0.00 0.00 0.12 B = bP/RT (5-24)

H2S 0.00 0.12 0.00

C, 0.02 0.12 0.08 The polynomial of Equation 5-21 may have one or three real roots. When
for apure component or a single-phase mixture only one real root exists, this
C2 0.06 0.15 0.07
root equals the compressibility factor of the phase presento The compress-
C3 0.08 0.15 0.07 ibility factor of a dilute gas is clase to unity. The compressibility factor of a
ic, 0.08 0.15 0.06 near critical pure component is clase to 0.33. Using the SRK-equation of
n-C. 0.08 0.15 0.06 state the compressibility factor of apure component at the critical point will
.~ always be found equal to 0.33. In the case where the polynomial of Equa-
i-C5 0.08 0.15 0.06 ~:
tion 5-21 has three real roots, one or more phases may be found. The num-
n-C5 0.08 0.15 0.06 ber of phases, and the amounts and compositions of each phase may be de-
n-C6 0.08 0.15 0.05 termined as described in Chapter 6.
Source: Reid el al. (1977). In the case of two phases, each component will have equal fugacities, f,
in both phases:
For a mixture, a and b are found as follows:
~., fr = ff" (5-25)
a = E E zza
j
(5-18)
where V and L refer to the vapor and liquid phases, respectively. The crite-
rion of equal fugacities indicates equality in the driving forces in the transfer
b = E z.b, (5-19) of components from one phase to the other. The fugacity is related to the
chemical potential, jl, as follows:
86 Properties oj Oils and. NaturalCases Equations oj State 87

.t= .t?+ RT In f
where .t0 is a standard-state
5-25 can be rewritten:
cfJr y = cfJ!-x or K = cfJrNr
(5-26) pure component, or a mixture of given composition, may be calculated as
the sum of two contributions, namely the ideal gas enthalpy and the resid-
ual enthalpy:
chemical potential for component i. Equation
H Hd + H""
= (5-32)
(5-27)
and similarly for the entropy:

where Xand y = mole fractions of component i in the vapor and liquid s = Sd + Sres (5-33)
phases, respectively
cfJ= fugacity coefficient The ideal gas terms at temperature T are usually calculated from the follow-
ing equations:

K = (y/x) is called the equilibrium ratio. The following general thermody- T

namic relationship exists for determination of the fugacity coefficient of Hd = E ZHid = E Z J
i i Tref
qd dT (5-34)
component i in the mixture:

v Sd= E z(Sid- R In z.)

In cfJ= - l/RT J 00 ((ap/anh,v,nj - RT/V) dV - In Z (5-28)

= E Z (J
i
T
Ttef
(qd/T) dT - R In P/Pref - R In z) (5-35)
where n, is the number of moles of type i. When the SRK-equation is used
Equation 5-28 yields: .
where Tref is a reference temperature, here OC, and Pref is the reference
pressure, here 1 atm, and C~ the ideal gas heat capacity, which is usually
In cfJ= (b/b)(Z - 1) - In Z + In [V/(V - b)]
approximated by a third-degree polynomial in the temperature
+ a/bRT [b/b - 2 ~ Zj(l- kj)(aaj)05/alln [~V + b)/V] (5-29)
C~ = C + C2T + C3T2 + C4T3 (5-36)

Values for the coefficients C-C4 of the lighter petroleum mixture constit-
The phase compositions are related to the total composition as follows:
uents are given by Reid et al., (1977). A procedure for estimating C-C4 of
heavy hydrocarbons is presented in Chapter 7. The residual terrns of Equa-
x, = z/(l + (K - 1)/J) (5-30)
tions 5-32 and 5-33 may be derived from the SRK-equation using the follow-
ing general thermodynamic relations:
y = Kz/(l + (K - 1)13) (5-31)

where Z = mole fraction of component i in the total mixture

H'es = - RT2 a In cfJ/aT (5-37)
13 = the molar vapor phase fraction
where cfJis the fugacity coefficient of the mixture and the derivative is taken
at constant pressure and total composition. Also
The SRK-equation may also be used for calculation of other thermody-
namic properties, e.g., the enthalpy and the entropy. The enthalpy, H of a sres = Hres/T - R In cfJ (5-38)
88 Propertes 01 Oils and Natural Gases Equations 01 State 89

A summary of the equations used for computing the thermodynamic prop- PRASE DEN SITIES
erties from the SRK equation of state is shown in the box:

Summary of Equations Used for Calculating Fugacity Coefficients, Enthalpies and

The phase densities may be determined from the SRK-equation by using
Entropies from the SRK-EOS (Edmister and Lee, 1984) Equation 5-7 or Equation 5-21. In general, good results are obtained for
vapor phase densities, whereas the liquid phase densities are somewhat un-
SRK-EOS: P = RT/(V - b) - a(T)/[V(V + b))
derestimated (Pedersen et al., 1984). Using the SRK-equation for density cal-
a;(T) = ao;",;(T)
culations in cases where the liquid phase densities are of importance is there-
",;(T) = [1 + m;(1 - T~5)F fore not recommended.
] \
m; = 0.480 + 1. 574w; - 0.176w;2 It is generally advantageous to use the same density correlation for both
gas and liquid phases (e.g., to avoid discontinuities in the near critical re- 1,
ad = 0.42747 R2 To;2/Po;
gion). Peneloux et al. (1982) have presented a consistent method for correc- !
b, = 0.08664 R T,;lP,;
tion of SRK-volumes using an equation of state of the following form:
a = L: L: (a;a)o'\z;z(1 - k;)
i j
P = RT/(Y - b) -:- a/[(Y + c)(Y + b + 2c)] (5-.39)
b = L: z.b,

where e is a constant; which for a mixture is calculated as follows:
Fugacity coefficient at constant composition, </>:

In </> = Z - 1 - In (Z - B) - (A/B)(1 + B/Z) e = L CZ (5-40)

i I

Z = rv IRT; B/Z = blV; A/B = a/bRT

Fugacity coefficient of component i in a mixture, </>;:

e, is the value of e for component i. The SRK-equation (Equation 5-7), and
In </>; = - In (Z - B) + (Z - l)b;lb
Equation 5-.39 give identical results for saturation points and K-factors, but 11
different results for gas and liquid phase volumes. The SRK-volume, V, and :
1 '
- (A/B) [~ (2a; ~ zaf5(l - k;)) - bJb] In (1 + B/Z) the Peneloux-volume, Y, are related as follows:

Enthalpy at constant
H;d = "
'-'
composition,
zH;d
' ,
H:
(id means ideal gas)
Y=V-c

The parameter e can therefore be regarded as a volume translation pa-

(5-41)
I'
.) ,
:i
"1
Hi d
= 1 T

Tref
c~ dT rameter. Martin (1979) has a discussion on the volume translation concepto
For non-hydrocarbons and for hydrocarbons < C7 the following expres- I
'i_,

sion is used for c: l',

H"" = RT[Z -1 - (AlE) [1 -:; :;]ln (1 + B/Z)]

e, = 0.40768 (RTcJPci) (0.29441 - (ZRA)) (5-42)

T da
dT
= - L: L: z;zm(a;a,T,)05(l
j
- k;)

H = H;d + H"" where (ZRA) is the Racket compressibility factor (Spencer and Danner, 197.3)
Entropy at constant composition, S:
of component i for which the following approximation is used:

S;d = L: z;(S;;d -
i
R In z;) = L: z,
i
[
1. ~T
T

lref
Cid
dT - R In -
P
Pref
- R In z,
] (ZRA) = 0.29056 - 0.08775 W (5-4.3)

Peneloux et al. (1982) suggest finding the c-value of paraffinic,

S,~ = In (Z - B) + (A/B) [:!:a dT
da] In (1 + B/Z)
naphthenic, and aromatic C7 + -components, respectively, from fifth-degree
S = S;d + S'" polynomials in the carbon number. Pedersen et al. (1984) have found that
this procedure works reasonably well for gas and gas condensate mixtures,
 90 Properties oj Oils and Natural Gases Equations oi State 91

while inaccurate results are obtained for heavy oil mixtures. The c-parame- The values of a and b of a C7 +-fraction are found from the following ex-
ter of C7 +-components (or fractions) may instead be found as the difference pressions:
in the results for the molar volume at atmospheric pressure and 15C, calcu-
lated with the SRK-equation and determined experimentally (Pedersen et In aC7+ = 3.8405985 X 1O-3MW - 9.5638281 X 1O-4MWlp
al., 1988). + 2.6180818 x 102/T + 7.3104464 x 1O-6MW2
Far from the critical point it is of less importance to get consistent results + 10.753517 (5-47)
for gas and liquid phase densities. Liquid density correlations developed
specifically for petroleum mixtures are then often superior to the Peneloux bC7+ = 3.4992740 X 1O-2MW - 7.2725403 p + 2.2323950
procedure. Examples of such procedures follow. X 1O-4T - 1.6322572 x 1O-2MWlp + 6.2256545 (5-48)

where MW = molecular weight of the total C7 +-fraction

The Alani-Kennedy Equation p = density (in g/cm'') of the total C7 +-fraction at 1 atm and
15C
The form of the Alani-Kennedy equation (1960) is: a and b of a mixture are found as molar averages. The Alani-Kennedy equa-
tion gives accurate results for liquid densities, but it is not applicable to va-
y3 - (RT/P + b)y2 + aY/P - ab/P = O (5-44) por phase densities.

For pure substances a = K exp (n/T) (5-45)

The Standing-Katz Procedure
and b = mT + C (5-46)
This was originally developed as a graphical procedure (Standing and
where K, n, m, and C are constants. Values for the lighter hydrocarbons are Katz, 1941). Analytical expressions (Pedersen et al., 1984) approximating
given in Table 5-3. It is suggested using the methane values also for N2, CO2, the graphical form petmit the correlations to be applied in computer calcu-
and H2S. lations also.
The procedure involves the following steps (the densities are expressed in
g/cm '):
Table 5-3
Constants ot the Alani-Kennedy Equation tor Pure Hydrocarbons 'l\'
l. (H2S + C3 +) Density. The density af standard conditions of the
~"I (H2S + C3 +) fraction is calculated as follows
Component K n m x 10' C
C, (70-300F) 9160.6413 61.893223 3.3162472 0.50874303 p(H2S + C3+) = 1: MWx/ 1: MWx/p (5-49)
C, (301-460F) 147.47333 3247.4533 - 14.072637 1.8326695 i i

C2 (100-249F) 46709.573 - 404.48844 5.1520981 0.52239654 The index i refers to H2S, C3, and heavier components. The pure com-
C2 (250-460F) 17495.343 34.163551 2.8201736 0.62309877 ponent densities can be taken from Table 5-4, whereas a measured C7 +
C3 20247.757 190.24420 2.1586448 0.90832519
density must be available.

ic, 32204.420 131.63171 3.3862284 1.1013834 2. Correction for C2. The weight fraction W2 of C2 of the (H2S + C2 +)
fraction is determined. The density at standard conditions of the
n-C. 33016.212 146.15445 2.9021257 1.1168144
(H2S + C2+) fraction is then calculated from the equation:
n-Cs 37046.234 299.62630 2.1954785 1.4364289

n-C6 52093.006 254.56097 3.6961858 1.5929406 p(H2S + C2+) = p(H2S + C3+) - Aa - Alal - A2a2 (5-50)
Units: P in psa: T in R; V in ft3/lb-mol; R = 10.7335 Ib-ft3/in.2R Ib-mol.
Source: Alani and Kennedy (1960). where Aa = 0.3158 W2 (5-51)
92 Properties 01 Oils and Natural Gases Equations 01 State 93

Table 5-4 b3 = 1 + 30 w(5 W - 1) (5-61)

Standing-Katz Pure Component Densities at 1 atm and 15C
b, = 1 - 60 W + 750 w} - 2,500 w3 (5-62)
Density
Component (g/cm3)

H2S 0.7970
5. Pressure Correction. The density Pp at the actual pressure P (in psia)
and 60C is found from the equation:
C3 0.5072

ic, 0.5625
(5-63)
Pv = Po - Co - Cjc - C2C2- C3C3
n-C. 0.5836

i-C5 0.6241 where Co = - 0.034674 + 0.026806 C4 + 0.003705 C5

n-C5 0.6305 + 0.000465 Cs (5-64)

n-C6 0.6850
C = - 0.022712 + 0.015148 C4 + 0.004263 Cs
Sauree: Standing and Katz (1941).
+ 0.000218 Cs (5-65)

.I
A = - 0.2583 W2 (5-52) C2 = - 0.007692 + 0.003521 C4 + 0.002482 e,
+ 0.000397 Cs (5-66)
I
A2 = 0.01457 W2 (5-53)
C3 = - 0.001261 - 0.000294 C4 + 0.000941 Cs (

a = 3.3 - 5.0p (H2S + C3+) (5-54) + 0.000313 C6 (5-67)

\

a2 = 1 + 15(p(H2S + C3+) Cj = 1 - 2(P - 500)/10,000 (5-68)

- 0.46) (2.5p(H2S + C3+) - 2.15) (5-55) (
C2 = 1 + 6(P - 500)[(P - 500)/10,000 - 1]/10,000 (5-69)
3. Correction for CO2. Thedensity of the (C02 + H2S + C2+) fraction at
standard conditions is calculated on additive volume basis using C3 = 1 - 12(P - 500)/10,000 + 30[(P - 500)/10,000]2 I

p(H2S + C2+) and a CO2 density of 0.8215 g/cm3. - 20[(P - 500)/10,000]3 (5-70)
4. Correction for C and N2 The weight fraction W of (C + N2) of the
total mixture is calculated. The density Po of the total mixture at stan- C4 = 3.4 - 5 Po (5-71)
dard conditions is then calculated from the equation:
Cs = 1 + 15(po - 0.48)(2.5 Po - 2.2) (5-72)
Po = p(C02 + H2S + C2+) - Bo - Bb (5-56)
Cs = 1 - 30(po - 0.48) + 187.5(po - 0.48)2
where Bo = 0.088255 - 0.95509 b2 + 0.007403 b3 - 312.5(po - 0.48)3 (5-73)
- 0.00603 b4 (5-57)
6. Temperature Correction. Finally, the density p of the total mixture at
B = 0.142079 - 0.150175 b2 + 0.006679 b3 the actual temperature T (in F) and pressure can be calculated
+ 0.001163 b, (5-58)
p = Pp - Eo - Eje - E2e2 - E3e3 (5-74)
b = p(C02 + H2S + C2+) - 0.65 (5-59)
where Eo = 0.055846 - 0.060601 e, + 0.005275 es
b2 = 1 - 10 W (5-60) - 0.000750 es (5-75)
 Equations o] State 95
94 Properties 01 otts and~Natural Gases

Table 5-5
E = 0.037809 - 0.049262 e4 + 0.012043 es
API Pure Component Densitles at 1 atm and 15C
+ 0.000455 e6 (5-76)
Density
E2 = 0.021769 - 0.032396 e4 + 0.011015 es Component (g/cm3)
+ 0.000247 e6 (5-77)
N2 0.804

E3 = 0.009675 - 0.015500 e4 + 0.006520 es CO2 0.809

- 0.000653 e6 (5-78) H2S 0.834

C, 0.300
e = 1 - 2[(T - 520)/200] (5-79) C2 0.356

e2 = 1 + 6[(T - 520)/200][(T - 520)/200 - 1] (5-80) C3 0.508

i-C4 0.563
e3 = 1 - 12[(T - 520)/200] + 30[(T - 520)/200]2 n-C4 0.584
- 20[(T - 520)/200P (5-81) te,
, 0.625

n-C, 0.631
e, = 3.6 - 5 pp (5-82)
C6 0.664
es = 1 + 15(pp - 0.52)(2.5 pp - 2.3) (5-83) Source: API (1982).

e6 = 1 - 30(pp - 0.52) + 187.5(pp - 0.52)2

- 312.5(pp - 0.52)3 (5-84)
p-values for non-hydrocarbons and for C-Ce are given in Table 5-5.
The Standing-Katz procedure gives accurate density results for liquid The measured densities at standard conditions are used for the C7 + -
phases with a large concentration of heavy components. The procedure is components.
not applicable to vapor phases. 4. Calculation of the density correlation factor, C, at standard conditions
and at the actual conditions. C is given by the following expression:

The API Method C = A + A2 x T; + A3 x T;2 + A4 x T;3 (5-86)

Calculation of the liquid density using the API method (1982) involves the Each of the coefficients A-~ in Equation 5-86 is determined by an
following steps: expression of the following form:

1. Calculation of the pseudocritical temperature (T~) and pressure (P~) A = B + B2 X P; + B3 X P; 2 + B4 X P; 3 + Bs x P; 4 (5-87)

for the mixture in question. These values are obtained as molar aver-
ages of the pure component critical temperatures and pressures. where the Bj-values are given in Table 5-6.
2. Calculation of the pseudoreduced temperature (T; = T/T~) and the 5. Calculation of the unknown density from the equatin:
pseudoreduced pressure (P; = P/P~).
3. Calculation of the average density for the mixture at 60F and 1 atm p = p(C/C2) (5-88)
from the following formula:
where the CI and C2 are the density correlation factors (Equation
n n \ 5-86) at standard conditions, and at the actual conditions, respec-
p = E xMW/ E(xMW/p? ) (5-85)
tively.
i~ 1 i~ 1
96 Properties 01 Oils and Natural Gases Equations 01 State 97

Table 5-6 Jensen, B. H., "Densites, Viscosities and Phase Equilibria in Enhanced Oil
Coefficients of Equation 5-87 for Calculation of Liquid Denslties Using the Recovery," Doctoral Dissertation, Technical Univ. of Denmark, Lyngby,
API-Method 1987.
Kirn, H., Lin, H.-M. and Chao, K.-C., "Cubic Chain-of-Rotators Equation
B, B2 B3 B. B5 of State," Ind. Eng. Chem. Fundam., 25, 1986, pp. 75-84.
A, 1.6368 - 0.04615 2.1138(10-3) - 0.7845(10-5) - 0.6923(10-6) Lee, B. I. and Kesler, M. G., "A Generalized Thermodynamic Correlation
A2 - 1.9693 0.21874 - 8.0028(10-3) - 8.2328(10-5) 5.2604(10-6) Based on Three-Pararneter Corresponding States," AIChE [., 21, 1975,
2.4638 12.8763(10-3) 14.8059(10-5) - 8.6895(10-6)
pp. 510-527.
A3 - 0.36461
Leland, T. W., Phase Equilibria and Fluid Properties in the Chemical In-
A. - 1.5841 0.25136 - 11.3805(10-3) 9.5672(10-5) 2.1812(10-6)
dustry, Frankfurt/Main: DECHEMA, 1980, pp. 283-333.
Source: API (1982). Martin, J. J., "Cubic Equations of State-Which?" Ind. Eng. Chem. Fun-
dam., 18, 1979, pp. 81-97.
Michelsen, M. L. and Heidemann, R. A. "Calculaton of Tricritical Points,"
Fluid Phase Equilibria, 39, 1988, pp. 53-74.
The API-density method is only applicable to liquid phases. Results for
Mollerup, J. and Rowlinson, J. S., "The Prediction of the Densities of Lique-
vapor and liquid phase den sities using each of the correlations mentioned
fied Natural G~s and Lower Molecular Weight Hydrocarbons," Chem.
previously are given in Chapter 9. Eng. Sci., 29, 1974, pp. 1373-1381.
Pedersen, K. S., Thomassen, P., and Fredenslund, Aa., "Therrnodynamcs of
Petroleum Mixtures Containing Hydrocarbons. 2. Flash and PVT Calcu-
lations with the SRK Equation of State," Ind. Eng. Chem. Process Des.
REFERENCES Dev., 23, 1984, pp. 566-573.
Pedersen, K. S., Thomassen, P, and Fredenslund, Aa., "Characterzaton of
Abbott, M. M., "Cubic Equations of State: An Interpretive Beview," Chap- Gas Condensate Mixtures," paper presented at the 1988 AIChE Spring
ter 3 in: Chao, K. C. and Robinson, R. L., Equations oj State in Engi- National Meeting, New Orleans, LA, March 6-10, 1988.
neering and Research, Advances in Chemistry Series 182, Am. Chem. Peneloux, A., Rauzy, E. and Frze, R., "A Consistent Correction for Red-
Soc., Washington, D.C., 1979. lich-Kwong-Soave Volumes," Fluid Phase Equilihria, 8, 1982, pp. 7-23.
Adachi, Y., Lu, B. C.-Y., and Sugie, H., "A Four-Parameter Equation of Peng, D.-Y. and Robinson, D. B., "A New Two-Constant Equation of
State," Fluid Phase Equilibria, 11, 1983, pp. 29-48. State," Ind. Eng. Chem. Fundam., 15, pp. 59-64.
Alani, H. G. and Kennedy, H. T., "Volumes of Liquid Hydrocarbons at Pitzer, K. S., "Volumetric and Thermodynamic Properties of Fluids. 1. The-
High Temperatures and Pressures," ]. Peto Tech., Nov. 1960, pp. 272-273. oretical Basis and Virial Coefficents," ]. Am. Chem. Soc., 77, 1955, pp.
American Petroleum Institute, Technical Data Book-Petroleum Reiining, 3427-3433.
API, New York, 1982. Redlich, O. and Kwong, J. N. S., "The Thermodynamics of Solutions. V. An
Benedict, W, Webb, G. B., and Rubin, L. C., "An Empirical Equation for Equation of State. Fugacities of Gaseous Solutons," Chem. Rev., 44,
Thermodynamic Properties of Light Hydrocarbons and their Mixtures. I. 1949, pp. 233-244.
Methane, Ethane, Propane, and Butane," ]. Chem. Phys., 8, 1940, pp. Reid, R. C., Prausnitz, J. M., and Sherwood, T. K., The Properiies o] Gases
334-345. and Liquids, 3rd ed., New York, McGraw-Hill, 1977.
Carnahan, N. F. and Starling, K. E., "Interrnolecular Repulsions and the Soave, G., "Equilbrium Constants from a Modified Redlich-Kwong Equa-
Equation of State for Fluids," AIChE i., 18, 1972, pp. 1184-1189. tion of State," Chem. Eng. Sci., 27, 1972, pp. 1197-1203.
Edmister, W. C. and Lee, B. I., Applied Hydrocarbon Thermodynamics, Spencer, C. F. and Danner, R. P., "Prediction of Bubble-Point Density of
Volume I, Houston: Gulf Publishing Company, 1984. Mixtures," ]. Chem. Eng. Data, 18, 1973, pp. 230-234.
Ishikawa, T., Chung, W K., and Lu, B. C.-Y., "A Cubic Perturbed, Hard Standing, M. B. and Katz, D. L., "Density of Crude Oils Saturated with
Sphere Equation of State for Thermodynamic Properties and Vapor-Liq- Natural Gas," Am. Inst. Min. Metal/. Eng., Meeting in Los Angeles, Oct.
uid Equilibrium Calculations," AlChE t., 26, 1980, pp. 372-378. 1941, Tech. Pub. No. 1397, pp. 159-165.
 98 Properties oj Oils and Natural Gases

Starling, K., Fluid Thermadynamic Properties far Lght Petroleum Systems,

Houston, Gulf Publishing Company, 1973.
Thomassen, P., Pedersen, K. S., and Fredenslund, A., "Adjustment of the
C7 + -Molecular Weight in the Characterization of Petroleum Mixtures
Containing Heavy Hydrocarbons," SEP Report 8617, Instituttet for
Kemiteknik, The Technical University of Denmark, 1986. Chapter 6
Tsonopoulos, C. and Heidman, J. L., "Hgh-Pressure Vapor-Liquid Equlb-
ria with Cubic Equations of State," Fluid Phase Equilibria, 29, 1986, pp. Flash Calculations
391-414.
Tzouvaras, N., "Equatons of State for Vapor-Liquid Equilibrium Data Re-
duction," Ph.D. Thesis, Rensselaer Polytechnic Institute, Troy, NY, 1986.
van der Waals, J. D., Doctoral Dissertation, Leiden, Holland, 1873. The flash process is illustrated in Figure 6-1. The feed stream has N com-
ponents, i, with the composition z., i = 1,2, ... , N (mole fraction). The feed
stream is initially in the single-phase region. After expansion through the
valve into the flash chamber, which here operates at fixed temperature, T,
(
and pressure, P, two or more phases are formed. In our illustration, a vapor
phase of composition y, and a liquid phase of composition Xi, i = 1, 2, ... ,
( N, are formed. The fraction, on molar basis, which enters into the vapor
(
phase, is called 3. The vapor phase is assumed to be in equilibrium with the
liquid phase, i.e., the flash chamber contains exactly one equilibrium stage.
(

(
. - Vapor
Yi
a moles

(
(1
Fe ed
m ole)

zi
r ~- T,P
(fixed)

(
i = 1,2 N
- Liquid
xi
(1-a) moles

Figure 6-1. Schematic diagram of a ~LE flash chamber.

A calculation of the equilibrium conditions of a given mixture at specified

pressure and temperature is called a (P,T)-flash calculation. It is used to de-
termine:
1. The number of phases
2. Molar amounts of each phase
3. Molar phase compositions
4. Phase densities
5. Molar enthalpy, entropy, and heat capacity of each phase
A flash problem may alternatively be stated as a (P,H)-flash, or as a (P, S)-
flash. In the former case, in addition to the molar composition, the pressure
and the molar enthalpy are specified; in the latter case the pressure and the

99
 Flash Calculations 101
100 Properties 01 Oils and Natural Gases

molar entropy are specified. When solving the flash problem in those cases, ergy of phase 1 may be approximated by a Taylor series expansion truncated
the temperature is calculated in addition to the properties mentioned above. after the first order term:
(P,T) , (P,H), and (P,S)-flashes each have a unique solution. This is in con-
trast to (T,H), (T,S), (P, (3), and (T, (3)-flashes, which may have two, one, or
no solutions. (T,H) and (T,S)-flash calculations will not be covered here, as GI = Go _ 1.J
~ el. - (aGi) (6-3)
they are seldom used in practice. Calculation of the pressures and tempera-
i= 1 ani n

tures corresponding to a fixed molar vapor fraction are treated as a phase

The Gibbs energy of the second phase is found to be:
envelope problem (see Phase Envelope Calculations), and not as a flash
problem. (6-4)
The equilibrium conditions are those where the Gibbs energy (G) is at a Gn = G(e, e2, e3, ... , eN)
minimum. The Gibbs energy is defined as follows:
The change in the Gibbs energy due to the phase split is hence:
N
G = E n, .t (6-1)
N N
i= 1
~G = GI + Gn - Vo = Ee ((.ti)n- (.t)o)= e E y ((.t)n- (.t)o) (6-5)
i= 1 i= 1
where ni and .tare the number of moles and the chemical potential of com-
N
ponent i, respectively. Procedures for computing G and .tifrom equations of
where e = E e, y is the mole fraction of component i in phase 11 and the :(
state are given in the Two-Phase (P,T)-F1ash section.
i=l
The chemical potential can be regarded as the "escapng tendency" of subindices o and 11 refer to the single phase and to phase 11, respectively. (
component i, and one way to escape is to form an acditional phase. Only Only one phase exists if ~G is greater than or equal to zero for all possible
i I
one phase exists if the total Gibbs energy increases for all possible trial com- i (
trial compositions of phase 11. The chemical potential, .t,may be expressed
positions of an additional phase. Two or more phases exist if it is possible to
in terms of the fugacity, f, as follows: I i
split the mixture into two phases having a total Gibbs energy that is lower
than that of the single phase. The final number of phases and phase compo- (6-6)
.t= .tp+ RT In f = .tP+ RT (In Z + In <I>+ In P)
sitions are determined to be those with the lowest total Gibbs energy. \

where .t0is a standard state chemical potential, <I>a fugacity coefficient, z a

mole fraction, P the pressure, and the subindex i stands for component i.
TWO-PHASE (P, T)-FLASH The standard state is in this case the pure component i at the temperature
and pressure of the system. Equation 6-5 may then be rewritten:
The calctilation of whether a given mixture at specified (P,T) splits into
two or more phases is called a stability analysis (Michelsen, 1982a). The ~G N ~
= E y (In y + In (<I>)n- In Z- In (<I>)o) (6-7)
starting point is the Gibbs energy (Go) the mixture would have if there was eRT i= 1
only one phase present:
where z is the mole fraction of component i in the total mixture. The stabil-
C, = G(n, n2, n3, ... , nN) (6-2)
ity criterio n can now be expressed in terms of mole fractions and fugacity
where ni stands for the number of moles of type i present in the mixture, and coefficients. Only one phase exists if:
N is the number of different components. The total number of moles
N N
E y (ln y + ln(<I>)n- In Z- In(<I>)o)~ O (6-8)
(= ~ ni) is called n. Secondly, the situation is considered where the mix-
i= 1

ture splits into two phases (1 and 11) with the compositions (n - e, n2 - e2,
n3 - e3, ... , nN - eN)and (e, e2, e3, ... , eN), where ei is small. The Gibbs en- for all trial compositions of phase 11.
 Flash Calculations 103
102 Properties oj Oils and Natural Gases

Equilibrium Equatons:
A minimum in the expression on the left hand side of Equation 6-8 will at
t' the same time be a stationary point. It can be shown that a stationary point
(6-15)
ti1, y <l>r = X<l>r (i = 1, 2, 3, oo., N)
must satisfy the equation:
r

Summation of Mole Fractions:

In y + ln(<l>)n- In Z - ln(<l>)o= k (6-9)

N
where k is independent of the component index. Introducing new variables,
Y, given by: E (y - Xi)= O (6-16)
= I

In Y = In y - k (6-10) In Equations 6-14 through 6-16 Xi,y, and Zare mole fractions in the liq-
, uid phase, the vapor phase, and the total mixture, respectively. {3is the mo-
'1 Equation 6-9 may be rewritten: lar vapor fraction. <l>v
and <l>L
are the fugacity coefficients of component i in
,1 the vapor and liquid phases, computed as shown in Chapter 5. Equations
l
i
In Y = In Z + In(<l>)o- ln(<l>)n (6-11) 6-14 through 6-16 contain (2N + 1) equations with (2N + 3) variables,
namely (Xl, X2,X3, oo., XN), (y, Y2, Y3, oo., YN),{3, T, and P. With T and P
Michelsen suggests using the following initial estimate for the ratio K, be- specified, the number of variables equals the number of equations, and
tween the mole fraction of component i in the vapor phase, and in the liquid Equations 6-14 through 6-16 can be solved for Xi, y, and {3. These three
phase: equations can be simplified by introducing the equilibrium ratio K = y;lx.
The following expressions may then be derived for Xand y:
(5
1
1
l' y
~= K =pexp.4
Pci 2 (1 -T
Te)) (6-12)
(6-17)
x, = z;l(1 + (3(K- 1))

where Tci is the critical temperature, and Pci is the critical pressure of com- y = Kx, (6-18)
, ponent L As initial estimates for Y, use Kz, f phase o is a liquid, and z;lK,
if phase o is a vapor. The fugacity coefficients, (<l>)11'corresponding to the and for K:
initial estimates for Y, are determined. Based on these fugacity coefficients,
new Yj-values are determined, and so on. For a single phase mixture this K = <l>f"/<l>r (6-19)
direct substitution calculation converges to either the trivial solution (i.e., to
two identical phases) or to Yj-values-Iulfillng the criterion: The (2N + 1) equations of Equations 6-14 through 6-16 may then be re-
N duced to the following (N + 1) equations:
E Y :5 1 (6-13)
= I (6-20)
In K = In <l>r - In <l>r (i = 1, 2, 3, .", N)
i which corresponds to a non-negative value of the constant k of Equation
N
11
6-9. A negative k-value would be an indication of the presence of two or
more phases. The molar composition of phase II obtained by solving Equa-
E (y - Xi)= E Z (K - 1)/(1 + (3(K- 1)) = O (6-21)
= 1
l tion 6-11 is then usually a good starting point in the calculation of the phase
l compositions. For two phases in equilibrium, the following set of equations
must be satisfied:
For a given total composition, a given, (T, P), and K estinated from the
stability analysis, an estimate of {3may be derived from Equation 6-21. New
1:
estimates of Xand y may be calculated from Equations 6-17 and 6-18, and
!
the K-factors are recalculated using Equation 6-20. A new {3-value is calcu-
Material Balance Equations:
lated from Equation 6-21. This direct substitution calculation may be re-
1.
peated until convergence. However, as described by Michelsen (1982b),
I {3y + (1 - (3) X= Z (i = 1, 2, 3, oo., N) (6-14)
!i

104 Properties oi Oils and Natural Gases Flash Calculatums 105

convergence may be accelerated by using, e.g., the general dominant eigen- ever, problems which cannot be converged in this manner occur relatively
value method followed by a Newton-Raphson iteration. frequently in multi-phase systems. According to Michelsen (1982b), New-
ton-Raphson iteration is not well suited for final convergence of difficult
problems. He instead recornrnends Gibbs energy minimization using the
MULTI-PHASE (P,T)-FLASH method discussed by Murray (1972).

A stability analysis as described in the previous section may be performed

on each of the two phases determined in a two-phase (P,T)-flash calcula-
tion. If the stability analysis shows that one or both of these phases are un-
stable, it is an indication that more than two phases existo A three-phase
(P,T)-flash calculation as sketched below may then be carried out. A stabl-
ity analysis on each of the three resulting phases may reveal that the system
is still unstable and a four-phase (P,T)-flash calculation may be performed.
This calculation scheme can be continued until the stage where the stability
analysis shows that all the calculated phases are stable.
For a system consisting of J phases, the analogue of Equation 6-21 is:
N
E z, (Kim - 1)/Hi = O (m = 1, 2, ... , J - 1) (6-22)
i= 1
(P,H) OR (P,S)-FLASH CALCULATIONS
When pressure and enthalpy (P,H) , or pressure and entropy (P,S), are used
as specification variables, the procedures outlined in prior sections are not
readily applicable since they require that the temperature, T, be known.
When performing (P,H) or (P,S)-flash calculations it is essential to provide a
good initial estmate-of the temperature. Figure 6-2 shows the enthalpy de-
pendence of temperature for a mixture of 50 % methane and 50 % ethane at
60 atm (from Michelsen, 1986a). The SRK equation as described in Chapter
5, is used for calculation of all thermodynamic properties, and the enthalpy
is chosen such that the enthalpy for an ideal gas equals zero at OC. Ideal

J-l H/R (K)

where n, = 1 + E {3m(Kim - 1) (6-23) o

m e I

{3mis the molar fraction of phase m. In Equation 6-23 Kim is defined as the
ratio of the mole fraction of component i in phase m to its mole fraction in
phase J. Initial estimates for the K-factors are available from the stability
-500
analysis for the preceding J-1 phase solution. Using a Newton-Raphsorr
iteration, the corresponding molar phase fractions, {3, may be found from
Equations 6-22 and 6-23. The phase compositions are subsequently found
from:

Yim= Zi Kim/Hi (m = 1, 2, ... , J- 1) (6-24)

-1000

YiJ = z/H (6-25)

where Yim,and Yu, are the mole fractions of component i in phase m and
phase J, respectively. The corresponding Fugacity coefficients may be calcu-
lated (from Equation 5-29 if the SRK-equation of state is used) and a new -1500 ' I I , , , ! !

estimate of the K-factors is established using an analogue of Equation 6-20. 210 220 230 240 250 260 270
As is the case for the two-phase (P,T)-flash calculations previously de- T( K)

scribed, the general dominant eigenvalue method may be used for accelera-
Figure 6-2. Molar enthalpy in dependence oftemperature lor an equimolar mix-
tion of the direct substitution procedure which has just been outlined. How- ture 01 methane and ethane at 60 atm. Source: Michelsen (1986a).
 106 Properties o] Oils and Natural Gases Flash Calculations 107

gas heat capacities are represented by third-degree temperature polynomials form at least two phases at equilibrium, and a stability analysis (see Two-
as described in Chapter 5. Phase (P,T) Flash Section) is not required. As initial estimates for the phase
It is observed that the enthalpy increases monotonically with tempera- compositions at T* , two phases of identical compositions are used, but with
ture, for this mixture. Hence, the temperature corresponding to any speci- different molar volumes (corresponding to the smallest and the largest root
fied enthalpy is easily located. The enthalpy-temperature variation is for the in the compressibility factor polynomial of Equation 5-21).
same mixture shown at 30 atm in Figure 6-3. In the temperature interval Michelsen (1986a) has described a general multi-phase flash procedure
223.7K < T < 240.1K, two different results may be obtained for the single- for solving a (P,H) specification for a given mixture. The procedure is out-
phase enthalpy of the mixture. (These results correspond to two different lined as follows for the case with only two phases.
real roots in the third-degree polynomial of Equation 5-21, for the compress- Two object functions are defined, namely:
ibility factor). The lower branch corresponds to a liquid state and the upper
branch to a vapor state. At temperatures below T = 233.0 K, the liquid N
state has the lower Gibbs energy (.e., it is the more stable state), and above g = L Z (K - 1)/l (6-26)
T*, the vapor phase has the lower Gibbs energy (and is the more stable = 1
state). At the transition temperature, T* ( = 233.0 K), the Gibbs energies of
the two phases are equal. It is found that the molar enthalpies of the two where L = 1 + (3 (K - 1) (6-27)
phases at T* are H/R = - 65 K, and HIR = - 455 K, respectively. If
the specified enthalpy is outside this interval, determining the correspond- and ~ = H - Hspec (6-28)
ing temperature is straightforward. For a specification in the interval
( - 1165 K to - 455 K) the transition temperature, T*, may be used as an where H is the molar enthalpy corresponding to the estimated phase cornpo-
initial temperature estimate. In the latter situation the system is known to sitions, and Hspecis the specified molar enthalpy. At convergence both g and
H/R (K)
~ are zero (g equals the left side of Equation 6-21). The molar enthalpy H
I o i I is calculated as follows:

II H = (3 HV + (1 - (3)HL (6-29)

11
where HV and HL are the molar vapor and liquid phase enthalpies, respec-
tively. These properties are calculated as shown in the Summary of Equa-

l'
-500

--~-
tions section in Chapter 5. The iterations proceed as follows:
l. Calculate (3 from Equation 6-2l.
2. Calculate In <pand a In <p/aT for each of the two phases (for the

-
!!
"" SRK-equation of state, <pmay be derived from Equation 5-29).
!I
L 3. Calculate K from Equation 6-20.
I
il
q ~ -1000
I
I I I 'L
I~

4. Calculate g from Equation 6-26.
1:
N

1, 5. Calculate ag/a(3 = - L Z (K - 1)2/l r (6-30)

i= 1
anda~/a(3 = HV - HL
-1500 I I I I I I I I I
210 220 230 I
I
240

T*
250 260 270 280

T(K)
6. Calculate ag =
aT
Z
1.J 1
V. (K a InaT K - (K -
1 1
l)T*/I-.)/I-.
1 ~ 1 ~ 1
(6-31)
1=
Figure 6-3. Molar enthalpy in dependence ot temperature for an equimolar mix-
ture of methane and ethane at 30 atm. Source: Michelsen (1986a). where Tt = (3 K a In K/aT (6-32)
108 Properties oj Oils and Natural Gases Flash Calculations 109
~\r:'

7. Calculate g2 from Equation 6-28. Initial testing establishes whether apure liquid water phase is separated
from the feed. This is done by comparing the chemical potential (see Equa-
tion 6-6) of water in a pure water phase:
8. Calculate ag2/OT = {3C~ + (1 - (3)q +
N
RT2 1: zH! (Ti" - (T*)2Iti) (6-33) = JLe.,+ RT(ln P + ln(<pw)pure)
(JLw)pure (6-37)
ie l

where T* * = {3K, (a In K/aT)2 (6-34) with the fugacity of water in the total mixture:

and C, v and CpL are the molar vapor and liquid phase heat capacities + RT(ln P + ln(<pw)mix+ In zw)
(JLw)mix = JL~v (6-38)
( = the temperature derivatives of the molar phase enthalpies).
9. Use a Newton-Raphson iteration procedure to determine the size of Zw is the mole fraction of water in the total mixture.
the next {3and temperature steps (fl{3 and flT).
10. Calculate new estimates for {3and for the K-factors: If (JLw) pure < (JLw) mix (6-39)

{3new= {3 + fl{3 (6-35) +eln z., > In (<pw)pure

or In (<Pw)mix (6-40)

_ a In K, fl T
1n Kinew- In K, + -- (6-36)
aT
11. CaIculate the new phase compositions from Equations 6-17 and 6-18.
12. Return to step 1.
A (P,S)-flash calculation may be performed using a procedure that is quite
analogous to what has just been outlined for a (P,H)-flash calculation.
In most cases the final equilibrium temperature T deviates less than 30K
from the initial temperature estimate, T*. In the case of difficult flash prob-
lems, (e.g., near critical fluids) it can be advantageous to use a different cal-
culation procedure. (P,H) and (P,S)-flash problems can both be expressed as
minimization problems. A (P,H)-flash may be solved by locating the mini-
mum of ( - S) subject to H = Hspec' Similarly a (P,S)-flash calculation may
be performed by locating the minimum of H subject to S = Sspec'
The (P,H) and (P,S)-flash algorithms outlined require approximately the
same computer time as the equivalent (P,T) algorithm.
THREE-PHASE (P,T)-FLASH WITH A LIQUID WATER PHASE
The need for performing three-phase flash calculations on petroleum
mixtures often arises when water is present, and where an almost pure liq-
uid water phase potentially may be formed. If the solubilities of other com-
ponents in the aqueous liquid phase are neglected, (P,T)-flash calculations
can be very much simplified (Michelsen, 1981) as compared with what was
described in the Multi-Phase (P,T)-Flash section.
apure liquid water phase will precipitate. The amount of water remaining
in the mixed phase is found as the amount fulfilling the criterion:
ln(<pw)mix+ In x., - ln(<pw)pure= O (6-41)
where x., is the mole fraction of water in the mixed phase, after removal of
the pure water phase. A stability analysis, as described under the Two-Phase
(P,T)-Flash heading, may then be carried out for the remaining mixed
phase. If it splits into two phases, the amount of the pure water phase must
be adjusted, considering that Equation 6-41 must be fulfilled for each of the
mixed phases.
Initially comparing the chemical potentials of water in the pure liquid
form with those of water in the mixed phase may alternatively reveal that
water does not condense from the feed. In that case, a conventional stability
analysis is performed for the total feed. If it is found that more than one
phase exists, a two-phase (P,T)-flash calculation is performed as described in
the section so named. Each iteration step is checked to see whether water
will precipitate from the mixed phases, i.e., whether Equation 6-39 is ful-
filled. The final result of the calculation may well be that three phases are
present, even though the initial test for water precipitation from the feed
was negative.
SIMPLIFIED TWO-PHASE (P,T)-FLASH
When using the SRK-equation of state, the binary interaction coefficients
of Equation 5-20 are usually assumed equal to zero for hydrocarbon-hydro-
 110 Properties o] Oils and Natural Gases Flash Calculatons 111

carbon interactions, while non-zero values are used for interactions between where F refers to the total mixture, or feed. Similarly, for the b-pararneter:
two non-hydrocarbons, and between a hydrocarbon and a non-hydrocar-
bono For mixtures containing up to 2 % non-hydrocarbons, binary interac- bF = (3 by + (l - (3) bL (6-51)
tion coefficients equal to zero are usually a good approximation. For such
mixtures, the two-phase (P,T)-flash calculation can be greatly. simplified al, and bL may be expressed in terms of these three parameters:
(Michelsen, 1986b).
r
When, with the SRK equation, all binary interaction coefficients equal aL = a - (3 ay
(6-52)
zero, Equation 5-20 can be reduced to: 1 - (3

a = (a')2 (6-42) b = bF - (3 by
L (6-53)
1-(3
N
where a' = 1: z, a' with a'= a?5 (6-43) Irrespective of the number of components, the flash problem has now
i= 1
been reduced to involve only three parameters, namely ay', by, and (3. The
calculation procedure is as follows:
and the expression for the fugacity coefficient given in Equation 5-29 can be
rewritten: l. Calculate aF' and bF
2. Estimate ay', by, and (3
In <p= qo + q"a' + Clab (6-44) 3. Calculate aL' and bL from Equations 6-52 and 6-53
4. Calculate the gas and liquid phase fugacities for each component
5. Calculate the K-factors ( = Y/Xi)
where qo = -In (:T (V - b)) (6-45) 6. Calculate the phase compositions using Equation 6-17 and 6-18
7. Evaluate the check-Iunctions:

(
q" = _ 2~ In ((V + b)) (6-46) el =
N

1: (y - Xi) (6-54)
bRT V
= 1
( N
e2 = 1..J
'\"'
y a' - ayr (6-55)
l(PV ) a" (6-47)
q~ = b RT - 1 + 2b q" i=l
N

The following relation exists between the mole fractions of a given corn-
e3 = 1: y b, -
i= 1
by (6-56)

ponent in two phases in equilibrium, and in the total mixture:

z, = (3 y + (1 - (3)x (6-48) If el, e2, and e3 are all equal to zero, the phase distributions at equilb-
rium have been determined. If not equal to zero, a Newton-Raphson itera-
Multiplication by a' and summation over all components yields the identity: tive correction of ay, by, and (3is performed, and the calculation returns to
step 3. '
N N N The calculation time is almost independent of the number of components.
1: z, a'= (31: y a'+ (1- (3)1: x, a' (6-49) Thus, for multi-component mixtures, much computation time can be saved
i= 1 i= 1 i= 1 as compared with the usual two-phase (P,T)-flash previously described. Pe-
dersen et al. (1985) have reported relative computation times for (P,T)-flash
or a{: = (3 a{r + (1 - (3)a{ (6-50) calculations on petroleum mixtures using the usual (P,T)-flash, and the
 112 Properties 01 Oils and Natural Gases Flash Calculations 113

k = O flash described in this section. The results are presented in Table To initialize a phase envelope calculation, it is convenient to specify a
6-1. fairly low pressure, e.g., P < 20 atm. Equation 6-12 may be used to find
Table 6-1 initial estimates of the K-factors, and using this simple K-factor model,
Relative Computer Time tor Flash Calculations on Petroleum Mixtures Equation 6-21 may be solved for the temperature T. Equations 6-20 and
6-21 are then solved simultaneously for T and In K, where now the actual
3 Hydro- 6 Hydro- 10 Hydro- 20 Hydro- 40 Hydro- K-factor model, (e.g., the SRK equation) is used. In addition, the deriva-
Program carbons carbons carbons carbons carbons tives of T and In K, with respect to P, can be calculated and used to generate
k;J*0 0.08 0.16 0.25 0.46 1.00 an estimate of the second point on the phase envelope at a slightly higher
k;J= O 0.05 0.06 0.08 0.09 0.15 pressure than the first one. From the third point, and on, the extrapolation is
based on the two last calculated points, and the corresponding derivatives.
*
k; O: Flash program with so me 01 the binary interaction coefficients different lrom zero.
Using this procedure the complete phase envelope can be constructed in a
k, = O: Flash program with all binary interaction coefficients equal to zero.
Source: Pedersen el al. (1985). stepwise manner, and even passing the critical point creates no problems.
Figure 6-4 shows the phase envelope calculated for a typical North Sea oil.
PHASE ENVELOPE CALCULATIONS

REFERENCES
The (P,T)-values which for a given mixture correspond to a fixed vapor
mole fraction, (3, are collectively called a phase envelope. Michelsen (1980)
has described a procedure for phase envelope calculations, and it is briefly Michelsen, M. L., "Calculaton of Phase Envelopes and Critical Points for
outlined in the following paragraph. Multicomponent Mixtures," Fluid Phase Equilibria, 4, 1980, pp. 1-10.
Michelsen, M. L., "Three-Phase Envelope and Three-Phase Flash Algo-
400.0 i i
rithms with a Liquid Water Phase," SEP Report 8123, Instituttet for
Kemiteknik, The Technical University of Denmark, 1981.
Michelsen, M. L., "The Isothermal Flash Problem. Part 1. Stability," Fluid
Phase Equlibria, 9, 1982a, pp. 1-19.
~ 300.0
Michelsen, M. L., "The Isothermal Flash Problem. Part II. Phase-Split Cal-
"' B = o culation," Fluid Phase Equilibria, 9, 1982b, pp. 21-40.
~ Michelsen, M. L., "Multphase Isenthalpic and Isentropic Flash Algo-
~
rithms," SEP Report 8616, Instituttet for Kemiteknik, The Technical Uni-
!
versity of Denmark, 1986a.

- -
B = 0.2
200.0
/
-- Michelsen, M. L., "Simplfed Flash Calculations for Cubic Equations of
State," Ind. Eng. Chm, Process. Des. Dev., 25, 1986b, pp. 184-188.

-
/' Murray, W., "Methods for Unconstrained Optmzation," Academic Press,
/

-- - - = London, 1972.

-
6 0.4
/ /'
Pedersen, K. S., Thornassen, E, and Fredenslund, Aa., "Therrnodynamics of
100.0
/ -: Petroleum Mixtures Containing Heavy Hydrocarbons. 3. Efficient Flash
,/
Calculation Procedures Using the SRK-Equation of State," Ind. Eng.
/ / .fo.6

-- -
./

/ 16 = I Chem. Process. Des. Dev., 24, 1985, pp. 948-954 .

/ /
./' /'
./'6 = 0.8
0.0 ' , , , I , I ,

-100.0 0.0 100.0 200.0 300.0 400.0 500.0 600.0

Temperaturo ('e),

Figure 6-4. Phase envelope of a North Sea oil mixture calculated using the SRK
equation. {:J is the mole fraction of the vapor phase.
 Characterization Procedures 115

In P, = 8.3634 - 0.0566/Se
- (0.24244 + 2.2898/SG + 0.1l857/SG2) X 10-3 TB
+ (1.4685 + 3.648/SG + 0.47227/SG2) x 10-7 T~
- (0.42019 + 1.6977/SG2) x 10-10 ~ (7-2)
Chapter 7
w = (In PBr - 5.92714 + 6.09648/TBr
Characterization Procedures + 1.28862 In TBr - 0.169347T~r)/(15.2518 - 15.6875/TBr
- 13.4721 In TBr + 0.43577Tt) (for TBr<0.8) (7-3)

w = - 7.904 + 0.1352K - 0.007465K2 + 8.359TBr

+ (1.408 - 0.01063K)/T Br (for T Br> 0.8) (7-4)
CLASSIFICATION OF THE CONSTITUENTS OF OIL AND GAS
MIXTURES The boiling point, T B, and the critical temperature, Te, are in R, and
SG s the 60F/60F specific gravity (the density of the fraction at at-
As described in Chapter 2, naturally occurring oil and gas mixtures will mospheric pressure and at 60F, divided by the density of water at the
often be too complex for a complete identification of components. The com- same conditions; i.e., se is approximately equal to the density of the
ponents of a reservoir fluid may instead be elassified as follows: fraction in g/crn" at atmospheric pressure and 60F). The critical pres-
sure, P is in psia. TBr = TB/Te, and PBr = PB/Pe, where PB s the pres-
Defined components are those for which there exists a quantitative gas chro-
sure where TB was measured. K is the Watson characterization factor
matographic analysis and for which Te, P and w are known. This in-
which equals T~3/se. Thomassen et al. (1986) have suggested that
eludes non-hydrocarbons and hydrocarbons up to at least C6.
Equation 7-3 be modified to:
TBP fractions cover the true boiling point fractions from C7 and up to max.
C30, each containing many different components. The density and the av-
w = (In PBr - 5.77881 + 6.05615/TBr + 1.37882 In TBr
erage molecular weight may be measured on each fraction.
- 0.173914 T~r)/(15.5523 - 15.7915/TBr
The TBP resulue consists of components that are too heavy to be separated
- 12.7855 In TBr + 0.43487 Tt) (for TBr<0.8) (7-5)
using a TBP distillation procedure. The density and the average molecu-
lar weight may be measured for the total sample., but neither boiling
2. Winn-Relations (Winn, 1957; Sim and Daubert, 1980):
point nor molecular weight distributions are determined.
Te = exp [4.2009 ~.08615 SG004614]/1.8 (7-6)

TBP FRACTIONS P, = 6.1483 X 1012 TB'2.3177SG24853 (7-7)

Because pure component properties are not available, equation of state TB and Te are in K, SG in 15C/15C, and Pe in Pa.
(EOS) calculations on mixtures ineluding heavy hydrocarbon fractions re-
quire a characterization procedure for determination of the EOS-parame- 3. Cavett (1964) Relations: *
terso Each TBP fraction may be represented as one pseudo-component. The Te = 768.071 + 1.7134 TB - 0.10834 x 1O-2T~
critical temperature (Te), and pressure (Pe), and the acentric factor (w) to be + 0.3889 X 1O-6T~ - 0.89213 X 1O-2TBAPI
used in the EOS-calculation can be found from the boiling point, the molec- + 0.53095 X 1O-6T~ API + 0.32712 X 1O-7T~APJ2 (7-8)
ular weight, and the density using empirical correlations. Examples of such
correlations are given below:
1. Lee-Kesler Correlations (Kesler and Lee, 1976): .
The term 0.53095 X 10-6 T~ API in Equation 7-8 was in Cavett's work originally
Te = 341.7 + 811SG + (0.4244 + 0.1l74SG)TB 0.53095 x 10-5 T~ API.
+ (0.4669 - 3.2623SG) X 105/TB (7-1)

114
116 Properties of Oils and Natural Gases
Characterization Procedures 117
loglO Pe = 2.829 + 0.9412 X 1O-3TB - 0.30475 X 1O-5T~ Moa
+ 0.15141 X 1O-8T~ - 0.20876 X 1O-4TBAPI
+ 0.11048 X 1O-7T~API + 0.1395 X 1O-9T~APJ2
- 0.4827 X 1O-7TBAPJ2 (7-9) -,
3.0+ I I ~\
In Equations 7-8 and 7-9, API equals 141.5/Se - 131.5, where se \
\
is the 60F/60F specific gravity. The critical temperature, Te, and the :\
boiling point, T B, of the fraction are given in F and the critical pres- (

sure, Pe, in psia.

4. Edmister's (1958) Acentric Factor:

2,Or
1.0
"" llHH
w = ~ [ log Pe ] _ 1 (7-10) O.
7 Te/TB-1 10 15 20 25
Carbon Number

where P, is in atmospheres. Figure 7-1. Molar disiribution of C7+ -fractions in a North Sea oil (molar composi-
tion in Table 7-1).
5. The correlations of Riazi and Daubert (1980a):
Table 7-1
Te = 24.2787 ~.58848 seO.3596 (7-11) Molar Composition 01 the North Sea Oi!
whose C7+-Composition is Plotted in Figure 7-1
r, = 3.12281 X 109 TB2.3125se2.3201 (7-12)
Density
Molecular (g/cm3) at
where Te and TB are in R, P, in psia, and se in 60F/60F. Component Mole % Weight 15C, 1 atm
N2 0.57
CO2 1.20
THE TBP RESIDUE C, 45.85
C2 3.68
C3 1.23
Neither boiling point nor rnolecular weight distributions are available for -C4 0.63
the TBP residue, i.e., Equations 7-1 through 7-12 are not readily applicable. n-C4 0.87
Figure 7-1 shows, for a typical North Sea oil whose analytical data are given te, 0.65
in Table 7-1, the mole fraction from C7 to C29 versus molecular weight. As is n-e; 0.43
C6 0.87 86.0
seen from the Figure, the composition of the C7 + -fraction is an almost con- 0.644
Cl 3.28
tinuous distribution in molecular weight. The boiling point fractions may be 93.0 0.733
Ca 3.55 107.0
represented using an appropriate analytical expression for the mole fraction 0763
Cg 3.44. 120.0 0.785
versus molecular weight (or carbon number). This analytical expression may C10 2.53 131.6 0.808
then be applied to estimate the molar composition of the TBP residue. Vari- Cll 2.48 147.6 0.815
ous distribution functions are suggested in the literature. Yarborough (1978) C'2 2.13 158.6 0.836
C'3 2.45 172.4
and Pedersen et al. (1984) have suggested for gas condensate mixtures to as- 0.850
C'4 2.20 184.8 0.861
sume a logarithmic distribution of the molar fraction Z versus the carbon c., 2.06 197.4 0.873
number CN: C'6 1.66 209.0 0.882
c., 1.69 226.5 0.873
In z = A + B X CN (7-13)
(table continued on next page)
118 Properties 01 Oils and Natural Gases Characterization Procedures 119
O.Or'---------- -,
Table 7-1
Continued

Density -2.5

Molecular (g/cm3) at
Component

c.,
c.,
Mole %
1.45
1.27
Weight

242.6
253.7
15C, 1 atm

0.875
0.885
-sJ

-- -
Measured mole fractions
Estimated mole fractions
e
o
e20 1.21 262.0 0.903 +-' <,
e; 0.90 281.1 0.898 u
'"
e,
-7.5 <,
<,
e22 0.95 293.2 0.898 4-
<,
e23 0.77 307.3 0.899 ~o <,
e24 0.75 319.6 0.900 .s 10.0 <,
<,
e25 0.66 333.3 0.905 .s <,
e26 0.57 345.9 0.907 <,
e27 0.66 361.1 0.910 .12.51- <,
e 0.62 373.5 0.915
<,
28
<,
e29 0.60 380.8 0.920
C30+ 6.14 624.0 0.953 , ,
-IS.J lO
I I

20
I

30
I

40 50 60
I

70 80

Carbon Number

A and B are constants determined from the measured mole fraction and the Figure 7-2. Measured and estimated (Pedersen et al., 1984) C7+-composition for
measured molecular weight of the plus-fracton, Figure 7-2 shows, for a gas the gas condensate of Table 4-6.
condensate (molar composition in Table 4-6), a plot of ln Z of the TBP frac-
tions versus the carbon number. Also sketched in Figure 7-2 is the distribu- 800.
tion estimated for the TBP residue. To get agreement with the experimental
A
I
mole fraction and the average molecular weight of the TBP residue, the con- 700J /
stants A and B of Equation 7-13 may for the TBP residue deviate slightly
from those determined by a least squares fit to the data. for the TBP frac-
tions. For oil mixtures, the composition of the TBP residue may alternatively
600.11- .'
.:
Measured TBP-data
Artificially generated TBP-data
/
/

/
be estimated by extrapolating the TBP distillation curve to 100 % distilled - - -. Extrapolated TBP-curve ,,-
off (Pedersen et al., 1984) by fitting a fifth-degree polynomial to the experi-
mental TBP data, plus the following artificially generated data:

TSOwt%= 1O-02127T;ii~~% se-o 6495 (7-14)

~'""l
+-'
'"
s,
QJ
O-
E
(!!.
400.

300.


/'
/'
/'
,/

TIOO,m\\t-,Cl
= 1OI.OS3
TO.7097seO.6717
50wt% (7-15)
200.0~



T50wt%,TSOwt%,and TIOOwt%are the temperatures in K at which 50,80, and
100 wt % of the crude oil has been distilled off. se is the specific gravity of 100l

the TBP sample that roughly corresponds to the C7 + -fraction of the total
0.0
mixture. The cumulative weight % distilled off versus temperature is shown 0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0 100.0

in Figure 7-3 for the North Sea oil of Table 4-7, Also sketched is the ex- Weight % distilled off
trapolated TBP curve. Mole fractions below 10 - 6 are neglected, and the Figure 7-3. Measured TBP-distillation data for the oil mixture of Table 4-7, and
heaviest component considered is Cso. estimated extrapolation to 100% distilled off. (Pedersen et al., 1984).
120 Properties of Oils and Natural Gases Characterization Procedures 121

Whitson (1983) suggested describing the C7vmolar composition as a 1 = 3.583 X 1O-3T~OI47(MWlp)-04787 for MW<200 (7-21)
gamma-distribution. The molar fraction, Z,of component i, with a molecu-
lar weight of MW, is given by the expression: 1 = 1.4 X 1O-3T~o9(MWlp)-O.3984 for MW>200 (7-22)

Z = P(MW) - P(MW - 1) (7-16) where MW is the molecular weight, TB is the boiling point in R, and p is
the liquid density in g/cm3 at 20C and 1 atm. Based on the refractive index,
where the specific gravity, and the molecular weight, the PNA distribution (in mole
%) can be estimated as described by Nes and Westerns (1951):
P(MW) = (MW - 1)),,-lexp( - (MW - 1))1(3)
(7-17)
(3"r(ex) v = 2.51 (n - 1.4750) - p + 0.8510 (7-23)

I' is the gamma function, MW is the molecular weight, and ex, (3, and 1) are w = p - 0.8510 - 1.11 (n - 1.4750) (7-24)
parameters defining the distribution. These parameters are determined
from the avalable analytical information. In cases where the analytical in- %A = 430 v + 3,660/MW for v>O (7-25)
formation is limited, ex is used as a regression variable, which is determined
from measured PVT-data. %A = 670 v + 3,660/MW for v -c O (7-26)

R = 820 w + 10,000/MW for wo-O (7-27)

CHARACTERIZATION PROCEDtJRES
R = 1,440 w + 10,600/MW for w-c (7-28)

Pedersen et al. (1984) use Equations 7-8 and 7-9 to determine Te and Pe, %N = R - %A (7-29)
and Equations 7-4 and 7-5 to determine w. The equations are used sepa-
rately for the paraffinic (P), naphthenic (N), and aromatic (A) compounds %P = 100 - R (7-30)
within a TBP fraction. Te, Pe, and w of the total fraction are found as molar
averages. The specific gravities of the P, N, and A components are found To use Equations 7-4,7-5,7-8, and 7-9 to determine Te, Pe, and w, TB and
from the specific gravity of the total fraction, and the following relations: SG of the pseudocomponents of the TBP residue must be estimated. Pedersen
et al. use a logarithmic distribution for the specific gravity versus carbon
SGN = 1.12 SGp (7-18) number. Experimental and extrapolated specific gravities versus carbon
numbers are shown in Figure 7-4 for theoil mixture of Table 4-7. For the
SGA = (1.21 + 0.008 (CN - 6)) SGp (7-19) boiling points, the values of Katz and Firoozabadi (1978) are used, up to
C45. When a logarithmic distribution according to Equation 7-13 is assumed
where CN is the carbon number. In those cases where no PNA analysis is for the TBP residue, the following relation is used for the boiling points (in
available, the PNA distribution of a given TBP fraction is estimated. The K) of the pseudocomponents of the TBP residue:
refractive index, n, of each TBP fraction is calculated from the specific grav-
ity, the normal boiling point, and the molecular weight, using the correla- TB = 97.58 MWO3323SG004609 (7-31)
tions of Riazi and Daubert (1980b):
When, for oil mixtures, the characterization is based on a TBP distillation
n= (~)O.5 (7-20) curve, 6 K boiling-point intervals per carbon number are assumed for the
1- 1 C45 + -components. The following expression is used for the molecular
weights of the components of the TBP residue:
1 is a characterization factor, which is found from one of the following corre-
lations: MW = - 4.0 + 14.0 CN (7-32)
 ~::~
~I\''''
122 Propertes oj Ols and Natural Gases Characterzaton Procedures 123/!r'-;'
f :"
t ,,(.

! \"

1.1 I i tion 7-10. Whitson uses his characterization procedure with the Peng- Robn- , ' ' ..
son equation of state (see Equation 5-8). Other examples of characterization J t\lXllJ f-j. ; o

- procedures are given by Edmister (1988). .. I

1.05

-
;

- - Pedersen et al. and Whitson only use the distribution functions as appro-. .

- --
E
+-' 1.0
priate procedures for selecting pseudo-components. Using continuous ther- \,;

-
ro

-
1..'-
modynamics (see e.g., Cotterman and Prausnitz, 1985), the distribution '-.
.95
/ functions are maintained in the phase equilibrium calculations. The molar
~ /'
composition is described by a molar distribution function in some variable,
+-' ,,/
'" .9 ./ .: Measured densities 1, often the molecular weight or the boiling point. The molar fractions of a
"" u
E
./ continuous mixture must add up to 1:
'<, Estimated densities
~ .85 .fIII'

c-,
J 1 F(I)dl = 1 (7-35)

+-'
v>
.8
e
Q!
el
For a naturally occurring petroleum mixture, non-hydrocarbons, and the
hydrocarbons e1-e4, are handled as discrete components. If the total mole
. 75

fraction of the continuous components is r and there are n discrete compo-

.7
t 10 20 30 40 50 60 70 80
nents, Equation 7-35 is changed to:
Carbon Number
n

Figure 7-4. Measured and estimated (Pedersen et al., 1984) C7+-densities for
i
1:
=
z, + r J
1 1
F(l)dl = 1 (7-36)
the oil mixture of Table 4-7.

where z, is the molar fraction of the -th discrete component. For two phases
Pedersen et al. use binary interaction coefficients equal to zero for interac- in equilibrium, with a molar vapor fraction of t the following material bal-
tions between two hydrocarbons. For interactions between an unlike pair of ances can be set up for the discrete components:
non-hydrocarbons, and between a non-hydrocarbon and hydrocarbons
ranging from el to e6, the interaction coefficients of Tahle 5-2 are used. For z, = ~Yi + (1 - ~) x, (7-37)
interactions between a e7 + -component and a non-hydrocarbon, the kij"val-
ues given in Table 5-2 for e6 are used. where Yiand x, are the mole fractions of component i in the vapor and liquid
Whitson (1983) assumes that the Watson characterization factor: phases, respectively. Similarly, for the continuous components:

K = TB1/3/se (7-33) rFFF (1) = ~rv FV (1) + (1 - ~)rL FL (1) (7-38)

is a constant for each e7 + fraction, and that it can be related to the molecu- F refers to the feed, V to the vapor phase, and L to the liquid phase. Cotter-
lar weight using the following formula: man et al. (1985) use continuous thermodynamics with the SRK-equation of
state (Equation 5-7). From C, on, the components are treated as three con-
K = 4.5579 MW015178 se - 084573 (7-34) tinuous distributions in molecular weight: one for the paraffins, one for the
naphthenes, and one for the aromatics. The a and b parameters entering
TB is in R, and se in 60F/60F. into the SRK-equation are, for the continuous distributions, determined
It is necessary to choose K such that the specific gravity of the total e7 +- from the molecular weight. Cotterrnan et al. have shown that when the
fraction equals the measured value. The boiling points of the e7 + -samples SRK-equation is used, the distribution functions for the two phases in equi-
may then be calculated using the definition of the Watson characterization librium can be related to one another. For example, if a gamma distribution
factor. Te and P, are found from Equations 7-11 and 7-12, and w from Equa- describes the compositon of one phase, the composition of the other phase is
124 Properties of Oils and Natural Gases Characterization Procedures 125

also described by a gamma distribution. One major drawback of this Feed: 40% CO2, 60% paraffins (continuous)
method is that the total mixture is not described by a gamma function, sim-
ply because the sum of two gamma functions is generally not a gamma func- 0.012 I I I I i i i i i

tion. Therefore, the material balance cannot be satisfied, and the flash
problem can only be solved approximately. Cotterman et al. suggest an ap- P 50 bar -- specified
T 512 K
proximate solution, where the total mixture is also described by a gamma sum of liquid
0.008
distribution. The parameters are chosen so that the following equation is and vapor
satisfied for r = 0, r = 1 and r = 2:

l1FMt = ~l1vM~ + (1 - ~)l1LM~ (7-39) 0.004

e (calcd~
O I
+-'
M, is the r-th statistical moment about the origin of the distribution F(I): :J
.D
\....

M, = 1 I'F(I)dI
1 (7-40)
+-'
Vl

el O
s,
To satisfy the material balance exactly, Equation 7-39 should have been sat- '" Feed
O
::E p = 10 bar
isfied for r = 0, 1,2, ... , oo , The errors introduced in the material balance T = 422 K
for two flash calculations on a mixture of 40 % CO2 and 60 % paraffins are 0.008
shown schematically in Figure 7-5. Cotterman et al. nave also suggested an
alternate flash calculation procedure, the quadrature method, which is ac-
, (
tually a pseudocomponent procedure. For each pseudocomponent, the mo-
lecular distribution, f, is described by the following expression: 0.006 'I

1(
s

1 f(I)dI E w (I
1
=
p = 1
p) f (Ip) (7-41)
o
O 100 150 200
s is the number of quadrature points, w is a weighting function, and Ip is a Molecular Weight
quadrature point. Equation 7-36 is then changed to:
Error in material balance indicated by shaded area
n m s

E z, + E l1k E
i=l k=l p=l
wk(Ip)fk(Ip) = 1 (7-42)
Figure 7-5. Examples of the errors in the material balance when using continu-
ous thermodynamics (Cotterman and Prausnitz, 1985).

The continuous distribution has thus been replaced by m pseudocornpo- a procedure whereby all hydrocarbon segments of the C7 + - fractions are
nents. given approximately equal importance. This is accomplished via a grouping
of carbon number fractions on weight basis. The C7 + -fraction is divided
into three (or more) groups that, by weight, are of approximately equal size.
GROUPING OF PSEUDOCOMPONENTS The group critical properties are computed as weight-mean values. Two ex-
amples of grouping are given in Tables 7-3 and 7-4, with analytical data in
The characterized mixture may consist of more than 80 components or Table 7-2. It is seen that N2, CO2, and CH4 (Cl) are given separate status,
pseudocomponents. Obviously, for extensive process calculations, it is neces- ethane and propane are grouped together (C2-C3), and so are the C/s, Cs's,
sary to drastically reduce this number. Pedersen et al. (1985) have developed and the Ca's.
126 Properties oj Oils and Natural Gases eharacterization Procedures 127

Table 7-2 Table 7-3

Analytical Data for North Sea Gas Condensate Characterization and Grouping for North Sea Gas Condensate
(Analytical Data in Table 7-2)
PNA-Distribution (Mole %) Input to Flash Program
Density
(g/cm3) at Molecular
Components Mole % Weight % Te(K) Pe (atm) w
Components Mole % 15C, 1 atm Weight P N A
N2 0.64 0.63 126.2 33.5 0.040
N2 0.64
CO2 9.16 CO2 9.16 4.19 304.2 72.8 0.225
C, 68.80
45.4
8.43 C, 68.80 38.83 190.6 0.008
C2
C3 5.11 C2-C3 13.54 16.85 335.7 45.2 0.123
ic, 0.81
C,-C6 3.94 9.20 450.7 34.3 0.225
n-C. 1.45
i-C5 0.52 C,-C1O 2.30 8.57 565.0 31.6 0.408
n-C5 0.53
C11-C'5 0.94 5.13 653.3 25.1 0.656
C6 0.63
C7 0.83 0.741 96 0.50 0.42 0.08 C,6-C'8 0.68 6.60 766.2 14.8 1.085
C8 0.95 0.780 107 0.45 0.38 0.17
Source: Pedersen el al. (1985).
C9 0.52 0.807 121 0.48 0.27 0.25
C10 0.26 0.819 134 0.47 0.30 0.23
C11 0.20 0.810 147 0.56 0.27 0.17
C'2 0.17 0.828 161 0.55 0.24 0.21 Table 7-4
C'3 0.16 0.849 175 0.54 0.22 0.24 Results of the Characterization and Grouping of the Oil Mixture from
C'4 0.15 0.857 190 0.49 0.27 0.24 Data Shown in Table 4-7
C'5 0.11 0.868 206 0.52 0.20 0.28 Input to Flash Program
C'6 0.086 0.872 222 0.55 0.19 0.26
C17 0.078 0.859 237 0.57 0.20 0.23
Components Mole % Weight % Te(K) Pe (atm) w
C'8 0.068 0.854 251 C,8.O.70 0.11 0.19
C'9 0.050 0.866 263 N2 0.34 0.10 126.2 33.5 0.040
C2o.339
C20
C21
0.046
0.035
0.873
0.876 .. CO2 0.84 0.41 304.2 72.8 0.225

C22 0.025 0.876 C, 49.13 8.79 190.6 45.4 0.008

C23 0.034 0.875
C2-C3 10.76 4.30 338.2 45.0 0.126
C24 0.023 0.877
C25 0.017 0.876 C,-C6 7.35 5.76 464.5 33.2 0.242
C26 0.018 0.878
C7-C'5 19.69 30.20 619.4 25.3 0.581
C27 0.014 0.882
C28 0.012 0.886 C,6-C26 7.75 23.85 760.2 15.0 1.046
C29 0.013 0.889
C27-C76 4.14 26.58 925.6 11.3 1.585
C30 0.047 0.908

fixed values, For C7 + -components the procedure suggested by Kesler and

CALCULATION OF THE IDEAL GAS HEAT CAPACITY Lee (1976) may be used (the calculated heat capacity is in Btu/lb):

As described in the SRK equation of state section in Chapter 5, the ideal C = - 0.33886 + 0.02827K - 0.26105CF + 0.59332wCF (7-43)
gas heat capacity is calculated from a third-degree polynomial in the tem-
perature (see Equation 5-36). For the lighter petroleum mixture constituents C2 = - (0.9291 - L1543K + 0.0368K2) 10-4
C7), the coefficients C-C4 of the temperature polynomial are attributed + CF(4.56 - 9.48w)10-4 (7-44)
128 Properties oi Oils and Natural Gases Characterization Procedures 129

C3 = - 1.6658 X 10-7 + CF(0.536 - 0.6828w) 10-7 (7-45) Sim, W. J. and Daubert, T. E., "Prediction of Vapor-Liquid Equilibria of
Undefined Mixtures,' Ind. Eng. Chem. Process Des. Dev., 19, 1980, pp.
386-393.
C4 = O (7-46)
Thomassen, P., Pedersen, K. S., and Fredenslund, Aa. "Adjustment of C7 +_
where CF = ((12.8 - K)(lO - K)/(lOw))2 (7-47) Molecular Weights in the Characterization of Petroleum Mixtures Con-
taining Heavy Hydrocarbons,' SEP Report 8617, Instituttet for Kerni-
and K is the Watson characterization factor defined in Equation 7-33. teknik, The Technical University of Denmark, 1986.
Whitson, C. H., "Characterizing Hydrocarbon Plus Fractions,' SPE [our-
nal, 23, 1983, pp. 683-694.
Winn, F. W., "Physical Properties by Nornogram,' Peto Refiner, 36, 1957,
REFERENCES pp. 157-159.
Yarborough, L. "Application of a Generalized Equation of State to Petro-
Cavett, R. H., "Physcal Data for Distillation Calculation, Vapor-Liquid leum Reservoir Fluds,' paper presented at the 176th National Meeting of
Equilibria," 27th Midyear Meeting, API Division of Refining, San Fran- the American Chemical Society, Miami Beach, FL, 1978.
cisco, CA, 'May 15, 1964.
Cotterman, R. L. and Prausnitz, J. M., "Flash Calculations for Continous
or Semicontinous Mixtures Using an Equation of State," Ind. Eng. Chem.
Process Des. Dev., 24, 1985, pp. 434-443.
Edmister, W. C., "Appled Hydrocarbon Thermodynamics. Part 4. Com-
pressibility Factors and Equations of State," Peto Refiner, 37, 1958, pp.
173-179.
Edmister, W. C., Applied Hydrocarbon Thermodynamics, Volume 2, Hous-
ton, Gulf Publishing Company, 1988.
Katz, D. L. and Firoozabadi, A., "Predicting Phase Behavior of Conden-
sate/Crude-Oil Systems Using Methane Interaction Coeffcients," f. Peto
Technol., 20, 1978, pp. 1649-1655.
Kesler, M. G. and Lee, B. l., "Improve Prediction of Enthalpy of Frac-
tions," Hydrocarbon Processing, 55, 1976, pp. 153-158.
Nes, K. and Westerns, H. A. van, Aspects of the Consttution of Mineral
Oils, Elsevier, New York, 1951.
Pedersen, K. S., Thomassen, P, and Fredenslund, Aa., "Thermodynamcs
of Petroleum Mixtures Containing Heavy Hydrocarbons. 1. Phase Enve-
lope Calculations by Use of the Soave-Redlich-Kwong Equation of State,'
Ind. Eng. Chem. Process Des. Dev., 23, 1984, pp. 163-170.
Pedersen, K. S., Thomassen, P., and Fredenslund, Aa., "Therrnodynamcs
of Petroleum Mixtures Containing Heavy Hydrocarbons. 3. Efficient
Flash Calculation Procedures Using the SRK Equation of State,' Ind.
Eng. Chem. Process Des. Dev., 24, 1985, pp. 948-954.
Riazi, M. R. and Daubert, T. E., "Simplfy Property Predctions,' Hydro-
carbon Processing, 59, 1980a, pp. 115-116.
Riazi, M. R. and Daubert, T. E., "Predicton of the Composition of Petro-
leum Fractions,' Ind. Eng. Chem. Process Des. Dev., 19, 1980b, pp.
289-294.
 Simulation of PVT-Experiments 131

Table 8-1
Experimental Constant Mass Expansion Results tor 'Mixture 1
(Molar Composition In Table 4-2) at T = 155C
(The Results are Plotted Together with Calculated Results in Figure 8-1)
Chapter 8
L1quid Vol.
Simulation of PVT- Experiments Pressure (% of Compressibility
(bar) Dew Pt. Vol.) Factor
597.1 1.3729
577.8 1.3450
The simulation results of this chapter are all based on the Soave-Redlich- 560.9 1.3208
540.5 1.2902
Kwong equation of state (Chapter 5) and the characterization procedure of
519.5 1.2596
Pedersen et al. described in Chapter 7. The procedure for adjusting the mo- 495.1 1.2244
lecular weight used for some mixtures is described in Chapter 10. AlI the 479.8 1.2027
density calculations of this chapter have been performed using the proce- 462.7 1.1779
dure of Peneloux et al. described in Chapter 5. The compositions of the mix- 449.9 1.1589
434.8 1.1367
tures referred to are shown in Chapter 4.
412.0 1.1043
393.0 1.0793
*388.0 0.00 1.0740
CONSTANT MASS EXPANSION 385.1 0.05
368.6 0.75
345.1 2.43
Table 8-1 shows the results of a constant mass expansion study (described
320.7 4.52
in Chapter 3) at T = 155C on mixture l. The results comprise saturation 300.5 - 6.11
point, liquid dropout curve, and single-phase compressibility factors. The 278.2 7.75
liquid dropout curve is plotted in Figure 8-1, together with calculated re- 255.6 9.06
sults. The results of using 6 and 20 hydrocarbon fractions to represent the 238.6 9.89
229.3 10.29
mixture are shown. With 20 hydrocarbon fractions twq, different kinds of 206.7 11.03
results are shown: in one case the measured molar composition is used, and 183.7 11.58
in the other case the molar cornposition has been changed corresponding to 161.3 11.80
a decrease in the molecular weight of the plus-fraction (C20 +) of 5.4 % . 146.2 11.89
With this adjustment, agreement with the measured dew-point pressure is Dew point
obtained. When adjusting the C2o+-molecular weight, the weight composi-
tion of Table 4-2 was maintained. A molecular weight adjustment was also which the characterized mixture contains 10 hydrocarbon fractions, and the
carried out when only 6 hydrocarbon fractions were used to represent the flash is performed using the kij = O flash versin described in Chapter 6.
mixture. It is seen from Figure 8-1 that too much liquid precipitation is pre- The calculated and the simulated results agree very well.
dicted in the pressure range of 250-400 bar, whereas the measured and the
calculated results agree well at lower pressures. To represent this mixture, it
matters little whether 6 or 20 hydrocarbon fraetons are used. Also it makes DIFFERENTIAL LIBERATION
little difference whether the molecular weight of the plus-fraction is ad-
justed or noto
Table 8-3 shows the results of a differentialliberation study (described in
Table 8-2 shows constant mass expansion results at T = 71.6C for mix-
Chapter 3) at T = 92.8C on mixture 2. The results comprise saturation
ture 3. The data include the bubble-point pressure, and the volumes at vari-
pressure, oil formation volume factors, gas/oil ratios, den sities of the oil
ous pressure stages relative to the volume at the bubble point. The results are
phase, and gas-phase compressibility factors. The results for the oil forma-
plotted in Figure 8-2. Also shown in Figure 8-2 are simulated results in

130
132 Propertes 01 Oils and Natural Gases Simulation 01 PVT-Experments 133

15.0 i Table 8-2

Experimental Constant Mass Expansion Results tor Mixture 3
Q) (Molar Composition in Table 4-4) at T = 71.6C
E
:>
(The Results are Plotted Together with Calculated Results in Figure 8-2)
o
e>
es' __
+-'
c:: Pressure Volume/Volume
2!. 10.0 (bar) at Sato Pt.
~Q)
C> 401.7 0.9811
4-
o 380.1 0.9836
.. 362.5 0.9853

Q)
343.8
325.3
0.9873
0.9895
E
:> 307.4 0.9915
o
:>
5.0
Experiment 291.4 0.9934
"O
20 CH-fractions 274.9 0.9955
:>
CT

--'
20 CH-fractions. MW adjusted 254.3 0.9981
6 CH-fractions. MW adjusted '239.0 1.0000
235.0 1.0033
215~1 1.0230
0.0
182.4 1.0678
0.0 100.0 200.0 300.0 '00.0
142.8 1.1651
Pressure (bar) 1.2946
113.3
Figure 8-1. Liquid dropout curvos lor mixture 1 at a constant mass expansion 91.7 1.4568
study at T = 155.0C. The experimental results are shown in Table 8-1. The den- 77.1 1.6193
sity is calculated using the procedure 01 Peneloux et al. (see Chapter 5). 65.7 1.8158
Saturalionpolnt

\
2.0 I i
Table 8-3
Experimental Dlfferen.tial Liberation Results tor Mixture 2

.
Experiment
(Molar Composition in Table 4-3) at T = 92.8C
10 CH-fractions. kij = o
+-'
c::
(The Results lor the Oil Formation Factor and the GOR are
'C; 1.5 Plotted Together with Calculated Results in Figures 8-3 and 8-4)
e,
c::

--
o

~---
+-' Di! Compro
-o
'-:> Pressure 011 Form Solution Density Factor
+-' Vol. Fact. GOR*'
(bar) (g/cm3) Z
'"
V1 1.0
.., 0.718
*267.9 1.396 132.9
'"
Q)
E
:> 239.6 1.365 119.4 0.726 0.889
o
2:. 209.4 1.329 103.7 0.735 0.888
~ ,.5
179.8 1.297 89.6 0.744 0.876
o
:>
150.0 1.265 75.7 0.753 0.880

112.9 1.227 58.6 0.765 0.894

0.0 100.0 200.0 300.0 '00.0 500.0 72.4 1.186 40.4 0.778 0.904
Pressure (bar)
33.1 1.144 21.9 0.791 0.949
Figure 8-2. Results lor the relative volumes 01 mixture 3 (see Table 4-4) in a con-
1.0 1.067 0.816
stant mass expansion study at T = 71.6C. The experimental results are shown
in Table 8-2. The density is calculated using the procedure 01 Peneloux et al. (see Salurationpoinl
Chapter 5). Std, m3 gas par m3 01 residualoilal 15C, 1 bar.
 . f) ~ <,
. ~J!...I\I(i
134 Propertes 01 Oils and Natural Gases
Simulation 01 PVT-Experiments 135 /;/ --,.~()~
<, - ("
-o
2.0 I .
tion factor are plotted in Figure 8-3 and the. results for the gas/oil ratio in ~ (BllUOlt,;j/ ;.'
Figure 8-4. Also shown in Figures 8-3 and 8-4 are the calculated results of-=>\ ;~
'"
::J
using 6 and 20 hydrocarbons to represent the mixture. Two different kinds or,. -- ti' P
"O

'"
Ql
results are shown, produced with the 20 hydrocarbon fractions. One corre- ",-'" ",flr)
'-
<'")
~ sponds to the measured composition. The other one is obtained by treating ~~,
o ~ --- - -, the molecular weight of the plus-fraction as an adjustable parameter to
which is attributed the value that makes the measured and the calculated
""'E
'-o saturation points agree. The latter procedure was also used when the mix-
1.0
+'
u
Experiment ture was represented using only 6 hydrocarbon fractions. The experimental
"-'" 20 CH-fractions and the calculated results agree very well. The best results are, in fact, ob-
Ql
E
::J
20 CH-fractions. MW adjusted tained when the mixture is characterized using only 6 hydrocarbon frac-
,.
o 6 CH-fractions. MW adjusted tions.
e;
o Table 8-4 shows, for mixture 4, the compositions of the gas phase libera-
+'
ro ted at each stage of a differentialliberation study at T = 141C. The results
E
'-
.l' are shown schematically in Figure 8-5. Also shown in Figure 8-5 are the cal-
I
culated results where the characterized mixture comprised 6 hydrocarbon
o o. 01 , , ! I ,

0.0 100.0 200.0 300.0 fractions. The measured and the calculated results agree very well.
Pressure (bar)
Figure 8-3. Results for the oil lormation volume factor 01 mixture 2 (see Table Table 8-4
4-3) in a differentialliberation study at T = 92.8C. The experimental results are
Experimental Results of Differential Liberation Study on Mixture 4
shown in Table 8-3. The density is calculated using the procedure 01 Peneloux et
al. (see Chapter 5). (Molar Composition in Table 4-5) at T =
141 o C.
Composition of Differentialiy Liberated Gas (Mole %)
200.0 I (The Results are Plotteq Together with Calculated Results in Figure 8-5)

P (bar) 317.3 292.4 249.3 214.9 174.5 138.4 91.4 49.3 16.9 1.0

N2 1.06 1.06 1.06 1.05 0.95 0.78 0.54 0.27 0.07 0.01
CO2 3.85 3.87 3.97 4.12 4.23 4.28 4.57 4.99 4.46 1.21
C, 71.31 71.93 73.67 75.72 75.64 75.01 71.39 64.44 41.40 7.15
o fiF
-;:/.;:- -' C2 9.01 9.00 9.13 9.37 9.58 10.27 11.58 14.75 20.13 9.39
-o
::J C3 5.55 5.50 5.42 5.32 5.32 5.76 6.79 9.33 18.31 14.93
"O
-;?''-'
-;?'/ / -C. 0.88 0.88 0.84 0.79 0.77 0.82 0.97 1.33 3.01 3.45
'"<l>
L 100.0
~ .~/.-. n-C. 1.92 1.91 1.79 1.61 1.56 1.67 2.05 2.70 6.72 9.34
.-o
<'")
E.
;'l".:::::-A
ic, 0.59 0.59 0.54 0.43 0.42 0.43 0.54 0.66 1.79 2.29
+' '" c, 0.74 0.75 0.66 0.50 0.49 0.49 0.66 0.75 2.11 6.39
'" '"
IX

-<'")
O> ~.~:;.....~"'" Cs 0.68 0.68 0.57 0.35 0.36 0.30 0.47 0.45 1.24 9.27
.-
o
E
.. ~ ....
/
Experiment
C7 0.84 0.79 0.60 0.25 0.28 0.16 0.33 0.26 0.65 14.80
"....
"O

'" +' Ca 0.80 0.70 0.45 0.14 0.14 0.03 0.10 0.07 0.11 12.79
'"'" -'" 20 CH-fractions C9 0.49 0.42 0.24 0.06 0.05 0.00 0.01 0.00 0.00 5.44
20 CH-fractions. MW adjusted ClO+ 2.27 1.92 1.06 0.29 0.21 0.00 0.00 0.00 0.00 3.55
6 CH-fractions. MW adjusted
0.0
0.0 100.0 200.0
Pressure (bar)
300
eONSTANT VOLUME DEPLETION

Figure 8-4. Results tor the gas/oil ratio 01 mixture 2 (see Table 4-3) in a differen-
tial liberation study at T = 92.8C. The experimental results are shown in Table Table 8-5 displays the liquid dropout curve, and the gas phase compress-
8-3. The density is calculated using the procedure 01Peneloux et al. (see Chapter 5). ibility factors, measured at a constant volume depletion experiment (de-
136 Propertes 01 Oils and Natural Gases Simula tia n 01 PVT-Experiments 137

100.0 I I
10.0 I I

-- .
ClJ

. . .--.
E
::J

.. 75.0
~
..
7.5
ClJ
e
o
E o
o..
e Measured mole % of C,
"
. ,
ClJ
e Measured mole % of C2-C6 Cl
.....
..o Measured mole% of C7+ 4-
,..,
....-
-, <,
...
o 5.0 ,..,

..'"
\...

,
.
e ----- Calculated results. 6 CH-fractions
ClJ
u \
e
o
u
ll!
::J
\
o....... 0--0
o \
"-.....~O----o----o---o :> 2.? \

.
"O

::J \
O"
\
\
_6--.
100.0
~
200.0 300.0 '00.0 0.0 0.0 100.0 200.0
"" 300.0 400.0
Pressure (bar)
Pressure (bar)
Figure 8-5. Results of a differential liberation study on mixture 4 (see Table 4-5) Figure 8-6. Liquid dropout curves for mixture 9 in a constant volume depletion
at T = 141C. Hydrocarbon concentrations in the liberated gas. The experimen- study at T = 82.5C. The experimental results are shown in Table 8-5. The den-
tal results are shown in Table 8-4. sity is calculated using the procedure of Peneloux et al. (see Chapter 5).
Table 8-5
Experimental Results of a Constant Volume Depletion Study
at T = 82.5C on Mlxture 9 (Molar Compositlon in Table 4-11) Table 8-6
(The Liquid Volume is in Percent of the 8aturation Point Volume. Experimental Results of a Constant Volume Depletion Study
The Liquid Dropout Curve is Plotted in Figure 8-6 Together with =
at T 82.5C on Mixtute 9 (Molar Composition In Table 4-11)
Calculated Results) (Molar Compositions of the Gas Mixtures Removed at Each Depletion 8tage)

Gas Phase Cumulative P (bar) 235 180 125 84 48

Pressure Volume % Compressibility Mole Fraction N2 0.08 0.08 0.09 0.09 0.08
(bar) Liquid Factor Removed CO2 2.45 2.46 2.47 2.49 2.52
281 0.0 0.893 0.0000 C, 82.97 83.67 84.28 84.35 83.88
C2 5.77 5.77 5.79 5.84 5.94
235 1.3 0.852 0.1191 C3 2.82 2.78 2.77 2.82 2.96
180 3.4 0.836 0.3077 -C, 0.54 0.53 0.52 0.54 0.57
n-C, 1.19 1.15 1.13 1.16 1.22
125 4.6 0.858 0.5205 te, 0.49 0.46 0.45 0.46 0.49
84 4.7 0.893 0.6841 e, 0.56 0.52 0.50 0.51 0.56
C6 0.65 0.57 0.53 0.53 0.58
49 4.4 0.932 0.8121 2.49 1.47
C7+ 2.01 1.21 1.20

scribed in Chapter 3) on mixture 9 at T = 82.5C. The liquid dropout curve with the measured dew point. The calculations were performed using 20
is plotted in Figure 8-6. Also shown in Figure 8-6 are calculated liquid drop- hydrocarbon fractions. The measured and the calculated results agree rea-
out curves, where the measured C7 vmolecular weight is used, and where sonably well after the adjustment of the C7 + -molecular weight. Table 8-6
the C7+-molecular weight has been decreased by 8.8%, to get agreement shows the measured molar composition of the gas removed at each depletion
138 Properties oj Oils and Natural Gases Simulation Di PVT-Experiments 139

100.0 .,--------------------------------,
stage. The results are illustrated in Figure 8-7, where the calculated results

...-------.-- - - . .. .. .
are also shown. It is seen from Figure 8-7 that it makes ver y little difference
for the gas-phase compositions whether the molecular weight of the C7 +-

.. 75.0
fraction is adjusted or noto Measured and calculated results for the cumula-
tive mole fraction removed at each depletion stage are shown in Table 8-5 .
'o" Measured mole % of C1 These results are plotted in Figure 8-8 together with the calc~lated results.
E

~ o Measured mole % of C2-C6

c: Measured mole % of C7+ Table 8-7
:!
+-'
50.0
Calculated results, measured MW
Molar Compositions of Separator Fluids of Mixture 5
.'".. (Molar Composition in Table 4-6)
+-' Calculated results, adjusted MW
e
'"
v
e Stage 1:T = 44C. P = 67.2 bar
o
u
25 ..0 Gas Liquid

Measured Calculated Measured Calculated

~ o -D o o Component Mole % Mole % % Dev. Mole % Mole % % Dev.

N2 0.13 0.14 - 0.01 0.01 -

0.00.0 50.0 100.0 150.0 200.0 250.0 300.0 CO2 2.65 2.69 1.5 1.35 1.24 - 8.1
Pressure (bar) C, 84.50 85.15 0.8 18.84 21.62 14.8
Figure 8-7. Results 01 a constant volume depletion study on mixture 9 (see Table C2 7.64 7.52 - 1.6 6.14 7.09 15.5
4-11) at T = 82.5C. Hydrocarbon concentrations in the removed gas. The ex-
perimental results are shown in Table 8-6. C3 2.83 2.63 - 7.1 5.12 6.19 20.9

C. 1.32 1.14 - 13.6 5.18 5.98 15.4

1.0 - 0.42
C5 0.53 - 20.8 5.04 5.27 4.6

C6 0.20 0.16 - 4.91 4.78 - 2.6

Experiment C7+ 0.20 0.15 - 53.41 47.82 - 10.5

"
'E";
E
.75 Calculated using~ measured MW Stage 2:T = 1SoC, P = 1 bar
a: '" Calculated using adjusted MW
Gas Liquid
c:
o
+-' Measured Calculated Measured Calculated
v
.... re
Component Mole % Mole % % Dev. Mole % Mole % % Dev.
u- .5

'" N2 0.01 0.03 - 0.00 0.00 -

o
:>:
CO2 3.72 3.03 - 18.5 0.00 0.03 -
'>"
+-' C, 51.92 53.67 3.3 0.00 0.28 -
'"
~ .25
=>
C2 16.60 16.82 1.3 0.18 0.60 -
u
C3 12.44 12.75 2.5 0.96 1.82 89.6

C. 8.64 8.65 0.1 :3.21 4.20 30.8

I ! I ! '-. ! .......! I
- 9.4 6.42
0.0 C5 3.92 3.55 5.67 13.2
0.0 50.0 100.0 150.0 200.0 250.0 300.0 350.0 '00.0
C6 1.44 0.99 - 31.3 6.89 7.31 6.1
Pressure (bar)
Figure 8-8. Cumulative mole Iractions 01 mixture 9 (molar composition in Table C7+ 1.30 0.51 - 60.8 83.09 79.34 - 4.5
4-11 removed as a lunction 01 pressure in a constant volume depletion study at % Dev. = 100 x (Calculated - Measured)/Measured.
T = 82.5C. The experimental results are shown in Table 8-5. (Only stated,for measured mole %'5 above 0.5.)
140 Properties of Oils and Natural Gases
An almost-perfect agreement with the experimental results is observed,
when the molecular weight of the plus-fraction is adjusted to match the
measured dew point.
SEPARATOR TEST
Table 8-7 shows the results of a two-stage separator test (described in
Chapter 3) on mixture 5. The results comprise measured and calculated gas
and liquid-phase compositions. The characterized mixture consisted of 10
hydrocarbon fractions. The measured and calculated gas-phase composi-
tions agree very well, whereas larger deviations are seen for the liquid
phases.
SUMMARY OF THE RESULTS OF THE PVT-SIMULATIONS
The PVT-simulations provide evidence that good agreement with mea-
sured PVT-results can be obtained using the SRK equation of state in con-
junction with the C7 + -characterizaton procedure-of Pedersen et al. (pre-
sented in Chapter 7). Six hydrocarbon fractions are, in most cases, sufficient
to represent the hydrocarbons of a reservoir fluido The procedure described
in Chapter 10, for adjusting the molecular weight of a plus fraction against
a measured saturation point, is ver y useful for obtaining reasonable results
for the gas/liquid ratios of a gas condensate mixture. An adjustment of less
than 10% of the molecular weight of the plus fraction does not influence the
phase properties significantly.
Chapter 9
Comparison Between Experimental
and Predicted Thermodynamic
Properties
The calculated results of this chapter are all based on the flash and phase
envelope algorithms"of Michelsen described in Chapter 6 and on the SRK
equation of state (Chapter 5) coupled with the C7 + -characterization proce-
dure of Pedersen et al. (see Chapter 7). The procedure, used in some cases
for adjusting the molecular weight of a plus-fraction, is described in Chap-
ter 10. The compositions of the mixtures mentioned are shown in Chapter 4.
OEW ANO BUBBLE POINTS (PHASE ENVELOPES)
Table 9-1 presents measured and calculated dew and bubble points for
mixtures 1-10. Calculated results, obtained using the standard flash (non-
zero binary interaction coefficients for interactions where non-hydrocar-
bons take part), and 20 hydrocarbon and 6 hydrocarbon fractions, respec-
tively, to represent the mixture, are shown. The simplified two-phase
(P,T)-flash (ki = O), described in Chapter 6, is also used to yield results. The
last column of Table 9-1 shows the needed adjustment to the molecular
weight of the plus-Iracton (maintaining the measured weight composition),
to get agreement with the measured saturation (dew or bubble) point. Re-
sults of using a characterized mixture consisting of 6 and 20 hydrocarbon
fractions are of comparable quality. Generally, it makes very little difference
whether the standard flash or the ki = O flash is used. For all mixtures ex-
cept mixture 7, an adjustment of less than 10% of the molecular weight of
the plus-fraction is sufficient to get agreement with the measured saturation
point. In Figure 9-1 the calculated phase envelope of mixture 10 is shown
together with four measured saturation points. The characterized mixture
consists of 20 hydrocarbon fractions. Figure 9-2 shows, for mixture 3, the
results of three different phase envelope calculations corresponding to the
results in Table 9-1.
141
 Table 9-1
Measured and Calculated Saturation Pressures for 10 Different Petroleum Mixtures
(The Molar Compositions are Given in Chapter 4)

Calculated Calculated Calculated

Measured p, with Psa' with p, with MW-
Mixture Temperature
Poo' 20 CH 6 CH kij = O adjustment
No. (0C) (bar) (bar) % Dev. (bar) % Dev. (bar) % Dev. (%)
1 155.0 388.0 398.3 2.7 386.4 - 0.4 392.4 1.1 - 5.4
2 92.8 267.9 259.5 - 3.1 263.5 - 1.6 251.4 - 6.2 9.1
3 71.6 239.0 235.9 - 1.3 243.6 1.9 233.1 - 2.5 2.6
4 141.0 340.3 346.5 1.8 347.2 2.0 339.5 - 0.2 - 8.3
5 129.0 464.0 463.7 - 0.1 458.5 - 1.2 457.6 -1.4 0.1
6 93.3 274.5 276.5 0.7 282.1 2.8 274.2 - 0.1 - 3.0
7 93.4 264.7 248.0 - 6.3 252.1 - 4.8 248.0 - 6.3 18.5
8 93.4 264.7 279.6 5.6 272.9 3.1 279.6 5.6 9.6
9 82.5 281.0 334.5 19.0 312.2 11.1 330.5 17.5 - 8.8
10 32.2 462.8 435.8 - 5.8 415.3 - 10.3 405.1 - 12.5 7.6
10 67.8 401.6 412.6 2.7 387.2 - 3.6 390.1 - 2.9 - 2.4
10 125.4 303.5 337.3 11.1 303.5 0.0 324.3 6.9 - 6.0
10 169.4 229.6 246.2 6.7 198.6 - 13.5 238.1 3.7 - 2.2
%AAD
% BIAS
-
-
-
-
-
- . 5.0
2.6
-
-
4.3
- 1.1
-
-
5.1
0.2
6.4
0.9
Notetion:
CH: Number of hydrocarbon fractions in the characterized mixture.
kij = o: The kij = O flash described in Chapter 6 has been used on a characterized mixture consisting of 20 CH-fractions.
% Dev.: 100 x (Calculated - Measured)/Measured.
MW-adjustment: The adjustment needed of the molecular weight of the plus-fraction (se e Chapter 10) to get agreement with the measured saturation point using 20
CH-fractions and standard k-values (see Chapter 7).
% AAD: % Average absol ute deviation.
% BIAS: % Average deviation.

Pressure (bar)
Pressure (bar)
,~ o g o g N g o
o
N
o
o
o
o o o
gr-~~~ -;o ~ ~o~ o~ ~o o o
o
o

o
o

o
o

-<
ru n
Q
n
Q.I
n
Q
t'Tl
)(
W
3 -< o
'O N 'O ro
::J
a.
n
e
n
e
n
e
ro
-s
3 o
~ ~ 3
'O
ro
o r-e-
e
Q
r-+
Q
.
Q
r+ ro ;;)
w ro ro ro :::J
C)
e
-s
ro o. o.. o. r+
N N
e
ro
rt

.., o
o

Ol rlg <> <:>
ID
a.
'O
:;y
g
o
W
en
ID
ID
::J
n ",
'" x
< o 'O
ID o e
n ..,
'"
O'
'O '" 3

~ ~
rt '"
ID
en N
o
o
~ o
:3 ~
X o
o~ ~
e N

o
ID o
o
144 Propertes 01 Ols and Natural Gases Comparing Experimental and Predicted Thermodynamic Properties 145

FLASH RESULTS :> o <J: al (") al C\J Cl

CI) M ~ M
~ ~
C\J N
~ ~
l1i al
O I
C\J

~ I I I I I
Measured and calculated liquid-phase mole fractions, at T = 93.4 DC and
lO
.. e

various pressures, are shown in Table 9-2 for mixture 7. The calculations .c 'g
... CI)
r-,
were performed, using 20 and 10 hydrocarbon fractions, respectively, to ... o;XU C\J C\J al ll) ll)

~ ~
~
Cl co ll) (") o
o ~
"5

represent the mixture. Table 9-3 shows measured and calculated phase com- c:: ll) M N N M
positions at T = 93.4DC and P = 201.1 bar. The calculated results for the
N
11
s
.
~<D

U '"
v v

O- o
totalliquid fr~ction agree well with the experimental observations, whereas Q.

:> ~ v co
"C

c:: E C\J C\J ID (")

larger deviations are found for the compositions of each of the two phases. It lO
o CI) N N
~ N
~ ~ ~
<i (") ll) <Xi
U O I
C\J
makes little difference whether the characterized mixture consists of 20 or of o CI)
1/)
;f. I I I I I

10 hydrocarbon fractions. o:t '"

s:
C'? Q.
'g
01 -
Table 9-2 11 '5
~x
..!!!U <J:
r-- C\J
co .n r-,
co
-e-
(")
(")
(")
ll)
t-,
::lo ~ ~

.t---
Measured and Calculated Liquid Mole Fractions ~ CT
::: ~N ll)
v
M N N N
v
lO
at T = 93.4C of Mixture 7 U
(Molar Composition in Table 4-9) .. QI

-
'g
::J CI)

><
;; ll)
v
ID
(")
ID
co
co ll)
~
v ID
V
Calculated Calculated
:B
1/) cri <i N N ~ M al

- '" v (")

Measured Liquid Mole Liquid Mole CI)

o :!!
Pressure Liquid Mole Fraction ~ Fractlon
(bar) Fraction Using 20 CH % Dev. Using 10 CH % Dev. (")-- :>CI) ~ ~ ~ ll)
,...:
r-, (") r-,
cf, QI ~ o -r-
(") ll) l1i <Xi
201.1 0.871 0.861 - 1.1 0.855 - 1.8 O I I I I
ll)
Qlo ;f. I I
173.4 0.775 0.788 1.7 0.784 1.2 :c:::
138.9 0.681 0.704 3.4 0.701 2.9 ~:. 'g
104.4 0.606 0.625 3.1 0.624 3.0 en CI)

70.0 0.537 0.550 2.4 0.550 2.4 e o;X

U
r--
eo
'<f
r-, o
ll)
v
f'-:
(")
(")
ID
(")
ll)
v
- 0.4
o "5

N (") ~ o 6 6 o
0.475 0.477
s ..oo
35.5 0.479 - 0.8 E e ~<D al

%AAD - - 2.1 - 2.0 Q. . U '"

% BIAS - - 1.5 - 1.2 E Q.
:>CI) al ID co (") r-, ID co
o E 6 o cri l1i N <i
% Dev. is defined in Table 9-1. o o O ~
(")
C\J

CH: Hydrocarbon fractions used in the calculation.

QI

:G
U
CI)
1/)
..~ I
I I I I
I

.r:. e
o- '"
s:
Q. 'g
O
.~

-
CI)
"C

ch "X alv (")

o ll) (") r-, :;
DEN SITIES QI ..!!!U r-,
~ r-, (") (")
C\J
<Xl u
'" i:lo N M ~ 6 6 o 6
-;

The experimental gas-phase compressibility factors (of Table 8-1), are

'5
lO

u
" N
U
al u
Q)
:S
S
plotted in Figure 9-3 for mixture l. Also shown are the calculated results of ~ 'g
1:)
Q)
"C

CI) ~ en
using the SRK-equation, and the procedure of Peneloux et al. (a11density e ;; ll)
ID
r--
~ ID
ll)
o
eo
ll)
(")
co
(")
al
o '"Q) :J
en
lO 1/) <i ~ ~ _ e

'" a;
correlations in this section are described in Chapter 5). The results cover the "C
6 6 6 .c O

pressure range 400-600 bar. The SRK-equation gives the better results at the ..
QI

::J
Q)
:!!
~
~~
"<3

en 1:) e
highest pressures, whereas at lower pressures, the better results are obtained Q) O
lO
QI e e.c
with the procedure of Peneloux et al. ::: Q)
e '2 '2 ~~
o
1:)
o en ~
Figure 9-4 shows the results for the gas-phase compressibility factor of the Q.
+ + -1:)
E ::::. ~ ~:t
gas liberated, in a differentialliberation study on mixture 2. The experimen- o
U eS eS eS .s I
;f.o
+
 Comparing Experimental and Predicted Thermodynamic Properties 147
146 Properties 01 O/s and Natural Gases
tal results are listed in Table 8-3. Also shown in Figure 9-4 are calculated
1.5
results where the procedure of Peneloux et al. is used. The characterized
mixture consists of 20 and of 6 hydrocarbon fractions, respectively. With 20
Experiment
hydrocarbon fractions, results produced using the measured molar composi-
<. Peneloux .,-
o 1.4 tion are shown, and also when the molecular weight of the plus-fraction is
+'
u
ro
u,
SRK

. . ... ...
...
............

adjusted to match the measured bubble point. The results with 6 hydrocar-

. .
o-,
+' ... bon fractions are obtained with an adjusted molecular weight. Figure 9-5
...
.a 1.3 ............ shows the corresponding results for the liquid-phase density in the same dif-
...
ferential liberation experiment. For both the gas and the liquid phases it

..
...
'"'"<lJ ...
<.
o.
... makes very little difference whether 20 or 6 hydrocarbons are used in the
... """
E
o
u 1.2
... calculations or whether the molecular weight of the plus-fraction is adjusted
<lJ or noto The results of liquid density calculations using 5 different density
'"ro
s: correlations are shown by Figure 9-6. The best results are obtained with the
e,

'"
ro
procedure of Peneloux et al., with the procedure of Standing and Katz, and
C,!) 1.1
with the Alani-Kennedy equation. Somewhat larger deviations are found

using the API-procedure, but the deviations from the measured results are
I I I I 1 I
stillless than 3 to 4 %. Much larger deviations are found when the liquid-
1.0
450.0 500.0 550.0 600.0 phase density is calculated using the SRK-equation.
350.0 400.0
Pressure (bar) For mixture 9, a ver y large deviation was observed between the measured
Figure 9-3. Gas phase compressibility factor of mixture 1 (molar composition in and the calculated dew point (see Table 9-1). This deviation can be elimi-
Table 4-2) in a constant mass expansion study at T = 155.0oC. The experimental
results are shown in Table 8-1. 1.0

1.25

1.1
Experiment
.9
s, 1. 15
20 CH-fractions
o
+' 20 CH-fractions (and 6 CH-fractions), MW
u
ro Experiment adjusted.
u; 1.1
'>, 20 CH-fractions
+' .8
'"..!::.E
-
.o
1.05
______20 CH-fractions, MW adjusted
__ . - 6 CH-fractions, MW adjusted 01
~--:.;:::.
'"'" 1.0
<lJ
<.
o.
>,
+'
.7

o
E
.95
'"e<lJ

-----------'
-'--=--::.----
--. .
u
'"
~
s:
e,
ro .9
.6
'"'"
<.O .85

.8
.5 I I I ,1 I

0.0 100.0 100.0 300.0

.75 300.0
0.0 100.0 100.0 Pressure (bar)
Pressure (bar)
Figure 9-5. Results tor the density 01 the liquid phase 01 mixture 2 (molar compo-
Figure 9-4. Results lor the gas phase compressibility tactor 01 mixture 2 (molar sition in Table 4-3) in a differentialliberation study at T = 92.8C. The experimen-
composition in Table 4-3) in a differential liberation study at T = 92.8C. The ex- tal results are shown in Table 8-3. The density is calculated using the procedure
perimental results are shown in Table 8-3. The compressibility lactor is calculated 01 Peneloux et al. (see Chapter 5).
using the procedure 01 Peneloux et al. (see Chapter 5).
 Properties of Oils and Natural Gases Comparing Experimental and Predicted Thermodynamc Properties 149
148

.85 , I nated by an adjustment of the molecular weight of the ClO+-fraction of

- 8.8 %. Figure 9-7 shows measured (see Table 8-5) and calculated gas-
Experiment
phase compressibility factors of the constant volume depletion experiment
Standing-Katz
on this mixture. Calculated results are shown with and without adjustment
<-,:::::-.::.::::-.-....:
.. Peneloux

.
81-',
'." .

. ......'
APl
- - SRK
of the molecular weight. The procedure of Peneloux et al. was used in the
density calculations. Better results are obtained with the adjusted molecular
weight than with the measured molecular weight, and thus the adjustment
ro
" ...... .

.
E
~ has in this case not only ser ved to improve the dew point result, but it has
00
. 75
", "
.......... ~ ..".'.",'<:::::...
~ . .. also improved the results for the gas-phase density and, as shown in Figure
+-'
c-,

'"
e
-. ..........

" ..........
8-6, for the gas/liquid ratio .
The liquid densities of mixture 6 are shown in Table 9-4, as measured in a
Cl
Q)

'" --- differentialliberation experiment at T = 93.3C. The results are plotted in

Figure 9-8, which also illustrates calculated results produced by using five

-'- ---'- -.
-.""-.....!
.7
different density correlations. The best results are obtained with the API-

.85
I
0.0
I

50.0
I

100.0
'-'- I

150.0
I

100.0
---,"-

150.0
I
300.0
Pressure
(bar)
procedure, but none of the calculated results deviate by more than about
5 % from the measured ones.
Table 9-5 shows measured liquid phase densities for mixture 8 at
T = 93.4C and various pressures. These measured results are plotted in
Figure 9-6. Measured and calculated liquid phase densities of mixture 2 (molar
composition in Table 4-3) in a differential liberation stndy at T = 92.8C. The ex-
perimental results are shown in Table 8-3. The employed density correlations are
described in Chapter 5. Table 9-4
Measured Liquid-Phase Denslties 01 Mixture 6 in a Differential
Liberation Study at T = 93.3C.
(The Results are Shown Graphically in Figure 9-8, Together with
Calculated Results)
...
Experimental
o Calculated using measured MW
+-' Pressure (bar) Density (g/cm3)
:;;
u..
.95
__ - Calculated using adjusted MW
c-, 389.8 0.677
+-'

-
.o ~
352.6 0.670

'"'"
Q)
...
o.
E
o
u
.9
~
~
~
" "-: /
/
/ .
/
319.7

288.3

'274.5
0.665

0.661

0.660
<: <, /
~ ... ,/
-o
s: .... .....
,/ 227.0 0.680
a..
-----
/

'"
. 85 193.7 0.693
'"
<.!)
148.1 0.711

109.9 0.727
Pressure
.8 I ! I I I ,
70.6 0.742
(bar)
0.0 50.0 100.0 150.0 100.0 150.0 300.0
31.4 0.759
Figure 9-7. Gas phase compressibility factor of mixture 7 (molar composition in 1.0 0.814
Table 4-9) in a constant volume depletion experiment at T = 82.2C. The experi-
Saturationpoint
mental results are shown in Table 8-5.
150 Properties of Oils and Natural Gases Comparing Experimental and Predicted Thermodynamic Properties 151

.90 i I .75

Exper iment
. 85 Standing-Katz ,
Peneloux <,
.7
.8 Alani-Kennedy -, <,
''-'-,-
--- API
M <, <, ..............
E __SRK M <,
~ <, <,

'"
E
-..;;::: .....
u .............
CJ>
...........................
---
----~
. 75
.... ~ . .

.
>.
.., ..........
. 65
<, ..,.>, ..--.~-:---:-::::---
V> .......
e <, ..... V>
........ ..
'"
D
--- <,
.....
e
'"
Experiment ........

.
D
.7 <,
Standing-Katz ........
'<, ........
......
--- ---
- __e __ ----

_/

# - ~

-..:=-:

-===-
.6
Peneloux --- =--,

.65
-<;

--- -.... .. _ - -_.-_.-

.. ..
API
Al an i-Kennedy
SRK
.6 [ I ! !
250.0 300.0 350.0 400.0 .55 I !
0.0 50.0 100.0 150.0 200.0
0.0 50.0 100.0 150.0 200.0 250.0
Pressure (bar) Pressure (bar)

Figure 9-8. Measured and calculated liquid phase densities of mixture 6 (molar Figure 9-9. Measured and calculated liquid phase densities of mixture 8 (molar
composition in Table 4-7) in a differentialliberation study at T = 93.3C. The ex- composition in Table 4-10) at T = 93.4C. The experimental results are shown in
perimental results are shown in Table 9-4. The employed density correlations are Table 9-5. The ernployer density correlations are described in Chapter 5.
described in Chapter 5.

Table 9-5
Measured Liquid-Phase Densities 01 Mixture 8 at T = 93.4C
(The Results are Shown Graphically in Figure 9-9: Together with
Calculated Results)
ENTHALPY
Pressure (bar) Density (g/cm3)

201.1 0.6565 No published data have been Iound for the enthalpy of petroleum reser-
173.4 0.6536 voir fluids. The data presented here are for hydrocarbon boiling point frac-
tions, shown in Table 9-6 as boiling point curves (Lenoir and Hipkin, 1973).
138.9 0.6538
To use the characterization procedure of Pedersen et al. it is necessary to
104.4 0.6753
"convert' the distillation data of Table 9-6 to concentrations of carbon nurn-
70.0 0.7160 ber fractions. This is done using the data presented by Katz and Firoozabadi
35.5 0.7209 (see Chapter 7) for boiling point ranges of carbon nurnber fractions. A usual
C7 + -characterization is then carried out. The ideal gas enthalpy was calcu-
lated using the procedure of Kesler and Lee described at the end of Chapter
Figure 9-9, together with the calculated ones. Five different density correla- 7. The residual enthalpy was calculated from the SRK-equation of state us-
tions were tried. Except with the SRK-equation, all the calculated results ing Equation 5-37. To facilita te cornparisons between measured and calcu-
deviate less than 5 % frorn the measured results. The best results are, in this lated results, the enthalpies given by Lenoir and Hipkin have been adjusted
case, obtained with the procedure of Peneloux et al., and with that of Stand- to the same base level as the one used in the calculations (i.e., the ideal gas
ing and Katz. enthalpy equals O at T = 273.15 K),
 IIII,I!

152 Propertes 01 Oils and Natural Gases Comparing Experimental and Predicted Thermodynamic Properties 153

111
I 60.0

50.0

(') 1(''<)1"1"'<1"11{) I{)lr.o

.~;:;'
VI E(.) C\J11{)
//
00
c: 1{)
00
1{)
00

-
(1)- ~ ~ ~ ~ al 40.0
N COI
....
!'I en 30.0
~!
" .c
11 Cl
::J
..o
~I
<.!)
I I I I I I I I I I <lJ
o
20.0 /
ni
"" E

/
-
<,
.c .."
..::, 10.0
Oli:"
"O' =~IO ~ 1'U;<1"1'-;,<1"1g 1'g<1"I~
~ I~'!f 10~ 11{)

//
o
.... ~'5c ~
'>,
o.. 0.0
en ID '0 ro
a.. .c
1/1
..::J
CII ;g -10.0

/ Measured
Cl -20.0
I ~ i :
./ - Calculated
i .:
c:: >-
-30.0

;: =i1: '<1"10>10>10> 11{) IllJ 11{) 11{)

co f'.. f'..
o
s: .ss:"" c: (.)
(I)[j,
c_ r.o
LO CX)
)
T"""

":
(")

":
ID
" ": " "
-40.0

,,/: en

-
c. -50.0
1/1 250.0 300.0 350.0 400.0 450.0 500.0 550.0 600.0
Z
""
ii',! ,'l' :; Temperature (K)
,1

1/1 I I I I I I I I I
! CII
a::
,.1 >-
lOa. ;gl~"OC:~Ol{)ool{)r.oo>r.oC\JC\J
-u.O)
Figure 9-10. Measured and calculated enthalpy results at 1400 psia tor the jet
naphtha Ior which the boiling point curve is shown in Table 9-6.
I!I a,(ij
CII:5
-.c c::
.3 w~.g
c:::_II{)IC\Jlr.oloIC\JI"IC\Jloolo
~~C\JC\JC\J(')(')'<I"
a.. 90.0 fl ----------'--------- -,
w
t!!!
l'l
,II" ",r, ,
~
j I 1/1
70.0
1
I,t'' ..::J
CII

'Inu X >-
'1: 50.0
'1 , : c: (.)
(I)[j,
C\JIO>I"I~I('
ID
f'..
O> o )IOI~IO
t--: ~
O'> o
r---: ~
T"""

~
T"""

~
C\J
aJ
c_
I I
i
E
::J
CII s:
"" o
<lJ
30.0

- - E E
'O
.. <,
.." Measur ed
g-I I I I I I I I "'"
l CII
a. z :r: 10.0 -- Calculated
,1" ..
- "O'=~IC\J
(1)1 -
'i',[ o "') O) u.
i;
v 10> 1('r-,) 1
C\J C\J C\J1"o(') 10>
(') I~('v) 1"io(') 1"(')
00
!I, 1/1
CII
c:::_
W 0.= C\J lO 00 C\J CD -10.0

! ::J..
> ID o a..
U r
m
c:: ::::- -30.0
.2 e

~
.~s
a.I
:;:; -s. "-u -50.0
I I I ! I I I I I
1/1 _"O -i e
e 3:~ 011{) ol~lgl~IRlgl~I~~ '"
250.0 300.0 350.0 400.0 450.0 500.0 550.0 600.0 650.0

Ec.~ '"
u....J., Temperature (K)

8 o
o
ID
..
.,
t' Figure 9-11. Measured and calculated enthalpy results at 30 psia tor the low-
-
::J
o
(f)
boiling naphtha tor which the boiling point curve is shown in Table 9-6.
"

:1
154 Properties o] Oi/s and Natural Gases

100.0 r------------------------------------------------------------,
80.0

60.0

Chapter 10
a'"
..!:.
->
40.0
Tuning of EOS-Parameters
"'"
- 20.0
,
el.

~ 0.0
+'
B
20.0
THE NEEDS FOR TUNING
Measured
.40.0
--- Calculated Accurate data for the phase behavior of oil and gas mixtures are needed
60.0 in, for example, the design of process plants, and in reservoir simulation
studies. Often, experimental PVT-data such as those described in Chapter 3
-80.0 I .,.. I I 1 I I I

250.0 300 350.0 400.0 450.0

500.0
Temperature (K)
Figure 9-12. Measured and calculated enthalpy resulls al 700 psia for lhe gas/oil
for which lhe boiling poinl curve is shown in Table 96.
SUMMARY OF THE RESULTS FOR THERMODYNAMIC
PROPERTIES
Dew and bubble point pressures can usually be predicted to within 5 % .
Six hydrocarbon fractions are suHicient to represent the eornponents of a pe-
troleum reservoir fluido The fast flash version (kij = O) for the mixtures in
this chapter (which all contain less than 15 % non-hydrocarbons) give results
comparable in quality to those found with the standard flash.
The SRK-equation and the procedure of Peneloux et al. may both, with
good results, be used for calculation of gas-phase densities. The SRK-equa-
tion predicts insufficient liquid-phase densities, however, good results are
obtained with the procedure of Peneloux et al., with the API-procedure,
with the procedure of Standing and Katz, and with the Alani-Kennedy
equation.
The enthalpy is accurately predicted using the procedure of Kesler and
Lee for the ideal gas enthalpy, and the SRK-equation for the residual en-
thalpy.
will be available, but in practice, not very many PVT-measurements are
550.0 600.0
carried out for a given mixture. It is, therefore, necessary to rely on thermo-
dynamic models when planning the production strategies for a given petro-
leum reservoir. This raises the question of what accuracy can be obtained
using a cubic equation of state for phase equilibrium calculations, for exam-
ple, at the conditions where oil and gas are being produced.
As seen from Chapters 8 and 9, a reasonable accuracy can usually be ob-
tained for saturation points, gas/oil ratios, and phase densities, using the
SRK equation of state, with a proper procedure for characterizing the unde-
fined, heavy hydrocarbons. The only major deviation between measured
and calculated results is found in the gas/liquid ratio for gas condensa te mix-
tures, at pressures slightly below the dew-point pressure. The calculated
amounts of liquids are, in most cases, somewhat larger than those observed
experimentally (see, for example, Figure 8-1). Unless a different thermody-
namic model is used, the only way to improve the agreement between the
measured and the calculated results is to adjust the parameters entering into
the equation of state. This kind of adjustment is called tuning.
PARAMETERS AVAILABLE FOR TUNING
As described in Chapter 5, the parameters entering into the SRK equation
of state are:

REFERENCE 1. The critical temperatures (Te), the critical pressures (Pe), and the acen-
tric factors (w) of each component of the mixture
2. The binary interaction coefficients between each pair of components
Lenoir, J. M. and Hipkin, H. G., "Measured Enthalpies of Eight Hydrocar-
(kij)
bon Fractions," t. Chem. Eng. Data, 18, 1973, pp. 195-202.
3. The molar composition
155
156 Properties 01 Oils and Natural Gases Tuning 01 EOS-Parameters 157

35.0 j I
For the welI-defined components of the mixture (N2, CO2, Cl, C2, C3, i-
C4, n-C4, i-C5, n-C5, and n-C6), accurate measured data are available for
Te, Pe, and w, and it is usually not considered justified to use these as tuning 30.0

parameters. Te, Pe, and w of the C7 + -fractions are usualIy determined from
empirical correlations (see Chapter 7). Each C7 + -fraction consists of many
~
::l ~--- '<,
........
<,
~ 25.0 ./
/'

---.---!- -,
different components, each having different critical temperatures, critical
pressures, and acentric factors. It is not obvious that the empirical correla- o
+'
e;
/
./

............. " \
e, ........ <, \
tions represent the best possible choice of mixture properties. Therefore, at ~ 20.0 ~ "' .... <,

least in principle it seems justified to use Te, Pe, and w of the C7 + -fraction as o
Q)
.... ',. \
(
tuning parameters.
4-
o " \

...
15.0 Experiment
The binary interaction coefficients, kij, are introduced to account for the
Q)
E
::l Std. Characterization
\ \
effect of polar forces in the interactions between two components. Usually, o \ \
t> "Tuned" C7+-Properties (

,
10.0
kij for hydrocarbon-hydrocarbon interactions are considered equal to zero, "O

\ \
kij ~ for C, - C7+
or at least clase to zero. Cubic equations of state were originally not in-
::l
O" \i \.
\
tended for use on components deviating from sphericity as much as the 5.0

heaviest components of a petroleum reservoir Huid do. One possible way to

compensate for the non-sphericity of the heavy hydrocarbons could be to ~
0.0 100.0 200.0 300.0 400.0
introduce binary interaction coefficients differing from zero for the interac- Pressure (bar)
tions including a heavy hydrocarbon.
Coats and Smart (1986) present a procedure where, normally, Te, Pe, and Figure 10-1. Constant volume depletion experiment tor gas condensate (molar
w for the heaviest fraction, and/or kij between methane and the heaviest composition in Table 10-1).
fraction, are adjusted using non-linear regression to laboratory data. The
data might include saturation pressures, densities, K-values, constant vol-
ume expansions, differentialliberations, etc. While the desired stated objec- (
Table 10-1
tives may be obtained using this approach, it is shown in the Dangers of
Molar Composition of North Sea Gas Condensate (
Tuning section that highly undesirable effects may result for properties not
included in the objective function for the regression. (
Density
The molar composition of the petroleum mixtures may be determined as Molecular (g/cm at 3)
described in Chapter 2. Inaccuracies in the analytical techniques will, of Component Mole % Weight 15C, 1 atm
course, influence the results of the phase equilibrium calculations. It is, N2 0.42
therefore, considered of interest to analyze the effect, on the results of the e02 2.98
phase equilibrium calculations, of a small variation in the least accurately e, 66.36
determined molar component concentrations. e2 8.44
e3 5.12
ic, 1.04
e, 2.35
EXAMPLES ON THE DANGERS OF TUNING te, 0.84
n-es 1.12

The experimentalliquid dropout curve of a constant volume depletion ex-

e6 1.36
e7 2.14 93 .743
periment is presented in Figure 10-1 for the gas condensate for which the ea 2.20 107 .753
molar composition is given in Table 10-1. It is a heavy gas condensate, e g 1.43 120 .776
which, at pressures > 200 bar and temperature around 136.1C (the tem- e1Q 0.60 134 .790
perature of the depletion experiment), has almost dentical gas and liquid el1 0.48 147 .795

phases. Thus, it exhibits near critical behavior. The problems of handling (table continued on next page)
 (~~

158 Propertes o] Ols and Natural Gases Tunng o] EOS-Parameters 159;;:JY

Q: (
-~~<'l()"
\'0
-c -- ('
Table 10-1
Th~1~
Results of the Characterization of the C7 + -Fraction of the Gas condensat~ t; \ ~\~\.I'.ln:C~.
o:. ;
Continued
of Table 10-1 \ _ ,
\. /J....'
Density ,, ",
Te Pe Acentric
Molecular (g/cm3) at
Component Mole % (K) (Atm) Factor
Component Mole % Weight 15C, 1 atm
C7-9 5.77 560.7 30.07 .4327
C'2 0.40 161 .807
0.39 174 .823 ClO_'6 2.72 658.3 21.00 .7479
C'3
C'4 0.34 188 .831 C'7-60 1.48 802.3 12.16 1.2993
C'5 0.30 202 .836
Sauree: Pedersen el al. (1988a) .
C'6 0.21 217 .843
C'7 0.23 234 .840
C'B 0.16 251 .845
C'9 0.14 259 .856 Table 10-3
Results of Tuned Characterization of the C7 + -Fraction of the Gas
C20+ 0.95 397 .890
Condensate of Table 10-1
Sauree: Pedersen el al. (1988a).

Te Pe Acentric
Component Mole % (K) (Atm) Factor

two almost identical phases are reflected in the deviation, as presented in C7_9 5.77 571.0 27.00 .3000
Figure 10-1, between the measured liquid dropout curve and the one calcu- ClO-16 2.72 629.0 21.00 .6000
lated using the characterization procedure of Pedersen et al., described in C,7-60 1.48 782.0 14.50 1.3500
Chapter 7. The data used to represent the C7+-fractions are given in Table Sauree: Pedersen el al. (1988a).
10-2. In order to improve the agreement between the measured and the cal-
culated liquid dropout curve, tuning of Te, Pe, and W of the C7+ -components
was attempted. Binary interaction coefficients of zero were maintained for 80
G>
the hydrocarbon-hydrocarbon interactions. By employing for the C7 +-com- ti>

'"
.J:: Experiment
ponents the tuned critical temperatures, critical pressures and acentric fac- a.
Std. Characterization
tors of Table 10-3, the agreement between the measured'and the calculated lb
"Tuned" C7+-Properties
liquid dropout curves was very-much improved, as compared with the re-
sults of using the parameters of Table 10-2. The liquid dropout curve calcu-
.,
'"
(!l

75
lated using the tuned parameters is also shown in Figure 10-1. Figure 10-2 -s
.J::
---- ...
... ...
shows measured and calculated results for the methane concentration at
"- <,
various depletion levels. Figure 10-3 portrays measured and calculated corn-
~
<,
-, -,
pressibility factors at the conditions of the depletion experimento Figures o
::E 70
<,
-, "- ...
10-2 and 10-3 show that the tuned predictions are slightly inferior to the re-
sults found using the characterization procedure of Pedersen et al. This
I <,
<,

<,
<,
means that the tuning has only served to improve the results for the liquid <,
-,
volume which was used in the tuning process, and less accurate results are
obtained for other thermodynamic properties. 65 LI ----~----~----~----~----~----~-- ~ __ ~

Figure 10-1 also shows the liquid dropout curve resulting from a tuning of O 100 200 300 400
the binary interaction coefficients for interactions between methane and the Pressure, bar
C7 + -components. It was found that the liquid dropout curve could be much
improved when, maintaining the parameters of Table 10-2, the following Figure 10-2. Experimental and calculated methane concentration correspond-
ing to the constant volume depletion experiment of Figure 10-1.
non-zero binary interaction coefficients were introduced:
 160 Properties of Oils and Natural Gases Tuning of EOS-Parameters 161

600.0
~ 1 10

U -
!\l
o

I
l.
\
U.
I 500.0 \ Exp. De~ Point

-.--
c-,
/ \ Std. eharacterization
I
\
.c / \ - - - kij ~ for el - e7+
400.0
/ \
.,''"" 1.00 1/
;:-
"
~ __ -f::_",

. / <, -c
c. / '"
:=:
""
<::,
E /
/, 300.0
o
t '-'"
.,
u "
Vl
Vl
Q)

'-
'"
~~~._---
e,
!\l 200.0
s:
el. 0.90

'"
!\l .....
o 100.0
l
O 100 200 300 400
Pressure, bar 0.0
-100.0 0.0 100.0 200.0 300.0 400.0

Figure 10-3. Measured and calculated gas phase compressibility factors corre- Temperature "C
sponding to the constant volume depletion experiment of Figure 10-1. Figure 10-4. Calculated phase envelope for the gas condensate mixture for
which the molar composition is given in Table 10-1.
CJ - C7~9 :kj = - 0.25
Table 10-4
CI - CIO~16 :k;j = 0.075 Molar Composition of the Oll Mixture Mixed with Gas Mixture 01
Table 10-5
The phase envelopes shown in Figure 10-4 are calculated using the C7 +_
Density
properties of Table 10-2 and Molecular (g/cm3) at
Component Mole % Weight 15C, 1 atm
l. kij"values of zero for all hydrocarbon-hydrocarbon interactions
2. The non-zero k(values indicated above and kj-values of zero for all Nz 0.31
COz 0.69
other hydrocarbon-hydrocarbon interactions.
C, 47.69
The phase envelopes eorresponding to the tuned parameters have a mini- Cz 6.64
mum in pressure at approximately 27C. There is no physical evidence for Cl 4.59
-C. 0.85
this behavior, and the example ShOWSthat equation of state parameters de- n-C. 1.74
termined for one temperature may lead to highly erroneous predietions at ic; 0.70
temperatures that are only slightly different. c, 0.46
Only the molecular weight, and the density at 15C and 1 atm, have been Ce 1.69
measured for the C7 + -fraction of the oil for whieh analytical data are pre- C7+ 34.64 234 .8688
sented in Table 10-4. Gas (of the eomposition listed in Table 10-5) was grad- Source: Pedersen el al. (lg88a).
ually added to this oil, and after each addition of gas, the saturation point
pressure at 111.1 C was measured. The addition of gas was continued until cally in Figure 10-5. When the characterization procedure of Pedersen et al.
the saturation point shifted from a bubble point to a dew point. The addi- is used on a C7 + -fraction for which no detailed composition is available, a
tion of gas was simulated using the characterization procedure of Pedersen logarithmie dependenee according to Equation 7-13 is assumed The molee-
et al. (Chapter 7). Experimental and simulated results are shown graph- ular weight of the C7 + -fraction of the oil described in Table 10-4 is 234,
162 Properties o] Oils and Natural Gases Tuning o] EOS-Parameters 163

Table 10-5 Table 10-6

Molar Composition of the Gas Mixture Added to Oil Mixture of Molar Composition of the Oil Mixture Used in the Separator Calculation
Mole %
Reported in Tables 10-8 and 10-9
Component

N2 0.55 Density
CO2 1.43
Molecular (g/cm3) at
C, 81.63 Component Mole % Weight 15C, 1 atm
C2 10.12
5.20 N2 0.79
C3
CO2 0.17
-c, 1.06
C, 45.90
n-C. 0.01
C2 5.97
Source: Pedersen et al. (1988a). C3 4.37
ic, 0.92
650.0 I n-C. 2.35
/
ic, 1.06
600.0 / n-Cs 1.46
/ C6 2.31
/
/ C7 4.07 91.8 .734
'- 550.0 /
-o
.D
Ca 4.65 104.9 .756
/
/ C9 3.40 121.0 .775
Q)
'-
~ 500.0
/
CID. 22.59 271.6 .881
/
~ /
'"-
o- / Source: Pedersen et al. (1988a).
/
... '50.0 /
e /
o /

/.
o- /
e 400.0 /
o / . match of the measured bubble point of the original oil is achieved. However,
/
';;; as is shown in Figure 10-5, the estimated saturation points of the
/
Experimental Bubble Points
B'- 350.0 /;
/
.t. Experimental Dew Point oil + gas mixtures are, in this case, much different from the experimental
m
~
V1
~ Std_ Characterization values. This means that the kij-values which are appropriate for the original
?
300.0 Q
Tuned C7+-proper~ies oil give saturation points for the ol + gas mixtures that deviate considerably
from the measured results.
250. o ! ! I ! I I J I I I I
As explained in Chapter 2, the composition of a reservo ir oil mixture is
0.0 .\ .2 .3 .4 .5 .6 .7 .8 .9 1.0
usually determined from the compositions of the gas and liquid phases re-
Mole Fraction of Gas Added per Mole Dil
sulting from a single stage flash to standard conditions (i.e., 15C and 1
Figure 10-5. Experimental and calculated saturation points at T = 111.1e for atm). This is also the case for the oil composition given in Table 10-6. The
mixtures of the oil of Table 10-4 and the gas of Table 10-5. final composition is, of course, dependent on the assumed ratio for the
amounts of gas and liquido The influence, on the calculated phase behavior,
of assuming an erroneous gas/liquid ratio, has been investigated for the oil
which roughly corresponds to that of a Cl7-fraction. The influence of using mixture of Table 10-6. Experimental and calculated bubble-point pressures
a distribution function to represent the C7 +-fraction may be illustrated by at T = 74.5C are shown in Table 10-7. The calculated bubble-point pres-
calculating the bubble-point pressure of the mixture of Table 10-4, where sure using the correct molar composition agrees very well with the measured
the properties of n-Cl7 are used to represent those of the total C7+-fraction. bubble point, whereas the bubble-point pressure calculated using the erro-
In that case, a bubble-point pressure at 111.1C is calculated, which is neous composition deviates by 8.70/0 from the experimentally determined
34.1 % lower than the measured value. This allows the introduction of a bubble point. Hydrocarbon-hydrocarbon k(values, which depend slightly
binary interaction coefficient of 0.145 for the interaction between C and on the molecular weights (MW) of the constituent components in the fol-
the C7 +-fraction having the properties of n-Cl7, in which case a perfect lowing manner, were introduced:
164 Properties 01 Oils and Natural Gases Tuning 01 EOS-Parameters 165

Table 10-7
Measured and Calculated Bubble-Point Pressures at T = 74.5C for the :> C'! C\J (') :> CXl al O
Q) v al ci Q) ..t
Oil Mixture for Data Given in Table 10-6 e e ~ ~ ~ e LO (')

Pressure (bar) % Dev.

-&.
o
. o
i!-
.
'R I I I
c:
a
.
o
o
jj.
'S- I I I

E ~ o.~
Experimental bubble o E
-
u a: C\J al C\J
o
u
CXl
o
C\J
o <O
point 235.0 I/J r-: en ~ al ~ o
~ O
CJ r-, C\J
C\J
I/J
~
(')

~ ~ ~
Calculated bubble o
Q)
(O o
233.6 - 0.6 e Q)
point based on
measured composition (O
e... e
O
... e
...
o
c:
w :8
Calculated bubble
o... E lO
:>Q) al o v
r-, ,.j r-:
Q)

:c W o
;
:>Q) o (')
,.j
al
,.j
214.5 - 8.7 ~ ~ .!:! e ~ ~ t!!! E
o .. e N
10
point based on Q)
I I I
:c a: 'S- I I I ~ ! 'S-
composition with
t!!! O
ti
lO
'O o o
erroneous GOR CJ :;; ..
Q)
ID
elO
Calculated bubble
'O
..:1

j
s:
u
-:1 ()
j s:

-
Q) U al
a: al C'! 10 >< U 10 C\J
points based on : o o
0.0
U 'O
'jj C!J J:'-
O (') C\J N
~ 'O al ~ o
(')
10

composition with 235.0 ><

(')
'jj ~ ~ ~
erroneous GOR using s 5..
k;-values different
s.. e s
- -
o e
o :>
from zero :>Q) ..
-
1/1 r-,
% Dev. = 100 x (Cale. - Exp.)/Exp.
o E~
o o e
..o.
'"
C\J
CXl
cri "",.j o E .~ e
Q)
C\J '"
C\J
10
,.j

1/1 E 'S- I I I u o o. 'S- I I I

Sauree: Pedersen el al. (1988a). o a: o m u..ni ~ E
(1); o o

, ni
... a:: u
CJ'C o...oe IDU
''0
(10-1) Q)= ()
'Q)
...~ Q,):;: ~ Q)
k = 0.00145 MW for MW>MW -o
.0--
j a: ~ ..t
r-,
LO
,.j
-
.o E ni
lO .
U ~ ~
C\J r-, LO

MW t!!! :
U I/J
lO O
~ t!!! . I/J al LO
~ C!
(')
10

Q) CJ CXl (') lO
Q) ~ ~ ~
o.. ::E ~ ::E
k = 0.00145 MW
MW
for MW>MW (10-2)
-o s.. M

--
'O
1i ~a: o 'O Q;

~.. o o 10 ~ al LO C
~O ci r-: ,.j ~ o
""~ o'"': os:'"
-
CXl V
With these binary interaction coefficients a perfect agreement with the QC!J al ~ 1/1 I/JID (')

o Q)
(')

~
lO
~ ~ ~ o
measured bubble point is obtained (see Table 10-7). Results of a three-stage .. ::E .. Q)

- ::E U)
ni
M <ti
separation process are available for the oil mixture of Table 10-6. The results ni
o. Q; o Ol

are summarized in Tables 10-8 and 10-9. Simulated results are shown for the Q) C ..
C\I Ol
~
correct composition, and for the composition with a gas/liquid ratio 5 % -
!Jl

o 6
z, C'! 10 C!
o'"
s: ni
C.
Q) 6 LO C\J 10
-o
:;
g
lower than what was measured. Results are shown using hydrocarbon-hy- 1/1
:: 1-
::;: cri
CXl ~
LO fE
<ti
-
en
o
~
1-
v
t-, ::;:
cri
CXl a

-
Ol
drocarbon interaction coefficients equal to zero, and determined from :1 Ol
E
1/1
Q)
~ 1/1
-o
Equations 10-1 and 10-2. The results in Tables 10-8 and 10-9 indicate that a:: 6 . :;
the erroneous analytical data are not accounted for by tuning to a measured
bubble point, using k-values different from zero, for hydrocarbon-hydro-
~
e
al
a:i
LO
N
C!
~
:o~~ 1/1
Q)
a::
e
...
lO
o al
LO CXl
LO
N
~~~
~~~co
U) Ol

a.. 10 C\J ~~ffi (') 10 C\J (fi:O~ __

a.~
a.. C\J
carbon interactions. ~~~
...e Cii "2ccij
(l) .- _ ::J .- _
The examples just given show that equation of state parameters tuned to a. U) Ol cr U) Ol

one specific property cannot be expected to give reliable predictions for

U)
"'-oOl
'" c: ~;2 ~ s
ti OJOlU)
c: ~ ti '"
ID
c:
Vl
~
E~~
:;:: (l)
other therrnodynamc properties. If the analytical data are inaccurate, tun- -
lO el
Q) -
f
Q) '" -o
a. Ol -o
. lO ~ C\J (')
"E~~
el
~ C\J (')
Ol

ing to the properties reported for these erroneous compositions may lead to lO_
g.!Jl
(tJ
-o
> ..
Ol Ol
lO lO
0.-
Q)!Jl
Ol
ln a; "
. "-

highly uncertain predictions.

c: O u o ~
!Jl
'" ~
5'if.g
!Jl 1,
~"
o
(J) (J)
166 Properties 01 Oils and Natural Gases
Tuning 01 EOS-Parameters 167

Table 10-10
Gani and Fredenslund (1987) have developed a two-part tuning proce-
Experimental and Calculated Dew Point Results for 10 North Sea Gas
dure that overcomes some of these problems:
Condensate Mixtures
l. A sensitivity analysis, to select the parameters to tune, and a feasibility
Temp. Exp. Dew Pt. Calc. Dew Pt. Adjust- Plus-
study, to check whether tuning can/should be done
No. (OC) (bar) (bar) % Dev. ment of MW fraction
2. If tuning is feasible, it is accomplished by regression; and evaluation of
the results takes place 1 96.6 282.0 311.6 10.5 - 8.8 e,,.
2 93.8 235.0 235.2 0.1 0.0 e+
1O

While this tuning policy often yields reasonable results, and often will 3 118.9 398.0 377.1 - 5/3 8.3 e+
1O

warn the user not to tune, it is also not completely safe. The dashed curve in 4 119.7 394.0 369.7 - 6.2 9.6 e+
1O

Figure 10-4 is drawn with kij-values for C1-C7+ computed from this proce-
5 150.3 381.5 396.3 3.9 - 5.4 e+
20

6 154.1 385.5 410.2 6.4 - 9.6 e+

1O
dure. The reason for the incorrect rise of the bubble-point branch is that the 7 131.5 367.0 378.7 3.0 - 9.0 e+
20

sensitivity of the low-temperature bubble points to the adjusted kifvalues 8 129.0 464.0 454.2 - 2.1 5.6 e+
20

was not taken into account. 9 136.1 386.4 375.6 - 2.8 8.4 e+
20

10 148.9 542.5 493.5 - 9.0 12.1 e+

20

%AAD 4.9 7.7

% BIAS - 0.2 1.1
THE MOLECULAR WEIGHT OF THE PLUS-FRACTION AS A % adjustrnent needed in Ihe molecular weighl 01Ihe plus Iraction lo get agreement wilh the measured
TUNING PARAMETER dew point maintaining Ihe measured weight composition.
% dev. delined as in Table 10-7.
% AAD = % average absolute deviation.
As explained in Chapter 2, the compositional analysis for oil and gas rnix- % BIAS = % average deviation.
Source: Thomassen et al. (1986).
tures is most often given in weight units. Conversion to molar compositions
requires knowledge of the molecular weights of each component and carbon
number fraction. The experimental inaccuracy in determining the molecu- Table 10-11
lar weights of the plus-fractions is of the order 5-10%. Deviations in that Measured Molar Composition of Gas Condensate 10 of Table 10-10
order of magnitude can considerably influence a calculated dew-point pres-
sure of a gas condensate mixture. An obvious application of this fact is to use Densily
Molecular (g/cm') al
the molecular weight of the plus fraction as an adjustabl parameter. Ad-
Componenl Mole % Weighl 15C, 1 atrn
justments within the experimental uncertainty can, e.g., be used to match a
measured saturation point, as suggested by Pedersen et al. (1988b). N2 .39

Table 10-10 presents experimental results for the dew-point pressures of 10

e0 2
3.47
e, 80.17
gas condensate mixtures. Also shown are the dew-point pressures calculated e2 6.28
using the characterization procedure of Pedersen et al. as described in Chap- e3 2.75
ter 7. The average absolute deviation between the experimental and the cal- -c, .43

culated dew-point pressures is 4.9%. To improve the dew-point calcula- n-C, .88

tions, Pedersen et al. attempted to treat the molecular weight of the plus
te, .31
n-es .35
fraction as an adjustable parameter, maintaining the measured weight com- C6 .54
position. Table 10-10 also shows the adjustments needed in the molecular C7 .72 96 .743
weights of the plus-fraction to get agreement with the measured dew point. ea .88 106 .761
The maximum adjustment required is 12.1 % (gas condensate 10). The mo- eg .55 118 .779

lar compositions of this mixture before and after 'ths adjustment are shown
e10 .33 132 .787
C .24 149 .788
in Tables 1O-1l and 10-12. Figure 10-6 shows the measured liquid dropout "
C'2 .20 163 .799
curve of a constant mass expansion experiment (described in Chapter 3), on
(table continued on next page)
gas condensate 1, in comparison with those calculated using:
168 Propertes 01 Oils and Natural Gases
Tuning 01 EOS-Parameters 169
Table 10-11 8.0 I i

Continued

Density Experimental Results

(j)

Molecular (g/cm3) at 5 6.0

Calculated Based on Measured MW
Component Mole % Weight 15C, 1 atm o Calculated Based on Adjusted MW
:>
..,
----
.."
C'3 .21 175 .820

.
e .
C'4 .18 194 .831 o
C,S .15 203 .833
e,
/~~"
C'6 .11 217 .838 "
(j)
4.0 -, -,

.
'"
4-
C17 .12 235 .836
C,S .09 247 .842 U
o

::>
" -, \
\
C'9 .08 255 .852 o-
\
C2D+ .57 396 .867 \
~ ,.0
Source: Thomassen el al. (1986). o \
r>
\
\

Table 10-12
Molar Composltion of Gas Condensate 10 of Table 10-10 After an
Adjustment of the Molecular 'Weight of the C20 + -Fraction by 12.1 %
0.0
0.0
I ,
100.0
,
200.0
..\ \ ,\
300.0
I
400.0
Pressure (bar)
- Density
Figure 10-6. Liquid dropout curve (constant mass expansion) lor the gas con-
Molecular (g/cm3) at
Component Mole % Weight
densate mixture of Tables 10-1 and 10-12.
15C, 1 atm
N2 .39 l. The measured molecular weight of the plus-fraction and
CO2 3.47
2. The molecular weight of the plus-fraction adjusted to get agreement
C, 80.22
with the measured dew point.
C2 6.28
C3 2.75 Figure 10-7 shows measured and calculated results for the compressibi!ity
i-C4 .43
factor of the gas phase liberated during the constant volume depletion ex-
n-C4 .88
ic, .31 periment. The results indicate that the adjustment of 12.1 % in the molecu-
c, .35 lar weight of the plus-fraction has ver y little influence on the calculated gas
C6 .54 phase density.
C7 .72 96 .743 Pedersen et al. find that an accurate molecular weight of the plus-fraction
Cs .88 106 .761
is less important for the bubble point of an oi! mixture than for the dew
C9 .55 118 .779
ClO .33 132 .787
point of a gas condensate mixture. Adjustment of the molecular weights is,
Cll .24 149 .788 therefore, a less effective tool for oil mixtures than it is for gas condensate
C'2 .20 163 .799 mixtures. On the other hand, there is also less need for improvements in the
C,3 .21 175 .820 calculations on ol mixtures, beca use accurate results can in most cases be
C'4 .18 194 .831 obtained based on the measured compositions .
C,S .15 203 .833
C'6 .11 217 .838
C17 .12 235 .836
C,S .09 247 .842 RECOMMENDATIONS REGARDING TUNING
C'9 .08 255 .852
C20+ .51 444 .867
Agreement between the results of PVT-experiments, and those obtained
Source: Thomassen el al. (1986).
with a cubic equation of state, may be improved by adjustment of the criti-
 /e:"';' ''\()~
170 Properties 01 Oils and Natural Gases
Tuning 01 EOS-Parameters 17
'<'
o: ~\al\Oll:~~~
./a:! - ~~\
\1

f
. 1--

Experimental Results '1

'1
-
/
Pedersen, K. S., Thomassen, P, and Fredenslund, Aa., "On the Dangers of\") \ PII _
Tuning Equations of State Pararneters," Chem. Eng. Sci., 43, 1988a, pp. \:,.;\... ,
269-278. .~
, ,...
~

/ h7A
'-
o --- Calculated Based on Adjusted MI< '1
+-1
u
1.2 '1 Pedersen, K. S., Thomassen, P, and Fredenslund, Aa., "Characterization of ' "
-o - - - Ca I cu I ated Based on Measured MI< '1
u; Gas Condensate Mixtures,' paper presented at the 1988 AIChE Spring
e-,
+-' '1 '" National Meeting, New Orleans, LA, March 6-10, 1988b.
'1
D
: Thomassen, P, Pedersen, K, S., and Fredenslund, Aa., "Adjustment of C7 +-
'1
~ 1.1 '1 Molecular Weights in the Characterization of Petroleum Mixtures Con-
QJ jI
'-el. '1 taining Heavy Hydrocarbons,' SEP Report 8617, Instituttet for Kemi-
>-
5
u teknik, The Technical University of Denrnark, 1986.
'/
~ '/
-o
.c;
'/
'/

/-
o.. 1.0
.'/
V>

'~
/
-c
co
./
- - --==----- -

0.9
0.0 100.0 200.0 300.0 400.0 500.0 600.0
Pressure (bar)

Figure 10-7. Gas phase compressibility factor of the gas condensate mixture of
Tables 10-11 and 10-12.

cal temperature, the critical pressure, the acentric factor, or the binary in-
teraction coeHicients. It is, however, unlikely that parameters tuned to one
specific property will provide accurate results for other properties. Pararne-
ters determined in a limited pressure and temperature range may lead to
highly inaccurate results when considered at pressures and temperatures not
included when performing the parameter estimation.
Rather than adjusting the equation of state parameters, it is recom-
mended treating the molecular weight of the plus-fraction as an adjustable
parameter. The experimental uncertainty on the molecular weight of the
plus-fraction is of the order 5-10 %, and it seems, therefore, justified to ad-
just the molecular weight of the plus fraction (e.g., against a measured satu-
ration point) within these limits.

REFERENCES

Coats, K. H. and Smart, G. T., "Application of a Regression-Based EOS

PVT Program to Laboratory Data," SPE Reserooir Engineering, May
1986, pp. 297-298.
Gani, R. and Fredenslund, Aa., "Therrnodynarnics of Petroleum Mixtures
Containing Heavy Hydrocarbons: An Expert Tuning Systern ,' Ind. Eng.
Chem . Research, 26, 1987, pp. 1304-1312.
 Viscosty 173

the molecular hard sphere diameter. The following dependence of 1/ on the

temperature and the molecular weight may then be derived:

1/ = constant (T1I2MW1I2/a2) (11-3)

a3 is often associated with the critical molar volume, Ve' If it is further as-
Chapter 11 sumed that Ve is proportonal to RTe/Pe, the following expression may be
Viscosity obtained for 1/ at the critical point:

n: = constant (MW1I2 p~/3/nI6) (11-4)

DEFINITION OF VISCOSITY
An external shear stress applied to a portion of a fluid will introduce a
movement of the molecules of the affected part of the fluid in the direction
of the applied shear stress. The moving molecules will interact with the
neighboring molecules. These will start moving too, but with a lower veloc-
ity than that of the molecules exposed to the stress. The dynamic vscosity, 1/,
of a Newtonian fluid is defined as the following rato for the flowing fluid:
1/ = - Txy/(aVx/ay) (11-1)
where Txy= shear stress
Vx = velocity of the fluid in the x-direction (the direction of the
applied stress)
avx/ay = the gradient of Vxin the y-direction (perpendicular to the
x-direction) .
Though kinetic gas theory is not applicable to the near critical region, this
expression plays an important role in vscosity calculations. It is convenient
to introduce the viscosity reducing parameter, ~:
~= T~/6/MW1I2 p~/3 (11-5)
~
VISCOSITY CORRELATIONS
Several methods have been suggested for estimating the viscosity of hydro-
carbon mixtures. In the simulation of processes related to oil and gas produc-
tion, vscosity correlations are needed that are applicable to a wide range of
hydrocarbon mixtures and process conditons. Methods limited to narrow
ranges of compositon and/or temperature and pressure will not be covered
here. Only calculaton procedures that can be used for both gas and liquid
phases, giving consistent results for each, are discussed.

THE VISCOSITY CORRELATION OF LOHRENZ ET AL.

KINETIC GAS THEORY
The viscosity correlaton that has the most widespread use in flow models
For a dilute gas, one may derive an approximate analytical express ion for for petroleum mixtures is probably the correlation of Jossi et al. (1962) in the
the viscosity, 1/. It is: form suggested by Lohrenz et al. (1964). Gas and liquid viscosites are re-
lated to a fourth-degree polynomial in the reduced density, o, = p/ Pe:
1/ = 1/3 n v MW L (11-2)
[(1/ - 1/')~ + 10-4]1/4 = al + a2Pr + a3Pr2 + a4Pr3 + a5Pr4 (11-6)
where n = number of molecules per unit volume
v =' average molecular speed where al = 0.10230
MW = molecular weight a2 = 0.023364
L = mean free path between two molecules. a3 = 0.058533
a4 = - 0.040758
It can be shown (see e.g., Hirschfelder et al., 1954) that the average speed, v, a5 = 0.0093324
is proportional to (RT/MW)1/2 and the mean free path.to l/(na2), where a is

172
 Properties oj Oils and Natural Gases Viscosity 175
174

By expressing the viscosity as a function of the density, the calculated vis-

TI' is the low-pressure gas mixture viscosity. t is the viscosity-reducing
rameter, which for a mixture is given by the following expression:
pa-
cosity becomes ver y sensitive to the results obtained for the density. Espe-
cially for high viscous fluids this may lead to severe errors for the calculated
viscosity (see the examples in the Comparison with Experimental Results

t
N
= [ .1: zTci
ll/6[ .1:N zMW1:
l- 1/2[ N l- 2/3
zPe (11-7) section) .
1=1 1=1 1= 1

where N is the number of components in the mixture, and Zthe mole frac- CALCULATING VISCOSITY USING THE CORRESPONDING
tion of component i. Pure components have a well-defined Pe, but it is not STATE S THEORY
obvious how to determine the critical density of a complex hydrocarbon
mixture. Lohrenz et al. (1964) suggest calculating the critical density of a The viscosity may al terna tivel y be calculated using a modified form of the
petroleum fluid as follows: corresponding states method. The starting point-is the properties at the criti-

p, = (Ve)- 1 = (t1= 1
(z.Vci) + ZC7+ VeC7+) - 1 (11-8)
cal point (e.g., the critical viscosity, Tle,the critical density, Pe, the critical
pressure, Pe, and the critical temperature, Te)' Away from the critical point,
the properties may be expressed in terms of reduced properties (TI" p" P" T"
i*C7-t
etc.), which are defined as the properties at the actual conditions, divided
where the critical molar volume (in ft3/lb mole) of the C7+-fraction is found by those at the critical point. A group of substances obey the corresponding
from the expression: states principIe with respect to viscosity, if the functional dependence of Tlr
on, e.g., o, and T, is the same for all substances within the group.
V,C7+ = 21.573 + 0.015122 MWC7+ - 27.656 X SGC7+
Tlr(P, T) = f(p" Tr) (11-13)
+ 0.070615 MWC7+ x SGC7+ (11-9)
In that case, compreheflsive viscosity data are only needed for one of the
In this expression, MW is the molecular weight, SG the specific gravity, and components of the group. That component is then used as reference sub-
the sub-index C7 + indicates that the mentioned property is an average value stance (o), and the viscosity of another component (x) within the group can
for the hydrocarbons with 7 and more carbon atoms. easily be calculated. In general, the critical viscosity (TIe) will not be known.
The dilute gas mixture viscosity TI*is determined as f.ollows (Herning and The inverse of t for which the expression is given in Equation 11-5, may be
Zippener, 1936): . used as an approximate value. When the corresponding states principle is
expressed in terms of p" and not P" it is convenient to express ~ in terms of
N N Ve, instead of Pe' The final expression for the reduced viscosity (Tlr)then be-
r* = 1: zTI*MWI/2/1: ZMWI/2 (11-10)
comes:
= 1 = 1

Tlr(P, T) = TI(p, T) (Te)-1/2 (Ve)2/3 (MW)-1/2 (11-14)

For the dilute gas viscosity of the individual components, TI~the following \
expressions are used (Stiel and Thodos, 1961): The viscosity of component x, at the temperature T, and a pressure where
it has the density P, is then found from the expression:

TI*= 34 X 10-5 ~ T094 for T; < 1.5 (11-11)

t n
Tlr(P, T) = (Te,/Teo)1I2 (Vex/Veo)-2/3 (MW,/MWo)I/2 ro(Po, To) (11-15)

where Po = P Peol Pe,

TI*= 17.78 X 10-5 ~ (4.58Tr - 1.67)5/8 for Tr > 1.5 (11-12)
T, = T Teo/Te,
~
Tlo= viscosity of the reference substance at temperature T, and den-
sity Po
where ~ is given by the expression in Equation 11-5.
176 Propertes of Oils and Natural Gases Viscosty 177

The corresponding states principie, in this simple form, may be used suc- The deviation from the simple corresponding states principIe is expressed
cessfully for, e.g., mixtures of hydracarbons of approximately the same in terms of a ratationaI coupling coefficient, Q' (Tham and Gubbins, 1970).
chemical structure and molecular weight. Oil and gas mixtures consist of a The final expression for the viscosity of a given mixture is the following:
large number of components, including paraffinic, naphthenic, and aro-
matic components. For oil mixtures, the molecular weights may range fram 1/m.(P,T) = (Te,mi.lTeo)-116(Pe,mixIPeo)213(MWmix/MWo)1I2
16 (methane), to about 1100 (Cso). It is not to be expected that methane and Q'mix/Q'o1/o(Po,
To) (11-19)
Cso should both belong to a group where the simple corresponding states
principie applies. Very comprehensive experimental viscosity data exist for P PeoQ'o
methane, and methane is therefore a natural choice as the reference sub- Po = P Q" (11-20)
c,mix rrux
stance. To apply the corresponding states principie to the remaining compo-
nents found in oil and gas mixtures, some modifications must be intraduced,
T TeoQ'o
as compared with Equation 11-15. Ely and Hanley (1981) have suggested (11-21)
T, = T e, mixQ'mix
expressing the deviations from the simple corresponding states principie in
terms of two shape factors, e(T" V" w), and <I>(T"V" w), which en ter into
The critical temperature and the critical molar volume for unlike pairs of
Equation 11-16 as follows:
molecuIes (i and j) are found using the following formulas:
"
1/x(p, T) = (eTex/Tco)1I2 (<I>VexlVco)-213(MWx/MWo)1I2
Tcij = (TciTej)1I2 (11-22)
1/o(po<l>,Tole) (11-16)

!
V el).. = 8 (Vel)13+ VeJ)13)3 (11-23)
Vr is the reduced molar volume and w is the acentric factor. Vr (and thereby
p) is itself found by a corresponding states model, and the determination of
e and <1> therefore involves an iterative procedure. To use the corresponding The critical molar volume of component i may be related to the critical tem-
states principie on a mixture, it is necessary to represent the mixture as a perature and the critical pressure as follows:
fluid consisting of one hypothetical pure component with a given Te, Pe, and
MW. Ely and Hanley adapted the mixing rules of Mo and Gubbins (1976) Vci = RZciT e;lP ci (11-24)
for Te and P, (the expression for Te is given in Equation 11-27, and the ex-
pression for P, in Equation 11-30). The average molecular weight is formed where Zci is the compressibility factor of component i at the criticaI point.
from considerations regarding the radial distribution function for a fluid in Assuming that Z is the same for all components, Equation 11-23 may be
non-equilibrium due to the presence of a shear. Baltatu (1982 and 1984) ex- rewritten:
tended the pracedure of Ely and Hanley to cover boiling point fractions of
petraleum mixtures also. The critical properties and the molecular weights Vcij = - 1 constant ((T~ ) 113 (T)
+ ~ 113)3
(11-25)
of these fractions are estimated using the method suggested by Riazi and 8 r, Pel
Daubert (1980).
Pedersen et al. (1984 and 1987) presented a corresponding states model for The critical temperature of a mixture is found fram the following expres-
viscosity calcuIations where the reduced viscosity is expressed in terms of the sion:
reduced pressure, P" and the reduced temperature, Tr:
1: 1: ZiZjT V
i j
cij cf

n, = f(P" Tr) (11-17)

Te,mix= 1: 1: Z;ZjVcij (11-26)
i j
where the following expression is used for the reduced viscosity:
where Zand Zjare mole fractions of components i and j, respectively. Using
1/r = 1/~ = 1/1(Te-1I6 p~13MWII2) (11-18) Equations 11-22 and 11-25, Equation 11-26 may be rewritten:
 Viscosity 179
178 Properties o] Oils and Natural Gases
Ths mixng rule is derved emprcally on the basis of available vscosty

"" [(T)113 (T )113

p data, and assigns (as one may expect) to the heaver components a relatvely
large nfluence on the mixture vscosty. The o-parameter of the mixture is
T . = 'r' 'T' ZiZj P:: + P:; T
j
[Tci Cj]112
found from the expresson:
C,OlIX
(11-27)
3
1: 1:
I J
z.z [(Tel)113+ (Tej)1131
Pel Pej
amix = 1.000 + 7.378 x 1O-3p;.847 MWmi~S173 (11-34)

and the o-parameter of the reference substance (methane) as follows:

For the crtical pressure of a mxture, Pe,mi the followng relation is used:
ao = 1.000 + 0.031 p;847 (11-35)
Pe,mix= constant Te,mi)Ve,mix (11-28)
The constants and exponents in Equatons 11-34 and 11-35 are estimated on
where the constant equals the one enterng nto Equaton 11-25 and Ve,mixis the bass of experimental viscosity data. In Equations 11-34 and 11-35:
ti
found as follows:
p; = Po(T TeofTe,mix,P Peo/P e,mix)1
Peo (11-36)
Ve,mix= 1: 1: ZiZjVeij
i j
(11-29)
By expressing the viscosity in terms of the reduced pressure, nstead of the
From Equations 11-27, 11-28, and 11-29, the followng expresson may be reduced densty as in Equaton 11-13, it is possible to perform a direct calcu-
laton of the viscosty. The calculations become much smpler, compared
derved for Pe,mix:
wth those of Ely and Hanley (1981), where the densty is calculated usng
an terative procedure. The use of methane as the reference substance pres-
113 3
ents problems when methane is in a solid form in its reference state. Peder-
P . = 8 "" 'r' 'T' ZiZj[(T)113
~ + (T
~ ) 1[TciTej]112
"i sen and Fredenslund (1987) have suggested a procedure for mproving the
(11-30) i
~f viscosity predictions in those cases. The reference vscosty correlation is
c.nux (~ ~ [(!::)", (!;)"']'l'
+ :1.
1,
~
based on the methane viscosity model of Hanley et al. (1975):
.i
') 1)'(p,T) = 1)o(T) + 1)l(T)p + ~1)'(p,T) (11-37)
l~!
The mxing rules of Equations 11-27 and 11-30 are those recommended by
Mo and Gubbns (1976). where 1)0' 1)1,and ~1)' are functions defined in the above reference. The
The mixture molecular weght is found as follows: methane densty is found using the BWR-equation in the form suggested by
McCarty (1974). In the dense liqud regon ths expression is mainly gov-
erned by the term ~1)' (p, T):
MW mix= 1.304 x 10 - 4(MW~;303- MW~303) + MW n (11-31)

where MWw and MW n are the weght average and number average molecu-
lar weghts, respectively:
~1)' (p,T) = expfj + i4/T) [exp [p01(h + i~/2)
N N
MW w = 1: ZiMWr/1:
i~ 1 i~ 1
ZiMWi (11-32) + 8pOS(is + ~ + ;2)] - 1.0] (11-38)

N where in the work of Hanley (1975), the coeffcients j-h have the followng
MWn = 1: ZiMWi
i~ 1
(11-33) values:
 -.*:-
/. '\

18;;>-----;
180 Properties oj Oils and Natural Gases Vscosty
.,
t = - 10.3506 HTAN = exp(~ T) - exp( - ~ T)
iz = 17.5716 exp(~ T) + exp( - ~ T) ".-
.-
b = - 3019.39 with ~ T = T - TF
j4 = 188.730 TF = freezing point of methane
s = 0.0429036 iIl':.

j6 = 145.290
h= 6127.68 j COMPARISON WITH EXPERIMENTAL RESULTS

f) is given by: ~.
Table 11-3 gives experimental viscosity results (measured as described in
(11-39) Chapter 3) for three North Sea oil mixtures whose compositions are given in
f) = (p - Pc)/Pc Tables 11-1, 4-7, and 11-2. The results are presented graphically in Figures
,~. 11-1 through 11-3, in comparison with the results calculated using the
Pedersen and Fredenslund (1987) have estimated an additional set of coeffi- ~ method of Pedersen et al. (1984, 1987) and the method of Lohrenz et al.
cients of Equation 11-38 using viscosity data measured at reduced tempera-
(1964). The latter results were obtained using densities calculated from the
tures below 0.4, i.e., corresponding to a methane reference state where [i
1~'
.,
methane is in a solid formo The data include viscosities of oil mixtures and
oil fractions (Baltatu, 1984). The estimation gives the following results: iil.~; Table 11-1
~ Molar Composition of the North Sea Oll (Mixture 1) for which Viscosity
~\,

~ Data are Shown in Table 11-3

~1)"(p,T) = expk + k4/T)[exp[po+2 + ;~2) ~.

Density
(g/cm3) at Molecular
Component Mole % 15C, 1 atm Weight

+ OP"(k, + ~ + ~:)]- 10]. (11-40) N2

CO2
0.41
0.44
C, 40.48
C2 7.74
with k = - 9.74602
C3 8.20
k2 = 18.0834 ic, 1.23
k3 = - 4126.66 n-C. 4.22
k, = 44.6055 ic, 1.43
ks = 0.976544 n-C, 2.21
C6 2.83
~ = 81.8134
C7 4.13 100 0.7294
k7 = 15649.9 Ca 4.31 106 0.7492
C9 3.13 121 0.7697
Continuity between viscosities above and below the freezing point of 2.439
ClO 135 0.7861
methane is secured by introducing ~1)" as a fourth term in Equation 11-37: 1.880 148 0.7919
C"
C'2 1.674 161 0.8037
1)(p,T) = 1)o(T) + 1)(T)p + Fl~1)'(p,T) + F2~1)"(p,T) (11-41) C'3 1.573 175 0.8191
C, 1.207 196 0.8331
C,S 1.232 206 0.8359
F = HTAN + 1 C'6 0.985 224 0.8429
2 Cl? 0.977 236 0.8400
c., 0.911 245 0.8458
C'9 0.585 265 0.8575
F = 1- HTAN C20+ 6.382 453 0.9183
2
2
182 Properties o] Oils and Natural Gases
Viscosity 183
Table 11-2
1.75
Molar Composition of the North Sea Oil (Mlxture 1) for which Viscosity
Data are Shown in Table 11-3
1.5 Experimental results
Density
\
Component Mole %
(g/cm') at
15C, 1 atm
Molecular
Weight 1.25
\ -----
Pedersen et al.
Lohrenz et al. using SRK density
\
Nz 0.33 ------ Lohrenz et al. using Peneloux density
COz 0.19 ~ \
.::. 1.0
C, 35.42
>,
\
Cz 3.36 +'

'" "
-- --
C3 0.90 '"o .75

-C, 0.69 ~
n-C,
ic,
n-Cs
0.26
0.26
0.14
>

.5 " <, "-._----:-

...-e--e-----e----e
C6
C,.
0.72
57.73 255 0.9165
.25 --------------

0.00.0
100.0 100.0 300.0 400.0
Table 11-3 Pressure (bar)
Measured Viscosities of Oil Mixtures for which Molar Compositions are
Given in Tables 11-1, 4-7, and 11-2 Figure 11-1. Measured and calculated viscosities at T = 92.8C of the liquid
(Below the Saturation Point Pressure, the Results for the Viscosity Apply to phase of oil mixture 1 for which the composition is given in Table 11-1. The mea-
the Liquid Phase) sured data are tabulated in Table 11-3.
3.0

Mixture 1 Mixture 2 Mixture 3

at T = 97.8C at T = 93.3C atT=71.1C
2.5
\
Pressure Viscosity Pressure Viscosity Pressure Viscosity
(bar) (cP) (bar) (cP) ~bar) (cP) \ Experimental results
-------- Pedersen et al.
401.9 0.359 389.8 0.469 345.7 2.64 \
376.8 0.352 347.7 0.447 276.7 2.44 e,
.::.
1.0
\ ----- Lohrenz et al. using SRK density

325.7 0.340 302.1 0.425 242.3 2.34 c-, '" ----- Lohrenz et al. using Peneloux density
275.3
251.3
0.322
0.316
285.9
275.6
0.413
0.406
207.8
*158.9
2.24
2.10
+'

'"o
u
'"
1.5
'" .""'-
.........

0.306 *274.5 0.404 145.5 2.21

> <;
223.8 1.0
*203.5 0.299 260.9 0.424 125.1 2.41 ... "-
............ <, --.--"
180.8
151.6
0.316
0.360
225.1
143.2
0.484
0.651
104.4
83.7
2.65
2.93 .5
~ -----~--_._.
-- -- ------ ~!..--~
126.8 0.404 107.9 0.761 63.0 3.29
101.1 0.448 64.7 0.911 42.4 3.70
0.0
77.1 0.512 35.8 1.051 21.7 4.23 0.0 100.0 100.0 300.0 400.0

58.2 0.570 7.9 4.82 Pressure (bar)

26.6 0.715 Figure 11-2. Measured and calculated viscosities at T = 93.3C of the liquid
Experimental Saturation Points phase of oil mixture 2 for which the composition is given in Table 4-7. The mea-
sured data are tabulated in Table 11-3.
 Viscosity 185
184 Properties of Oils and Natural Gases
4.0 I I
11.0 i I

Experimental results Experimental results for crude o I with

10.0 \ Lohrenz et al. using SRK density TB = 87.6 "C

8.0
'" ".." Lohrenz et al. using Peneloux density

Procedure of Pedersen et al.

N
~

E
E
3.0

.,
o Experimental results for crude oil with
TB = 212.6 "C
Pedersen et al.(1987)

- - ---
>,

'-.
a.
~
>,
+'
'" '- '<, --
+'

8'"
'"
2.0 '-.".-......:~
Ba Itatu (1984)

'"ou
'"
6.0
-"----'- ".
u "'-'-...;::::, _
".
4.0

1.0
~- ---=- ------------
-.-....--e--- e -----
<V
e;

" i.o [
L -7-"=- = -=- D
--------
"J ,

0.00.0 15.0
I 50.0
1 75.0 Temperature
0.00.0 100.0 100.0 300.0 400.0 Pressure ("C )
(bar)
Figure 11-4. Measured and calculated viscosities 01two Oklahoma crude oils at
Figure 11-3. Measured and calculated viscosities at f = 71.1 e 01 the liquid
atmospheric pressure. The measured data are tabulated in Table 11-4.
phase 01oil mixture 3 lor which the composition is given in Table 11-2. The mea-
sured data are tabulated in Table 11-3.
SRK-equation and the procedure of Peneloux (both of which are described
Table 11-4
Measured Vlscosities in Atmospheric Pressure of Oklahoma Crude Oil
in Chapter 5). The results obtained with the procedure of Pedersen et al.
Boiling Point Fractions (Baltatu, 1984) agree ver y well with the experimental observations. Larger deviations are
found using the procedure of Lohrenz et al. The results obtained using den-
Average Boiling
Exp. Kinematic sities calculated from the Peneloux procedure are found to be worse than
API Gravlty 01 Oil Temperature Viscosity those obtained using densities calculated from the SRK-equation. This is
Point 01 Oi!
Sample (Oe) (mm2/s)
Sample (Oe) surprising, beca use the Peneloux liquid densities are known to be more accu-
0.0 0.770 rate than the SRK densities (see Chapter 9).
10.0 0.692
Presented in Table 11-4 are measured viscosities for two Oklahoma oil
20.0 0.627
87.6 63.3 0.567
boiling-point fractions (i.e., mixtures of hydrocarbons with the boiling
30.0
40.0 0.523 points within a narrow interval) at atmospheric pressure (Baltatu, 1984).
50.0 0.484 The results are plotted versus temperature in Figures 11-4, which also shows
0.0 3.020 the results obtained using the procedures of Pedersen et al. and Baltatu. The
10.0 2.420 calculated resultsof both procedures are in good agreement with the mea- .
20.0 1.990 sured results.
30.0 1.741
42.3 40.0 1.440
212.6
50.0 1.256
60.0 1.100 REFERENCES
80.0 0.880
100.0 0.738
Baltatu, M. E., "Prediction of the Liquid Viscosity for Petroleum Frac-
API gravity = 141.5 - 131.5/specific gravity. tons,' Ind. Eng. Chem. Process. Des. Dev., 21, 1982, pp. 192-195.
3
Kinematic viscosity (mm2/s) = dynamic viscosity (cP)/density (g/cm )
 186 Properties of Oils and Natural Gases

Baltatu, M. E., "Predicton of the Transport Properties of Petroleum Frac-

tions," paper presented at the 1984 Winter National Meeting AIChE, At-
lanta, Ceorgia, 1984.
Ely, J. F. and Hanley, H. J. M., "Prediction of Transport Properties. 1. Vis-
cosity of Fluids and Mixtures," Ind. Eng. Chem. Fundam., 20, 1981, pp.
323-332.
Hanley, H. J. M., McCarty, R. D., and Haynes, W. M., "Equation for the
Viscosity and Thermal Conductivity Coefficients of Methane,' Cry-
ogencs, 15, 1975, pp. 413-417.
Herning, F. and Zippener, L., "Calculation of the Viscosity of Technical Cas
Mixtures fram the Viscosity of the Individual Cases," Gas u. Wasserfach,
79, 1936, pp. 69-73.
Hirchfelder, J. O., Curtiss, C. F., and Bird, R. B., Molecular Theory of
Cases and Liquids, John Wiley & Sons, New York, 1954.
Jossi, J. A., Stiel, L. 1., and Thodos, C., "The Viscosity of Pure Substances
in the Dense Caseous and Liquid Phases,' AIChE l., 8, 1962, pp. 59-63.
( Lohrenz, J., Bray, B. C., and Clark, C. R., "Calculating Viscosities of Res-
ervoir Fluids from Their Compositions,' l. Peto TechnoZ., Oct. 1964, pp.
1171-1176.
McCarty, R. D., "A Modified Benedict-Webb-Rubin Equation of State for
Methane Using Recent Experimental Data," Cryogencs, 14, 1974, pp.
276-280.
Mo, K. C. and Cubbins, K. E., "Conformal Solution Theory for Viscosity
.and Thermal Conductivity of Mixtures," MoZ. Phys., 31, 1976, pp. 825-
847.
Pedersen, K. S., et al., "Viscosity of Crude Ol,' Chem. Eng. Sc., 39, 1984,
pp. 1011-1016.
Pedersen, K. S. and Fredenslund, Aa., "An Improved Corresponding States
Model for the Prediction of Oil and Cas Viscosities and Thermal Conduc-
tivities,' Chem. Eng. Sc., 42, 1987, pp. 182-186.
Riazi, M. R. and Daubert, T. E., "Smplify Property Predictions," Hydro-
carbon Process., 59, 1980, pp. 115-116. .'
Stiel, L. 1. and Thodos, C., "The Viscosity of Nonpolar Cases at Normal
Pressures,' AIChE i.. 1961, pp. 611-615.
Tham, M. J. and Cubbins, K. E., "Correspondence Prnciple for Transport
Properties of Dense Fluids, Nonpolar Polyatomic Fluids,' Ind. Eng.
Chem. Fundam., 9, 1970, pp. 63-70.
Chapter 12
Thermal Conductivity
When a temperature gradient exists in a fluid, the fluid molecules will try
to eliminate the temperature gradient by transferring energy from regions of
higher temperature to regions of lower temperature. The thermal conduc-
tivity is defined as the proportionality constant, A, in the following relation
(Fouriers law):
(12-1 )
q = -A (~)
where q is the heat flow per unit area, and (dT/dX) is the temperature gra-
dient in the direction of the heat flow.
Most of the procedures that have been suggested for calculation of the
thermal conductivity of hydrocarbon mixtures (e.g., Reid et al., 1987) are
only applicable to mixtures of well-defined components, or may only be
used for either gas or liquid phases. This chapter deals with calculation pra-
cedures that are applicable also to petroleum mixtures containing undefined
heavy hydrocarbon fractions, and that give consistent results for both gas
and liquid phases.
KINETIC GAS THEORY
As for viscosity (see the Kinetic Cas Theory section in the previous chap-
ter), one may for the thermal conductivity, A, of a dilute gas, derive an ap-
proximate analytical expression. It is given by:
1
A=-nvLC (12-2)
3 v

where n = number of molecules per unit volume

v = average molecular speed
C, = molar heat capacity at constant volume
L = mean free path between two molecules

187
 ~\
!/rflY"" ~f'~

188 Properties of Oils and Natural Gases Thermal Conductivty 89/ ..... \\

( ~ 1 nIIUQ1Et;A ) ~
If it is assumed that the average speed, v, is proportional to (RT/MW)1I2, where Po = P Peo/Pex '. ~\ __ ~
the mean free path proportional to 1I(na2) where a is the molecular hard- r, = T Teo/Tex ~ V$\:-,.. ~?;"'"
sphere diameter, and C, is proportional to the molecular weight, the follow- "o = thermal conductivity of the reference substance at the tempera~~.1' '-,
'i:;" 'r 41
o'J
ing equation may be derived: ture T o an d pressure P o '--,.."

As was the case in the discussion of viscosity, methane appears to be a con-

" = constant x T1I2/(MW1I2a2) (12-3)
venient choice as a reference substance, however, some corrections must be
introduced as compared with the simple corresponding states principie ex-
If it is furthermore assumed that a3 is proportional to the critcal molar
pressed in Equation 12-7. The thermal conductivity of poi yatomic sub-
volume, Ve, and that Ve is proportional to RTe/Pe, the following expression
stances (Hanley, 1976) can be separated into two contributions, one due to
may be derived for " at the critical point:
transport of translational energy, and one due to transport of internal en-
"e = constant x P~/3/(MW1I2Te1l6) (12-4) ergy:

This expression is used in the calculation of the thermal conductivity using x= "tr + "int (12-8)
the corresponding states theory. .
Christensen and Fredenslund (1980) have suggested that the correspond-
(
ing states theory should be applied to the translational term only. They use a
term a to correct for the deviations from the simple corresponding states (
model. The final expression for calculation of the thermal conductivity of a
(
CALCULATION OF THE THERMAL CONDUCTIVITY USING THE component mixture at the temperature, T, and the pressure, P, is the follow-
ing: (
CORRESPONDING STATES THEORY
(
"mix(P,T) = (Te,mix/Teo) -1/6 (Pe,mix/Peo)2/3
(MW mix/MWo) -1/2
The most accurate procedures for calculation of thermal conductivities of (
X (amix/(l!o)("o(To,Po) - "int,o(To)) + "int,mix(T) (12-9)
hydrocarbon mixtures are those based on a corresponding states method sim-
(
ilar to that used for viscosity calculations (see Chapter 11). In terms of pres-
sure and temperature, the corresponding states theory applied to thermal where T; = ,a,)
TI T e,mlXmlX and Po = PI (pe,mlXmlX
,a,)
(
( Teoao Peoao
conductivity can be expressed as follows:
, (
Te,mixand Pe,mixare found from Equation 11-27 and Equation 11-30, re-
x, = f(P" Tr) (12-5)
spectively. The critical temperatures Te, and critical pressures Pe, of the indi-
vidual components are found using the characterization pracedure of Peder-
where f is the same function for a group of substances obeying the corre-
sen et al. described in Chapter 6. The mixture molecular weight MW mixis
sponding states principie. For the reduced thermal conductivity, "" the fol- (
lowing equation is used: found from the Chapman-Enskog theory as described by Mo and Gubbins
(1976):
"r(P, T) = ,,(P, T)/(Te-1I6P~/3MW-1I2) (12-6)

The nominator of the expression at the right side of Equation 12-6 may be
MWmix = ~
8
[E Ei j
(ziZj(lIMWi + lIMWj)1I2(TeiTej)1/4)1

regarded as an approximation of the thermal conductivity at the critical -2

point (see Equation 12-4). Using the simple corresponding states theory, the
thermal conductivity of component x at the temperature T, and the pressure
P, may be found from the following equation:
"x(P, T) = (Tex/Teo)-1I6(pex/Peo)2/3(MWx/MWo)-1I2,,0(Po, To) (12-7)
((Te;lPci)1I3+ (Tc/PCj)1I3)2 Te~~~~P::!iX (12-10)
]
where Z are mole fractions, and i and j component indices. The interna! en-
ergy contributions to the thermal conductivity ("int,o, of the reference sub-
stance, and "int,mixof the mixture) are both given by:
 Thermal Conductivity 191
190 Properties o} Oils and Natural Gases
j4 = 72.835897919
Aint= 1.18653 7}(C~ - 2.5R)f(Pr)/MW (12-11) ~ = 0.74421462902
(
j6 = - 2.9706914540
f(Pr) = 1 + 0.053432 Pr - 0.030182 P; - 0.029725 P~ (12-12) h = 2.2209758501 X 103
(

( 7}is the gas viscosity at the actual temperature and a pressure of 1 atm, C~d
the ideal gas heat capacity at the temperature T (qd is found as described in
( the sections on the SRK Equation of State and the Ideal Gas Heat Capac-
ity). R is the gas constant. The correlations 12-11 and 12-12 are founded on
the assumption that the internal energy contribution to the thermal conduc-
( tivity is closely related to diffusion of molecules possessing an internal energy
( deviating from the average level. Tham and Gubbins (1970) have reported
o-values for N2, CO2, and the hydrocarbons < C7 Pedersen and Fre-
( denslund (1987) extended the model to cover heavy hydrocarbons also. They
( use the following expression for a of component i, derived from thermal
conductivity data:
( results were obtained for LlA"(p, T):
(
(
ai = 1 + 0.0006004 p;;043
Pedersen and Fredenslund (1987) have introduced a "low temperature
term" in Equation 12-16, paralleling that in the discussion of viscosity (see
Equation 11-41). Their final expression for the thermal conductivity of
methane is the following:
A(p,T) = Ao(T) + A(T)p + FLlA' (p,T) + F2LlA"(p,T)
+ LlAc(p,T) (12-17)
F and F2 are defined in Equation 11-41. Based on thermal conductivity
data measured at reduced temperatures below 0.4, i.e., corresponding to a
methane reference state, where methane is in' a solid form, the following

where Pri =
MW~OB6

Po(T Tco/Tci, P Pco/Pci)/Pco

(12-13)

(12-14)
LlA"(p,T) = exp(I + 14/T) [ exp (pO (12 + ;~2)
(
( amix is found using the following mixing rule: + 8p" 11, + ~ + ~)) - 101 (12-18)

(
amix = EE ZZj(aiaj)O.5 (12-15) where 1 = - 8.55109
( i j 12 = 12.5539
13 = - 1020.85
( Using this mixing rule, the components having small o-values, i.e., small
14 = 238.394
molecules, are attributed more importance than those having larger a-val- 15 = 1.31563
(
ues. The smaller molecules are more mobile than the larger ones, and
16 = - 72.5759
( thereby contribute relatively more to the transfer of energy than do the 17 = 1411.60
larger ones. The calculation of the thermal conductivity of the reference
(
substance, methane, is based on a model of Hanley et al. (1975) which has
( the following form: Baltatu et al. (Baltatu, 1984; Baltatu et al. 1985) have suggested that the
thermal conductivity is calculated from a somewhat different corresponding
A(p,T) = Ao(T) + A(T)p + LlA'(p,T) + LlAc(p,T) (12-16) states model than the one previously described. They also calculate the ther-
mal conductivity of a mixture as the sum of a translational and an internal
In the dense liquid region the major contribution to this expression comes contribution:
from LlA'(p, T), which has the same functional form as the expression for
Ll7}' (p,T) given in Equation 11-38. The coefficients h-h have the following Amix(p,T) = A: (p,T) + A~ix(T) (12-19)
values:
\
A: (p, T) is the translational contribution to the thermal conductivity of a
h = - 7.0403639907 hypothetical pure fluid, x, with the density, p, and at the temperature, T.
( h = 12.319512908 A~ix(T) is the internal contribution to the thermal conductivity of the mix-
3 = - 8.8525979933 X 102
 Properties 01 Oils and Natural Gases Thermal Conductivity 193
192

ture at the temperature, T. The translational contribution is found by apply- a ~

ing the corresponding states principie: ;::
el)
~I(")I"'IC\!
e r- __ C\l
C\!I~
IX) (") M
~
C\!
Ol ~
'"~ .,
:f I I I I

(12-20)
A;(p,T) = A~(Po,To)FA ~~-i-+-t-t--+-~-+--~~
Gil
where A~(Po,T o) is the translational contribution to the thermal conductivity e Sl
5l:u ..<.. ~IC\!I~ ~If'-: a Ol "!~
of the reference substance, o (methane), at the temperature To, and the den- (") IX) Ol Ol U') ci M
(") IX) ~ a C\J a IX) ~
.. ~ ~ ~ ~ ~
'C :::
sity Po' FAis found from the expression: cf
..<
FA = (MW o /MW x )112 fl/2 h X,O
-2/3 (12-21) 11I
X,O
Q) ~I~I~I~
e a
'<l:1f'-: C\J
M
IX)
M
<O
N
.,
~ C\J
:;: IX) Ol
I ~ ~
-s. I
I

In this equation: ~::::J el)

"O
~I I I I I I I I I I
fx,o= h~; E E zzjfjhj
i j
(12-22)
O
cu
e
o
~I E
!::
81 ..< ~
Sl
f'-:1c>1~
(") Ol
'<t
i'- i'-
C\J
;;;I~
~ (")
r-,
,...: ~
U')
Ol
Ol
(")
C\J
(
E
Q)
.. ~ ~ ~ ~ ~
(
and hx,o = E E zzjhj (12-23) s:
....
i j "O

Q)

... -
; ,.<? '<t a
where fj = (ffj)1/2 (12-24) ::::J E f'-:Ic> '"
N U
... cu
c..
>e-
w::: ~I"'I~
U') Ol
(") IX)
~ ~
<O
~
Ol
~
~
C\J N
~
N
Ol
ci
~
Q)O ..
1 (12-25) :c"O
and h 1).. = -8 (h~/3
1
+ h/3)3)
~ 111
e
:;
en

The parameters f and h are, for component i, calculated as follows:

~
E
.:
Q) I-g ~I~I~
(")
i'-
C\J
(")
IX)
(")
<O
Ol
C\J
~I~
(")
(") Ol
(") C\J
(")
(")

M
(")
(")
(")

M
~
(")

c
Ol
(")
(ij

E
~
E
ID

.~
f = 8 Te/Teo (12-26) ~

-s >< x
W ~
(12-27) o -~> f'-:IC\!I~ "'I'<l:
U') IX) IX)
,...:
=-
:;
h = <I>
Vd/Veo Q,
I!
U') i'- i'- U')
'<t
U')
(") ~ ~ ~
en
~
e U') U') '<t (")

CJ (ij
.: E-
where 8 ando<I>are shape factors identical to those entering into Equation 111
U1Q)~
Ea>
11-16 for calculation of the viscosity. Te is the critical temperature, and Ve a. ~.~ :::.
'<l: r-,
E '<l:1~ C\! ,...: ~ a. ::l
the critical molar volume. The properties are found from the correlations o !'Sl c>1~1~
U') v
IX)
Ol
<O
'<t IX)
a C\J
~
(")
<O
IX) <O
I ~ 1
O U')
(") (") '<t U') U') U') U') <O ~I~
mentioned in Chapter 11 under the Corresponding States Theory heading as '>ro !:: al
::J ..
those used by Baltatu to extend the viscosity model of Ely and Hanley to c,~~
U ~ "O
"O _
apply to mixtures containing heavy hydrocarbons. The density Po, and the ti c~ ~ ~
;::

u
temperature T of Equation 12-20 are found as follows: :::J
~ s g--5 ~
'C
al a ~,
~ al r-, U)
Q)~
c: al E .~ tial_ ~~~ .~

To = T/fx,o (12-28)
Q,
~ c:h 1-_
al"C
::J
(/l
o ~
e Qio IO
c::: . o Q)
.- a.
<t- -

Po = phx,o; E '" al- QI

c: ~ al- '><" u ~
~ ::: x ~
:::J 8
~'~IO e~ ~ E
Q)
...J c: al c:
Q)::E'" ~ ti Ql-
QI _ ::J ~~ >-
The internal contribution to the thermal conductivity of Equation 12-19 ~ eID g e E ~ (/l
(/l al ::J
(/l u E ~
ctS
E'~ 11
g ~ >
T""" U
:;

is calculated as follows: ~ Q,
tU
(!J
al
:.::
el::J
::J ~
'"
(!J
ala
:'::C\J
.!!!~
o~ e '"
<~ JI~~~
~

I I I I I I I I I I~~ ~ 8
 194 Properties of Oils and Natural Gases
Thermal Conductivity 195

A,~x(T)= 1: 1: zz/\j(T)
j
(12-29)
Christensen, P. L. and Fredenslund, Aa., "A Corresponding States Model for
the Thermal Conductivity of Gases and Liquds," Chem. Eng. Sci., 35,
where Aj(T) = 2 [l/A "(T) + l/A "(T)]- 1 (12-30) 1980, pp. 871-875.
Hanley, H. J. M., McCarty, R. D., and Haynes, N. M., "Equation for the
For apure component, i, the internal contribution to the thermal conduc- Viscosity and Thermal Conductivity Coefficients of Methane,' Cry-
tvity is found from the expression: ogenics, 15, 1975, pp. 413-417.
Hanley, H. J. M., "Predction of the Viscosity and Thermal Conductivity
Coefficients of Mxtures,' Cryogenics, 16, 1976, pp. 643-651.
A"(T)= 1.32 (C:~i (T) - ~ R)r,(T)/MW (12-31) Mo, K. C. and Gubbins, K. E., "Conformal Solution Theory for Viscosity
and Thermal Conductivity of Mxtures,' MoZ. Phys., 31, 1976, pp. 825-
where C~~i = ideal gas heat capacity of component i at the temperature T 847.
(C~ may be found as described in the section on the SRK- Pedersen, K. S. and Fredenslund, Aa., "An Improved Corresponding States
Equation of State in Chapter 5, and the Calculation of the Model for the Prediction of Oil and Gas Viscosities and Thermal Conduc-
Ideal Gas Heat Capacity, in Chapter 7.) tivities,' Chem. Eng. Sci., 42, 1987, pp. 182-186.
R = gas constant Reid, R. C., Prausnitz, J. M., and Poling, B. E., The Properties oj Gases and
r' (T) = dilute gas viscosity (see section on Viscosity Correlation of Liquids, McGraw-Hill, New York, 1987.
( Lohrenz et al. in Chapter 11) of component i at the tempera- Tham, M. J. and Gubbins, K. E., "Correspondence Principie for Transport
ture T Properties of Dense Fluids. Nonpolar Polyatomic Fluids," Ind. Eng.
(
MW = molecular weight of component i Chem. Fundam., 9, 1970, pp. 63-70.

COMPARISON WITH EXPERIMENTAL OBSERVATIONS

It has not been possible to find thermal conductivity data for petroleum
reservoir fluids or any other rnult-component hydrocarbon mixtures. The
most interesting data are those published by BaItatu (1984), covering ther-
mal conductivity data for petroleum fractions, and each containing hydro-
carbons with the normal boiling point within a narrow temperature inter-
val. Some of these data are shown in Table 12-1, in a comparison with
resuIts obtained using the calculation procedures of Baltatu et al., and Pe-
dersen et al. Calculated results of comparable quality are obtained using ei-
ther of the two calculation procedures.

REEERENCES

Baltatu, M. E., "Prediction of the Transport Properties of Petroleum Frac-

tons," paper presented at the 1984 Winter National Meeting AIChE, At-
lanta, Georgia, 1984.
Baltatu, M. E. et al., "Therrnal Conductivity of Coal-Derived Liquids and
Petroleum Fractions,' Ind. Eng. Chem. Process Des. Dev., 24, 1985, pp.
325-332.
 r Surface Tension 197

(Xc is the Riedel parameter (Riedel, 1954):

I
I (Xc = 0.9076 (1 + TBr In Pc)/(l - TBr) (13-4)

P, is the critical pressure in atm, T, the critical temperature in Kelvin,

Chapter 13 and TBr the reduced boiling point (TB/Tc)' For mixtures, T, and P, will not
always be known, and it is also questionable whether the surface tension of
Surface Tension a mixture is related to the critical properties in the same way as is the surface
tension of apure component. For a mixture, it is more convenient to express
the surface tension in terms of phase properties, which can be measured, or
accurately calculated; as can, for example, phase densities. To transform the
Knowledge of the oil/gas surface tension is needed to understand the
temperature dependence to density dependence, the following approxima-
pore-level flow processes in a reservoir. The molecules located in the surface tion is used (Fisher, 1967):
of a liquid in equilibrium with a gas are exposed to forces differing from
those acting between molecules in the bulk liquid or gas phases. The attrac-
PL - Pv = B'(~c - T)1i (13-5)
tion from a neighboring gas molecule is less than the attraction from a
neighboring liquid molecule. The surface layer is in tension and tends to
where PL and Pv are the liquid and vapor phase densities, respectively. {3 is a
contract to the smallest area compatible with the mass of material, con-
constant, which, for most pure components, has a value between 0.3 and
tainer restraints, and external forces. The surface tension, a, is defined as
0.5. The coefficient B' has the dimension (density/temperature). It is trans-
follows: ~
forrned into a dimensionless constant B by rewriting Equation 13-5 to (Lee
and Chien, 1984):
a = (aGlaAh,v,N (13-1)
(PL - Pv)1 Pc = B(l - Tr)1i (13-6)
where G = Gibbs energy
A = surface area
where Pc is the critical density, and T, the reduced temperature (T/Tc)' Sub-
T = temperature stitution of Equation 13-6 into Equation 13-2 gives:
y = molar volume
N = number of molecules !J.IiIOy _ )
alilO = '''c_ c (PL - Pv (13-7)
B
The surface tension can be interpreted as the resistance from the liquid
against an imposed increase in the surface area.
where Yc is the critical molar volume, and PL and Pv are the molar densities
of the liquid and vapor phases, respectively. The coefficient:
ESTIMATION OF THE SURFACE TENSION USING THE
NgIOyc
MACLEOD-SUGDEN CORRELATION
B

Brock and Bird (1955) have shown that the surface tension of pure non-
is called the Parachor [P]. Equation 13-7 can then be rewritten to:
polar compounds can be approximated by the following relation:
alilO = [P](PL - Pv) (13-8)
a = Ac (l - Tr)O (13-2)
which is often referred to as the Macleod-Sugden relation (Macleod, 1923,
where (J = 11/9, T, is the reduced temperature ( = T/Tc) and
and Sugden, 1924). The equation can be extended for application to rnix-
Ac = p~/3 n /3 (0.133 (Xc - 0.281) (13-3)
tures, as proposed by Weinaug and Katz (1943) who use {31(J = 1/4:
 Surjace Tension 199
198 Properties of Oils and Natural Gases

Table 13-1
al/4 = .E (,odP]x - ,ov[PJy) (13-9) Values ot the B-parameter (Lee and Chien, 1984) Entering into Equation
i 13-11 tor Determining the Parachor

where i is a component index, and x and y are liquid and vapor mole frac- Compound B-parameter
tions. Water 4.310
A second modification of Equation 13-8 with application to mixtures has Carbon dioxide 3.505
been presented by Lee and Chien (1984). The surface tension is found from Nitrogen 3.414
Methane 3.403
the following expression:
Ethane 3.591
Propane 3.602
al/4 = ,odP]L - ,ov[P]v (13-10) n-Butane 3.652
n-Pentane 3.690
The liquid and vapor phase Parachors, [P]L and [P]v are found from the n-Hexane 3.726
expressions:

A1/4 A1/4 Ve RTe/[Pe(3.72 + 0.26(ae 7.0))] (13-12)

[PJL = eL X VeL [P]V = eV X VeV (13-11)
= -
B ;
Bv
L
and that the Parachor of the C7 + -fraction is calculated from the following
where VeL = .E XVe equation (Nokay, 1959):
i

IOglO[P]= - 8.93275 + 3.68894 logo(Te/Seo6676) (13-13)

BL = .E xB
Te is the critical temperature in R, and se the C7 + -specific gravity
(60F/60F). Knowing fP] it is possible to determine B of the C7+ -fraction
2/3 1/3
AeL = PeL TeL (0.133 aeL - 0.281) by rearranging Equation 13-11:

PeL = .E xPcj ,. B = A~/4Ve

(13-14)
i [P]
TeL = .E xTcj The liquid and vapor phase Parachors may now be calculated from Equa-
i tion 13-11. If the phase densities are not known experimentally, they can be
calculated as shown in Chapter 5, and the surface tension may finally be
aeL = .E xacj calculated from Equation 13-10.
i

x stands for liquid mole fractions, and i is a component index. The same
mixing rules are applied to the vapor phase, i.e., Xis replaced with the va-
por mole fraction, y. B-values for a number of oil and gas constituents are
given in Table 13-1. The critical properties of the same components are
given in Table 5-1. The properties entering into Equation 13-11 for the C7 +-
fractions may be determined using the C7 + -characterizaton procedure of
Pedersen et al. described in Chapter 7. Te and Pe of the total C7 + - fraction
are found as molar averages.
Pedersen et al. (1988) have suggested that the critical molar volume of the
C7 + -fraction is estimated from the following equation of Riedel (1954):
ESTIMATION OF THE SURFACE TENSION FROM THE VISCOSITY
It has been experimentally observed, e.g., by Pelofsky (1966) and by Pe-
dersen et al. (1988), that for pure hydrocarbons, and for oil mixtures at at-
mospheric pressure, there is an almost linear dependence between the recip-
rocal of the viscosity, r, and the logarithm of the surface tension:
In a = In A + B/r (13-15)
 200 Properties of Oils and Natural Gases Surface Tension 201

where A and B are constants depending on composition only, Pedersen et al.

have suggested that the A and B parameters entering into Equation 13-15
~ r-,
for a hydrocarbon mixture should be determined from the following rela- Cl ~ Cl
cD
lO
M
(5 e ~, I io
tions: 1'1:I
GI

E W[(Clp) + (C2N) + (C3A)] en

~~ I I
A =

(13-16)
-
.c

z
..
o ~
.,zEle
E
,...: C\I
r-,
oi
C\I
lO
,...:
C\I
~I<':!lO
<O
C\I C\I
al
C\I
o
,...:
C\I
v
q:
C\I

B E W[(C4pMW~7) + (CsNMW?) + (C6AMw~g)] (13-17) o

-
=

~
en v ~

;r:J
lO C\I f'-.
O
O
~ N cD
~
,...: ~
..
GI 01 I
where i = a given carbon number fraction :::1 I

w = weight fraction
ti)
ti)
..
GI
o..
I I
MW = molecular weight ~
U ., ZElf'-.
,...:
f'-. v <OI~
~
O al '<1:
,...:
.;: O <Xi f'-. lO lO
GI ., E C\I C') C\I C\I C\I C') C\I C\I
.c
Q.
Table 13-2 ti)
O
Coefficients Entering into Equations 13-16 and 13-17 for Calculating :> en al <O
E ~ al N M :::1 1 <Xi
lO
oi
C\I
,...:
Surface Tension of a Hydrocarbon Mixture (Pedersen et al., 1988) <;j ~
(Surface Tension is in mN/m ano, 1) in cP)
M ti) ti)
~~I
11
I
Numerical Value I e
.2 ..
GI ~ o
EIC\I
Z cri
~ C')
oi ~I<':!
en <O
C\I ~
en
C\I
<Xi
Coefficient M c;:;
:::1
ti) E C\I C\I C\I C\I C') C\I C\I

e, 3.118 x 10 GI e )(
e2 2.805 x 10 :a~:E
!!! GI
x tq r-, o r-,
e, 3.673 10
.e..
U C\I

~t!J ~ ~I 1 <Xi oi q:
N
- 3.282 x 10-2 ~ ~
e,
e, - 6.096 x 10-6 :::1
en
..
I I I
e,
e-
e,
eg
-
-
6.098
3.641
2.191
1.229
x
x
10-4
10-'
--
o
ti)

:ti);
11 I . I <':!
~ "z al
E C\I
r-,
C\I
C')
o
C')
'<I:IC\I
al
C\I
f'-.
C\I
N
C')
r-,
al
C\I
lO
<Xi
C\I

GI
a:
e in o
'O
GI
al
el> ;:;;
al
el> ~ ~
lO

el>
el> el>
;j ~c:: ~
P, N, and A are the mole fractions of paraffinic, naphthenic and aromatic :; 'O I ci. c.i c:: c.i .2 ~ I ci, uc.2c: .2~
components, respectively. The recommended values for the coefficients Cl- U
'l:I fII
e al
!;2II:
>< "j .Q "j O
W
010
::l
o .;:::0';:::
'" O

'rr" '"
><
W (3- ~::l
o '"
0 ~ (3 g
::l '"
O

g. ~
I a.
)(
Cgare given in Table 13-2. At higher pressures, where the vapor phase contri- ~ W
cr ::l ::l
rr rr rr
W a.
'O W W W W
I bution must be taken into 'consderation, Pedersen et al. recommend that the e
x
Q)
111 1'1:I
following formula be used:

ln a = ln A + BI Ll1'/ (13-18)
-
ea
e
GI
E
.;:
..
al
::l
)(
c::
O
:;:
'00
Q)
:o v
c::
O
"';:::; ID
"Ci515 ll)
u
Cii
.!.

o
x
8.~ el> 8.~ '" o
GI
Q. :li Ec::~ E c:: ~
o.s 0.S)2 >< O .- O .- :>
where Ll1'/ = ( 1'/L - 1'/v W <5 o o Q)
O
i I I I~

The subindices L and V refer to the liquid and vapor phases, respectively.
202 Properties of Oi/s and Natural Gases
Surface Tension 203

COMPARISON WITH EXPERIMENTAL OBSERVATIONS In the literature, experimental surface tension data are scarce for petro-
leum mixtures at elevated pressures. Simon et al. (1978) presented high-pres-
Table 13-3 gives experimental and calculated results for two North Sea oil
sure data for surface tensions of a reservoir fluid to which considerable
mixtures. Calculated results of using Equations 13-9, 13-10, and 13-15, re-
amounts of carbon dioxide were added. The composition of the reservoir
spectively, are shown. The best results are obtained with Equations 13-15
fluid is given in Table 13-6, and the measured results for the surface tensions
and 13-10. The logarithm of the experimental surface tension is plotted
are presented in Table 13-7. Lee and Chien (1984) have calculated the sur-
against the recprocal of the viscosity in Figure 13-1.
face tensions using Equations 13-9 and 13-10, respectively. Their results are

Table 13-4
Table 13-5
Molar Composition of on Mixture for whlch Surface Tension Results are
Molar Composition of 011 Mixture for whlch Surface Tenslon Results are
Shown in Table 13-3
Shown in Table 13-3

Molecular Density, (g/cm3)

Molecular Density, (g/cm')
Component Mole % Weight at 15C, 1 atm.
Component Mole % Weight at 15C, 1 atm.
C,-Cs 3.70 51 - C,-Cs 12.35 61.7 -
C6 3.25 80 0.644
C6 4.66 82.0 0.685
C7 8.38 93 0.733
C7 11.76 98.7 0.737
C8 7.28 107 0.763
C8 7.96 109.6 0.754
C9 7.28 120 0.785
C9 7.84 121.9 0.774
C'O 5.26 131.6 0.808
C1Q 5.73 133 0.789
Cll 5.17 147.6 0.815
Cll 5.06 148 0.794
C'2 4.44 158.6 0.836 ~
C'2 4.72 163 0.806
C'3 5.11 172.4 0.850
C'3 4.07 177 0.819
C'4 4.58 184.8 0.861 4.41
C'4 190 0.832
C'S 4.29 197.4 0.873
.~ C'S 3.50 204 0.834
C'6 3.45 209.0 0.882
C'6 2.59 217 0.844
C'7 3.53 226.5 0.873
C17 3.06 235 0.841
C'8 3.01 242.6 0.875
C'8 2.15 248 0.847
C'9 2.64 253.7 0.885 2.45
C'9 260 0.860
C20 2.51 262.0 0.903
C20 0.81 269 0.874
C2, 1.87 281.1 0.898
C2, 1.98 283 0.870
C22 1.97 293.2 0.898
C22 1.31 298 0.872
C23 1.60 307.3 0.899
C23 1.10 310 0.875
C24 1.56 319.6 0.900
C24 0.84 322 0.877
C2S 1.38 333.3 0.905
C2S 1.19 332 0.881
C26 1.19 345.9 0.907
C26 0.96 351 0.886
C27 1.38 361.1 0.910
C27 0.76 371 0.888
C28 1.30 373.5 0.915
C28 0.81 382 0.895
C29 1.26 380.8 0.920
C29 0.59 394 0.898
C30+ 12.80 624.0 0.953
C30+ 7.33 612 0.935
 'pamseaw(pamseaw - paiajnojao x 001 = UOIeIAaa %

Z'v ~ 588'0 ZZZ - 809'0 5LL'0 8' ~OZ 17'175 OZ/08

6'8 - 888'0 0'917 - 9617'0 6~6'0 ~'99~ 17'175 OZ/08
["6 - ~66'0 5'917 - L85'0 L60' ~ Z'68~ 17'175 OZ/08
v' ~ - 8ZV'0 6'58 - 8Lc:'0 17817'0 6'L8~ 17'175 517/55
UO!le!lI<la % O ~ -E ~ UO!lenb3 UO!le!/laa % 6-E ~ UO!lenb3 paJnseaw (req) d (::>0) 1 I!O/"O::>
palelnole::> palelnoleo O!leH 6u!x!W
(WlNw) UO!SUal aoe,JnS

ap!xo!o uoqJeO pue (9-E ~ alqe.l U! uoursoduiorj I!O J!oI\Jasa~e ~osaJnlX!W

~ouojsuej aoepns aln JO~("86~ 'Ua!4:> pue aal) sllnsa~ palelnole:> pue (8L6~ 'Ie la UOW!s) pamseaw
z-cr alqe.l

"-
u

e
-- e

~
E
u
C,
e
.2 l!)

c.; t-, N
~ io r::. N <O <') N en O
-e-
al
<D

~ ~
O O f'-. <t <') <O
O 00 <')
r-,
lO
N <') o
at'" -r- <t N lO <t N N
O ~ II
~ <') ~ ~
O O
O ~ O O O O O O
6 6
G> O 6 6 6 O O O O O O O O
e
"O
::E .,
oc.
:
.D
:J
.D

1i
Z-

"
o
.,
;

c:
o
~ -e
VI .),

o
o-

l5 '"
U
.D
e
/e
...'" ....'" G>
c:
o
Q.

oo '" ~
+
E
u
o
eS .s .s c3 c3 c3'
o

L- L-------~----~~
(W/NW Ul D) (D)U
206 Properties of Oils and Natural Gases
Surface Tension 207
Table 13-8
Molar Composition of Reservoir Fluld for which Surface Tenslon Results REFERENCES
are Given In Table 13-9

Brock, J. R. and Bird, R. B., "Surface Tension and the PrincipIe of Corre-
Component Mole %
sponding States," AIChE t.. 1, 1955, pp. 174-177.
N2 0.03
Firoozabadi, A., Katz, D. L., Sonoosh, H., and Sajjadian, V. A., "Surface
CO2 2.02 Tension of Reservoir Crude Oil-Cas Systems Recognizing the Asphalt in
C, 51.53 the Heavy Fraction," SPE 13826, SPE Reservor Engneerng, 3, 1988,
C2 8.07 pp. 265-272.
C3 5.04 Fisher, M. E., "The Theory of Equilibrium Critical Phenomena," Report on
-C. 0.83 Progress in Physics, A. C. Stickland, editor, 1967, p. 616.
n-C. 2.04 Lee, S.-T. and Chien, M. C. H., "A New Multicomponent Surface Tension
ic, 0.84 Correlation Basedon Scaling Theory," SPE/DOE Fourth Symposium on
r-e, 1.05
Enhanced Oil Recovery, Tulsa, Oklahoma, April16-18, 1984, pp. 147-
158.
C6 1.38
Macleod, D. B., "On a Relation Between Surface Tension and Density,"
C7 27.17
Trans. Faraday Soc., 19, 1923, pp. 38-42.
e7 molecular weight: 217.
Nokay, R., "Estimate Petrochemical Propertes," Chem. Eng., 66, 1959, pp.
Source: Firoozabadi el al. (1988).
147-148.
Pedersen, K. S., Lund, T., and Fredenslund, Aa., "Surface Tensions of Petro-
leum Mixtures," f. Can. Peto Tech., in press 1988.
Table 13-9
Measured and Calculated Results at T :::;82.2C for Surface Tension Pelofsky, A. H., "Surface Tenson. Viscosity Relation for Liquids,' f. Chem.
(dyn/cm) of Reservoir Fluid of Composition Given in Table 13-8
Eng. Data, 11, 1966, pp. 394-397.
Riedel, L., "A New Universal Vapor-Pressure Equation. 1. The Extension of
Pressure Measured Equation 13-10 Equation 13-15 the Theories of Corresponding States,' Chem. Eng. Tech., 26, 1954, pp.
83-89.
(bar) (mN/m) (mN/m) % Dev. (mN/m) % Dev.
Simon, R., Rosman, A., and Zana, E., "Phase-Behavior Properties of CO -
263.0 1.3 1.3 0.0 1.3 0.0
228.5 2.3 2.0 - 13.0 2.2 - 4.3
Reservoir Oil Systems,' Soco Peto Eng. s..Feb. 1978, pp. 20-26.
2

Sugden, S., "The Variation of Surface Tension with Temperature and Some
194.0 3.3 3.1 - 6.1 3.3 0.0
Related Functions,' f. Chem. Soc., 1924, p. 32-41.
159.6 4.6 4.7 2.2 5.2 13.0
" Weinaug, C. F. and Katz, D. L.,. "Surface Tensions of Methane-Propane
% Dev. = 100 x (calculated - measured)/measured. Mixtures," lnd. Eng. Chem., 35, 1943, pp. 239-246.
Sources: Firoozabadi el al. (1988) and Pedersen el al. (1988).

also given in Table 13-7. Also, in this case, Equation 13-10 is superior to
Equation 13-9. Equation 13-15 is not applicable due to the high concentra-
tion of non-hydrocarbons (C02).
Firoozabadi et al. (1988) have presented experimental results for the sur-
face tension of the reservoir fluid of Table 13-8, for pressures ranging from
160 bar to 263 bar. The results are shown in Table 13-9 in comparison with
the results obtained using Equations 13-10 and 13-15. Positive agreement
with the measured results is obtained with both correlations.
 Wax Formaton and Inhibition 209

CH3 CH2 CH2 CH2 CH2 CH2 CH2 CH2 CH2 CH3
\/\/\/\/\/\/\/\/\/
CH2 CH2 CH2 CH2 CH2 CH2 CH2 CH2 CH2

NORMAL PARAFFIN

Chapter 14 CH2 CH2 CH2 CH2 CH2 CH2 CH2 CH2 CH2
Wax Formation and Inhibition /\/\/\/\/\/\/\/\/\
CH3 CH2 CH2 CH2 CH2 CH2 CH2 CHz CH CH3

\CHz

WAX COMPOSITION
/
CH3

The majority of crude oils contain a certain proportion of heavy hydro-

ISOPARAFFIN

carbon compounds, which may precipitate as a waxy salid phase if the oil is
cooled below a certain temperature. Wax precipitation may take place at
temperatures far above the freezing point of water, and is therefore a poten-
tial problem when petroleum mixtures are transportd in pipelines, e.g., un-
\)\I\/\;\I\lVV\
CH2 CHz CH2 CH2 CH2 CH2 CH2 CH CH2
derwater, or treated in a process planto Wax precipitation may result in plug-
ging of pipes and process equipment. When designing pipelines and
I I
CH2 CH2
separation plants it is, therefore, of importance to be able to determine the
conditions where wax precipitation takes place, and the amount of wax ~~
likely to formo CYCLOPARAFFIN
Wax precipitated from oil mixtures consists primarily of closed packed lat-
tices of aligned paraffinic and naphthenic molecules, shown schematically
Figure 14-1. Typical formulae of petroleum waxes.
in Figure 14-1. The most comprehensive wax-composition analyses have
been made on waxes extracted from petroleum mixtures. They can be di-
vided into two main categories: paraffinic waxes and microcrystalline Table 14-1
waxes. The basic characteristics of each category are shown in Table 14-1 Typical Compositions and Properties of Paraffin and
(Gilby, 1983). Microcrystalline Waxes

Paraffin Microcrystalline
COMPUTAT.ION OF WAX FORMATION Waxes Waxes
Normal paraffins (%) 80-95 0-15
Branched paraffins (%) 2-15 15-30
The chemical potential of component i in a liquid phase at temperature T eycloparaffins (%) 2-8 65-75
may be expressed as follows:
Melling poinl range (Oe) 50-65 60-90
Average molecular
L
POi p,9 weighl range 350-430 500-800
RT = R~ + lo (xn+') (14-1)
Typical carbon number
range 18-36 30-60
eryslallinity range (%) 80-90 50-65
and the chemical potential of component i in a wax (salid) phase as follows
Source: G ilby (1983).
(Prausnitz et al., 1986):
 Wax Formaton and Inhibition 211
210 Propertes 01 Oils and Natural Gases

<I>
is the volume fraction of component i, and is found as follows:
J-Lr_ J-Lf (14-2)
RT - RT + In (s-yr)- .6Hf (1 - T/TD/RT
<l>r = xV s sV
1: xV
1
(14-7)
In Equations 14-1 and 14-2:
1: xV/
<I>-
j
J

i = a component index
V is the molar volume. Won gives values for the liquid phase solubility pa-
J-L0 = a standard state chemical potential
rameters, OL,up to C40. These values are shown in Table 14-2 .
.6Hf = the molar enthalpy of melting
Tf = the pure component melting point temperature
x = the mole fraction in the liquid phase Table 14-2
s = the mole fraction in the wax phase Liquid Phase Solubility Parameters Recommended by Won (1986)
-yL = activity coefficient in liquid phase
-ys = activity coefficient in wax phase bL bL bL
Component (cal/cm3)o .s Component (cal/cm3)o .s Component (cal/cm3)o .s
R = the gas constant.
(
CO2 7.12 C'5 7.96 C29 8.24
( C, 5.68 C'6 7.99 C30 8.25
If at a given temperature the change in Gibbs energy associated with the li~ C2 6.60 C17 8.02 C31 8.26
formation of a wax phase is negative for all possible wax phase composi- C3 6.65
( C'8 8.05 C32 8.27
tions: C4 6.65 C'9 8.07 C33 8.28
( C5 7.02 C20 8.09 C34 8.29

( .6G = I.>M + 1:nrJ-Lr - 1:n1J-L1 ~

i i i
O (14-3) ~~
r
. C6
C7
7.25
7.41
C21
C22
8.11
8.13
Ca5
C36
8.30
8.31
C8 7.53 C23 8.15 Ca7 8.32

no wax is formed, i.e., the temperature T is higher than the cloud point
C9 7.63 . C24 8.17 C38 8.33
C10 7.71 C25 8.18 Ca9 8.34
(Twax)' In Equation 14-3 n means mole number, Z is the mole fraction of C11 7.78 C26 8.20 C40 8.35
\ component i in the total mixture and F refers to the total mixture (feed). C'2 7.83 C27 8.21
The cloud point is defined as the temperature at which the first small C13 7.88 C28 8.22
( C'4 7.92
amount of solid (wax) is formed. At thermodynamic eqnilibrium between a
( liquid and a wax phase, the following equation must be satisfied for all com-
( ponents:
According to Won, the solubility parameters of the components in the
.6J-L/RT (J-Lr - J-Lr)/RT= O (14-4)
= solid phase, os, are found from the following expression:

Won (1985, 1986) uses the following expression to find the activity coeffi-
(or)2 = (oh2 + .6Hf (14-8)
cient of component i in both the liquid phase, and in the wax phase:
V

V(b - D)2
ln-y = RT (14-5) where .6Hf = O.1426MW . Tf (14-9)

MW is the molecular weight of component i.

where D is Hildebrand's solubility parameter of component i (Prausnitz et By comparing Equations 14-1, 14-2, and 14-4, the conditions for thermo-
al., 1986) and dynamic equilibrium of Equation 14-4 can be rewritten:

b= 1: <I>D (14-6) s/x = (-yr/-yf)exp(.6Hf(l - T/TD/RT) (14-10)

 ::'"';~~
~~II\tO~

~~
" tI;- 1
..
q - .
212 Properties 01 Oils and Natural Gases
Wax Formation and Inhibition 213( ~ ( n\tUO~

The cloud points are much toa high when predicted using the model of
i y\DI~-.
replaced by Vin Equation 14-17. The following expression may now be de- ...
~~I~
Won (see section on Comparisons Between Calculated and Experimental rived for the Gibbs energy of mixing: \ 1-, e
Results). To account for that it is necessary to drastically reduce either the
liquid phase activity coefficients or the melting enthalpies. Hansen et al. L\Gmx = kT E E nvXj+ kT E n.Inv, *
(14-18)
(1988) have chosen the former alternative. They have suggested using poly- j i<j i
mer solution theory as described by Flory (1953) for the liquid phase. The
following general thermodynamic relation is used: By partal differentiation of this express ion with respect to ni, it is possible
to derive an expression for the chemical potential of component i in the lq-
L\Gmix = L\Hmix - TL\Smx (14-11) uid phase:

where L\Gmix, L\Hmix, and L\Smix are the changes of mixing of the Gibbs en- (IIJ..
,.,...1 - 119)/RT
r-l = lnv, 1 -t 1
1 - v. - J...
1 " (MW/MW)v.JI J
ergy, the enthalpy, and the entropy. For an ideal mixture: j*i

L\Hmix = O (14-12) + E VjXjk*i

E Vk- E E (MWlMWj)VkVjXkj (14-19)
.. j*i j*i k<j
L\Smx = - k E nln(nl E nj) (14-13)
i j The molecular interactions are divided into three classes:
1. PN-PN
For a real mixture, Flory uses the following expressions: 2. PN-A
3. A-A
L\Hmix = kT Ej E
<j
nvXj; Xii = O (14-14)
where P, N, and A represent paraffinic, naphthenic, and aromatic mole-
cules, respectively. It is assumed that the interactions between paraffinic
where k = the Boltzmann's constant components with other paraffins are identcal to those with naphthenic
Vj = the volume fraction of component j components, and vice versa. The interacton coefficients are found fram the
Xj = a binary interaction parameter between components i and j. following relations, based on experimental wax formation data for real pe-
traleum mixtures (Hansen et al., 1988):
The volume fraction of component i can be approximated as follows:
PN-PN: Xlj = 2.17884 X 1O-3IMW - MWjl (14-20)
v = MWnl E MWjnj (14-15)
PN-A: Xj = - 1. 76960 x 10 - 4 (MW . MWj)
j (14-21)

For the binary interaction coefficients, the following relation applies A-A: Xj = - 8.53350 X 1O-3IMW - MWjl (14-22)
(Flory, 1953):
where the molecular weight of component i, MW, is larger than the molec-
Xj = Xj(MWlMW) (14-16) ular weight of component j, MWj (see aIso Equation 14-16).
Both the liquid and the wax phases have so far been treated as bulk
The entrapy of mixing is for a polymer soluton expressed as: phases, and interfacial contributions are neglected.
According to Mandelkern (1964), however, crystallization consists of two
L\Smix = - k E n.lnv, (14-17) basic processes, nucleation and crystal growth .. Nucleation is a necessary
condition for a crystal to grow. Crystalline structures below a certain size
will be unstable and decompose, because the surface tension of the crystal
The term (nl~nj), employed in the ideal case (Equation 14-13), is contributes to the chemical potental of the crystalline phase. The definition
of the surface tension, as expressed in Equation 13-1, may be used to find the
214 Properties oj Oils and Natural Gases
contribution from the surface tension to the change in Gibbs energy when a
crystal forms. Assuming a spherical crystal of radius r, this contribution
equals:
.::lGo= 47fr20 (14-23)
As the volume of the crystal equals 4/3 7fr3, the corresponding change in
chemical potentialbecomes:
2 The polarity of the polymer increases
= --47fr 0 V = 30 V/r
.::lJl..o (14-24)
4/37fr3
where V is the molar volume of the crystal.
The contribution from the surface tension to the Gibbs energy will, in the
absence of other solid particles, cause supercooling. If the temperature is
lowered, the size of the spontaneously formed crystalline structures in-
creases, and at some stage crystallization takes place. The smallest radius
where the formed crystals are stable is called the critical radius, r".
Hansen et al. (1988) have modified Equation 14-2 to include the contri-
bution from the surface tension:
-'2--s = -'2-- + In s - .::lH{(1 - T/TD/RT + 30 VI(r*RT) (14-25)
RT RT
In this equation, all activity coefficients in the solid phase are assumed
equal to unity, i.e., the solid phase is assumed to be ideal. The cloud point
may now be located by inserting into Equation 14-4 the expression for JI..L
given in Equation 14-19 and the expression for JI..s give~ in Equation 14-25.
The critical radius, r ", is assumed equal to 10-8 m, and the surface tension
of the wax particle is found from Equation 13-15. In the limit, as 1/-+ 00, o
(in Equation 13-15) approaches A, and in the work of Hansen et al. o is as-
sumed equal to the paraffinic value of A (= ci, i~ Equation 13-16).
WAX INHIBITORS
A variety of different additives, usually polymeric, have been shown to
modify the wax crystal size and crystal structure in a way that lowers the
pour point of the oil (the temperature where the oil .may no longer be
poured). Materials that have this effect include polyalkyl acrylates and
methacrylates, low molecular weight polyethylene waxes, and ethyl-vinyl
acetate (EVA) copolymers. The EVA copolymers are prabably the most com-
monly used of these wax inhibitors. According to Gilby (1983), the main
structural features which influence their performance as wax inhibitors are:
Wax Formation and lnhibition 215
l. Ratio of methylene to vinyl acetate (usually expressed as the vinyl ace-
tate content).
2. Average molecular weight.
3. Hydrocarbon chain branching.
Vinyl Acetate Contento As the vinyl acetate content increases, three impor-
tant trends can be identified:
The degree of crystallinity is reduced
At very high levels of vinyl acetate, chain stiffness is increased
The EVA copolymers used as wax inhibitors are low crystalline compounds
usually having an EVA content between 20 and 40 weight percent.
Molecular Weight. The solubility of EVA copolymers in oil mixtures de-
creases with the molecular weight of the copolymer. On the other hand, the
high molecular weight copolymers are those exhibiting the greatest effect as
wax inhibitors. The EVA copolymers used as wax inhibitors usually have an
average molecular weight between 10,000 and 30,000 (the heaviest oil mix-
ture constituents have molecular weights of the order 1,200).
Hydrocarbon Chain Branching. EVA copolymers having low levels of hy-
drocarbon chain branching are usually more effective as wax inhibitors than
EVA copolymers of similar vinyl acetate content and average molecular
weight, but containing higher levels of hydrocarbon branches.
The EVA copolymers include in their structure polymethylene segments
identical with those of the molecules forming the wax phase. When, prior to
the wax formation, an ordering of the paraffinic molecules takes place in the
liquid phase, the paraffinic branches of the EVA copolymers may enter these
pre-crystalline structures. This is illustrated in Figure 14-2. The vinyl ace-
tate groups of the EVA copolymer are very unlike the paraffinc branches,
and disturb further structuring of the paraffinic molecules. This is the rea-
son why the EVA molecules act as pour point depressants.
Table 14-3 shows the effect on the pour point of adding various amounts
of EVA copolymer to various crude oils. The pour point can be lowered by
approximately 25C, by adding between 25 and 100 ppm EVA copolymer to
the crude oils. The addition of 200 ppm of the inhibitor, on the other hand,
increases the pour point.
It is not absolutely clear frorn the literature whether the wax inhibitors
mentioned depress only the pour point, or whether they also depress the
cloud point.
 Wax Formation and Inhibition 217
216 Properties 01 Oils and Natural Gases

I I COMPARISONS BETWEEN CALCULATED AND EXPERIMENTAL

CH2 CH2 CH2 CH2 CH2 CH2 CH2 RESULTS

~/\/\/V\/\/\/\
I CH2 CH2 CHz CHz CHz CHz CH2 I In Tables 14-4 and 14-5, molar compositions are shown for two oil mix-
I I tures for which the cloud points have been measured at atmospheric pres-
sure. The results are given in Table 14-6. In addition, the cloud points have
J\?\~\/\A/\?\/H2
CH CH2CH2 CH2 CH2 CH2 CHz ~2 CH2 CH2 CH2 CH /\/\/\/\A been calculated, using the model of Won (1985, 1986), and using the model

/ I i I of Hansen et al. (1988). (Both models are described in the Computation of

Wax Formation section). The mixtures were characterized using the proce-
dure of Pedersen et al. as described in Chapter 7. When using the model of
\ ~\/\/\J\/\A/\/2 \ Won, the liquid-phase solubility parameters given in Table 14-2 were extra-
co I CH2 CH2 CH2 CH2 CH2 CH2 CH2 I c=o polated to components heavier than C40, using the following formula:
/ I I /
H
CH3 I I C 3

I I o~ = 7.971 + 0:36132 In MW (14-26)

I I
AMORPHOUS I CRYSTALLINE REGIONS I AMORPHOUS

I REGlON The calculated results are given in Table 14-6. The cloud points are much
REGlON
.1 too high when calculated using Won's procedure. On the other hand, very
I

Table 14-4
Figure 14-2. The wax inhibition mechanism.
Molar Compositlon of Oil Mlxture 1 for Measured Cloud Point Given in
Table 14-6

Table 14-3 Density

Effect of Addltlon of EVA-Polymer on the Pour Point of Varlous Crude Oils Molecular (g/cm3) at
Component Mole % Welght 15'C, 1 atm

Amount of Pour Point C. 1.47 58

on No. EVA (ppm) 'C C5 17.73 72

O 30 e. 6.01 86
1 e7 13.97 93 .743
9
1 100 e. 10.74 107 .753
2 O 8 e9 8.65 120 .776

2 25 - 18 e lO 5.90 134 .790

e" 4.72 147 .795
O - 1
3 c., 3.96 161 .807
3 25 - 26 c., 3.86 174 .823

4 O 21 c., 3.39 188 .831

25
c., 2.92 202 .836
200
4 e,. 2.05 217 .843
EVA characteristics: VAcontent: 28% en 2.25 234 .840
Molecular weight: 13,000 c., 1.62 251 .845
Hydrocarbon branches per 1,000 carbon atoms: 9 1.36 259 .856
C'9

No details have been given about the oil mixtures. 9.41 379 .890
C2()+

Source: Gilby (1983).

 218 Properties o] Oils and Natural Gases
Wax Formation and Inhibition 219
Table 14-5
Molar Composition ot Oil Mixture 2 tor Measured Cloud Point Given in REFERENCES
Table 14-6

Density
Flory, P. J., Principies 01 Polymer Chemistry, Cornel! University Press, New
York, 1953.
Molecular 15C, 1 atm
Component Mole % Weight (g/cc) eilby, e. W., "The Use of Ethylene-Vinyl Acetate Copolymers as Flow Im-
e4 .11 58 provers in Waxy Crude Oil," Chemicals in the Oil Industry, special publ-
es 3.01 72 cation, Proceedings, Manchester, 1983.
e6 5.06 86 Hansen, J. H., Hanningssn, H. P., Pedersen, K. S., and Fredenslund, Aa.,
e7 4.55 93 .741 "A Thermodynamic Model for Predicting Wax Formation in Crude Oils,'
ea 7.21 104 .762 AIChE i.. 1988.
e9 6.02 121 .780
Mandelkern, L., Crystallization 01 Polymers, Mceraw-Hill, New York,
e1Q 3.45 134 .795 1964.
en 4.90 149 .801
c., 3.61 164 .815 Prausnitz, J. M., Lichtenthaler, R. N., and de Azevedo, E. e., Molecular
c., 4.56 176 .828 Thermodynamics 01 Fluid-Phase Equilibria, 2nd ed., Prentice-Hal!,
e" 4.06 188 .838 1986.
c., 4.71 203 .845
c., 2.96 214 .853
Won, K. W., "Continuous Thermodynamics for Solid-Liquid Equilibria:
c., 3.78 232 .848 Wax Formation from Heavy Hydrocarbon Mixtures," Paper 27A pre-
( c., 3.23 248 .851 sented at AIChE Spring National Meeting, Houston, TX, March 26,
c., 3.85 259 .861 1985.
(
e20+ 34.92 446 .922 Won, K. W., "Thermodynamics for Solid-Liquid- Vapor Equilibria: Wax
Phase Formation from Heavy Hydrocarbon Mixtures," Fluid Phase Equi-
libria 30, 1986, pp. 265-279.
Table 14-6
Measured and Calculated Cloud Points at Atmospheric Pressure
(The Molar Compositions 01 the Oil Mixtures are Given in Tables 14-4 and
14-5)

Calculated Cloud Point (OC)

Measured

Mixture
Cloud Point Model of Won
. Model of Hansen
et al.
No. (OC) Twax .1T Twax .1T
1 16 56 49 17 1
2 25 83 58 24 - 1
IIT = Calculated cloud point-measured cloud point.

accurate results are obtained with the procedure of Hansen et al. Observe
that the experimental data of Table 14-6 were included in the data base used
when the values of the interaction parameters (Equations 14-20 through
14-22) were estimated. Due to lack of experimental data it is not possible to
check the predictive capabilities of the model by Hansen et al.
 , "",""
> \~

~
,'1i
Gas Hydrates 221 1;
\~
;

LN(P)
.
H + L2
I L1 + L2

I
Chapter 15 .I
I. L---
Gas Hydrates _-------7-- 1f
F
--- /.
H + G
. /
Gas hydrates are solid, semi-stable compounds which sometimes have
been found plugging natural gas transmission pipelines and other gas han-
/
dling equipment.
Gas hydrates consist of geometric lattices of water molecules containing
cavities occupied by light hydrocarbons, or other light gaseous compounds 273.15 T

(e.g., nitrogen or carbon dioxide). Gas hydrates may be formed where the
mentioned gaseous compounds are in contact with water at temperatures Figure 15-1. Coexistence curve for hydrate (H), gas (G), liquid hydrate forming
below approximately 35C. Gas hydrates may cause plugging of not only compound (L2), ice (1), and liquid aqueous solution (L1) in the case of apure hy-
drate forming compound (- ). Coexistence curve when the hydrate forma-
pipes, but also nozzles, and separation equipment. H is therefore important
tion is inhibited by methanol (_. -' -l. Vapor pressure of the pure hydrate forming
that the gas hydrate formation conditions are taken into consideration when
compound (----------).
designing separation plants and pipelines for use in the oil and gas industry.
Figure 15-1 illustrates the phase behavior of apure hydrate forming gas.
The fully drawn curves are three phase loci. Line AB represents equilibria LN(P)

between hydrate (H), gas in gaseous form (G), and ice (1). Line BC repre- L1 + L2
H + L2
sents equilibria between hydrate (H), gas in gaseous form (G), and liquid
water (L1). Line CD represents equilibria between hydrate (H), gas in liq- -----.~... 'o
uid form (L2), and water in liquid form (L1).
.- \
For a mixed gas, the situation may become slightly more complicated,
due to the possible intersection of the hydrate curve with the phase envelope
.- .- .- G+L1+L2

/
1
I
H + G + L2 /
of the gas mixture. This is illustrated in Figure 15-2. The lines AB, BC, and /
/
DE correspond to the lines AB, BC, and CD in Figure 15-1. At the line CD
in Figure 15-2, hydrate (H), gas in gaseous form (G), gas in liquid form
,./'/
.'-c,/

(L2), and water in liquid form (Ll) are in equilibrium.

...../"". I G+L1

_..-
TYPES OF HYDRATES

Water can form two different types of hydrate lattices, which are referred
to as structure 1 and structure I1, respectively. Both types of lattice contain
numerous small and large cavities. In a stable hydrate, some of the cavities
are occupied by gaseous compounds called guest molecules. Some physical
constants for the two different types of hydrate lattices are given in Table
15-1.
273.15
Figure 15-2. Coexistence curve for hydrate (H), gas (G), liquid hydrate forming
compounds (L2), ice (1), and liquid aqueous solution (L1) in the case with a rnlx-
ture of hydrate forming compounds ( ). Phase envelope for the hydrate
forming compounds (----------). Critical point (o).
 /Ir'

I
/

~:t\
222 Propertes o] Oils and Natural Gases ~"
\;'i
Gas Hydrates 22:l,

Only gases of rather rnodest size and appropriate geornetry can enter the
cavities. Table 15-2 indicates which of the gaseous cornpounds found in typi-
cal petroleum mixtures may enter hydrate cavities. The cavities may contain
one kind of gaseous compounds only, or they rnay contain gas rnolecules of
different chemical species. Not all cavities need to be filled in a stable hy-
drate.
HYDRATE COMPUTATION
Hydrates are formed when the hydrate state is energetically favorable,
compared to a pure water state (water in ice, liquid, or vapor forms). The
transformation from a pure water state to a hydrate state can be regarded as
consisting of two steps:

Table 15-1 pure water (a) - empty hydrate lattice ({3)- filled hydrate lattice (H)
Some Physical Constants of the Two Possible Water Structures in
Hydrates where a, {3, and H are used to identify each of the three states considered.
The {3-state is purely hypothetical, and only considered to facilitate the hy-
Property Structure I Structure 11 drate calculations. Whieh state is energetically favorable depends on which
Number 01 H20 state has the lowest chemical potential. The difference between the chemi-
molecules per unit cel! 46 136 cal potential of water in the hydrate state (H), and in a pure water state (a)
Number 01 smal! can be expressed as:
cavities per unit cel! 2 16

Number 01 large .tH - .t0 = (.tH - .ti3) + (.ti3 - .t0) (15-1)

cavities per unit cell 6 8
o The first term on the right side (.tH - .t(3) can be regarded as the stabiliz-
Cavity diameter/A
Small 7.95 7.82
ing effect on the hydrate lattice, caused by the adsorption of the gas mole-
Large 8.60 9.46 cules. This latter effect depends on the tendency of the gas molecules to en-
ter the cavities in the hydrate lattice. van der Waals and Platteeuw (1959)
have proposed expressing this tendency using a simple gas adsorption model.
The difference between the chernical potential of water in the empty hy-
Table 15-2 drate lattice and the filled one is expressed as follows:
Gases of Interest to Natural Gas Hydrate Formation, and Their
Occurrence in the Different Cavities of the Water Structure

(.tH - .ti3) = RT ~ ViIn (1 - ~ YKi) (15-2)

Structure I Structure 11
Small Large Small Large
Compound cavities cavities cavities cavities i = 1, 2, ... , NCAV
Methane + + + +
K = 1, 2, "', NCOMP
Ethane - + - +
Propane - - - + where Viis the number of cavities of type i, and YKidenotes the probability
Butane - - - +
that a cavity of type i is occupied by a gas molecules of type K. NCAV is the
number of cavities per unit cell in the hydrate lattice (see Table 15-1), and
Isobutane - - - +
NCOMP is the number of gaseous cornpounds present that may en ter a cav-
Carbon ity in the hydrate lattice. The probability YKi is calculated using the
dioxide + + + + Langmuir adsorption theory:
Nitrogen + + + +
Hydrogen
sullide + + + + y Ki = CKJK+ + ~ C) (15-3)
224 Properties o] Oils and Natural Gases Gas Hydrates 225

j = 1,2, ... , NCOMP Table 15-3

A and B Parameters for Calculating Langmuir Constants of Equation 15-5
where fK is the fugacity of component K. CKi s a temperature-dependent
adsorption constant, specific for the cavity of type (i), and for the gas corn- Small cavity Large cavity
A x 103 B A X 103 B
ponent (K). The adsorption constant accounts for the water-gas interactions
Gas Structure (Klatm) (K) (K/atm) (K)
in the hydrate lattice. In the work of van der Waals and Platteeuw and sev-
erallater works (e.g., Parrish and Prausnitz, 1972, Anderson and Prausnitz,
C, 0.7228 3187 23.35 2653
1986), the adsorption constant C is calculated from a cell potential w (r):
0.2207 3453 100.0 1916

C(T) = 4 ll/kT J; exp ( - w(r)/kT)r2 dr (15-4)

C2

C3
.
0.0
0.0

0.0
0.0
0.0

0.0
3.039
240.0
5.455
3861
2967

4638

In this equation k is Boltzmann's constant, T is the absolute temperatura, -C. 0.0 0.0 189.3 3800

and r is the distance from the lattice wall (the water molecule) to the guest C. 0.0 0.0 30.51 3699
(gas) molecule. The lower the cell potential in a given distance, r, from the N2 .. 1.617 2905 6.078 2431
lattice wall, the more likely it is to find a guest molecule at that position. 0.1742 3082 18.00 1728
The cell potential depends on the size of the molecule, and the nature of the CO2 I 0.2474 3410 42.46 2813
gas-water interaction. The latter is expressed in terms of a characteristic en- 11 0.0845 3615 851.0 2025
ergy, and a characteristic distance (the distance Frorn the lattice wall in
H2S I 0.0250 4568 16.34 3737
which the cell potential becomes zero). The distance and energy parameters 11 0.0298 4878 87.2 2633
can be found from experimental hydrate-formation data by minimizing the Source: Munck el al. (1988).
deviation between experimental and calculatated hydrate-formation condi-
tions.
A somewhat simpler approach for calculating C is to use the following where R is the gas constant, and dH and d V are the changes in molar en-
temperature dependence (Parrish and Prausnitz, 1972, and Munck et al., thalpy and molar volume associated with the transition. The following ex-
1988): pression may be obtained for the difference between the chemical potentials
of water in the fl and in the o-states at the temperature, T, and the pressure,
CKi = (AKJT) exp (BKJT) (15-5) P:

For each compound (K) capable of entering a cavity of type (i), AKi and
BKi must be determined from experimental data. The A and B values recom-
mended by Munck et al. are' given in Table 15-3.
.ti3- .ta = d.t(T,P)
RT RT
d.t(To,Po)
RT
r dH dT +
r, RT2
rPo
d V dP
RT
(15-7)
The fugacity of component K, fK, may be determined by a flash calcula-
tion, as explained in Chapter 6, using the SRK equation of state as described
in Chapter 5. where (To, Po) indicates a reference state where d.t is known. In Equation
The second term on the right side of Equation 15-1, (.ti3 - .ta), is equal to 15-7 it is assumed that dH is independent of pressure, and the temperature
the difference between the chemical potentials of water in the empty hy- dependence of the second term on the right side of Equation 15-6 has been
drate lattice (the fl-state), and water in the form of liquid or ice (the a- approximated by the average temperature:
state). An expression for this difference in chemical potentials can be derived
using the following thermodynamic relation: T = (T + To)/2 (15-8)

d(:~)= _ dH dT + dV dP
RT2 RT
(15-6)
If the reference pressure, Po, is chosen equal to zero, Equation 15-7 can be
rewritten to:
226 Properties oj Ols and Natural Gases Gas Hydrates 227

{t!3- p." = ~{t(T,P) = ~{t(To,P = O) _ JT ~H dT + P ~ ~

(15-9)
i.e., at equilibrium at those conditions, no hydrate can exist, and the water
RT RT RT 'r, RT2 . RT will be in the form of either liquid or ice.
Hydrates are not always formed from a condensed phase. The water mol-
~H may be calculated from the difference, ~Cp, in the molar heat capacities ecules of a gas phase may also form hydrate lattices. The hydrate formation
of the (3 and the o-states: process may, in that case, be regarded as consisting of the following succes-
sive steps:
T
~H(T) = J 'r, Cp(T)' dT (15-10) water in gaseous form (G) - water in condensed form (a) -
empty hydrate lattice ({3) - filled hydrate lattice (H)
The constants needed in the calculation of ~{t of the (3 - a transition at a
given temperature and pressure are given in Table 15-4. The difference between the chemical potential of water in a hydrate
phase and in a gas phase may be expressed as follows:

Table 15-4
{tH- p'e = ({tH- {t") - ({t" - {te)
Physical Constants Used for Evaluating Equation 15-9
= ({tH- {t(3) - ({t/3 - p.") - ({t" - {te)

Property Unit Structure I Structure 11

= ({tH- {t(3) - ({t/3 - {te) (15-14)

t>/4o(liq) J/mol 1264 883

t>Ho(liq) J/mol - 4858 - 5201 In this equation ({tH- {t(3) is known from Equation 15-2 and ({t!3 - p.,,)
t>Ho(ice) J/mol 1151 808
from Equation 15-9. The problem that remains is, in arder to calcula te
t>Vo(liq) cm3/mol 4.6 5.0
3.4 ({t!3 - {te) and thereby {tH- {te, finding an expression for ({t'" - {te). The dif-
t>Vo(ice) cm3/mol 3.0
t>Cp(liq) J/mol/K 39.16 39.16 ference between the chemical potential of water in the o-state and in gas-
Sources: Erickson (1983). Munck el al. (1988). eous form may be found from the following expression:

({t" - {te) = RT(ln f~ - In f~) = RT In (f~If~) (15-15)

Using this procedure, the difference in chemical potentials ({tH- p.") be-
tween water in a hydrate state, and in a pure water state, snay now be cal- where f~ is the fugacity of water in the gas phase, and f~ the fugacity of
culated. In Figures 15-1 and 15-2 the unbroken lines indicate corresponding water in the o-phase. The fugacity of water in the gas phase may be ob-
values of temperature and pressure where: tained using an equation of state (e.g., the SRK-equation which s described
in Chapter 5):
{tH- {t" = O (15-11)
f~ = 4>wywP (15-16)
At those conditions, the hydrate state and the pure water state are equally
favorable. To the left of the unbroken lines: P s the total pressure, and 4>w and Yw are the fugacity coefficient and the
mole fraction, respectively, of water in the gas phase. The fugacity of water
p.H - p." < O (15-12) in apure liquid form is calculated from the equation:

and the water will, at equilibrium,
structure I or a structure 11 hydrate
has the lowest chemical potential,
nents as potential guest molecules.
be in a hydrate formo Whether this is a v"'P
f~ = p~t exp RT (15-17)
depends on which of the two structures
in the presence of the actual gas cornpo-
To the right of the unbroken lines: p~t is the saturation pressure of water, and may be found from the equation:

{tH- {t" > O (15-13) In p~t = a + bIT (15-18)

 228 Propertes 01 Oils and Natural Gases Gas Hydrates 229

a and b may be estimated from experimental data. Munck et al. have pro- where ~T = C
posed to use the following values: K = constant that depends on the nature of the inhibitor and on
the heat of formation of the hydrate
Liquid water: a = 17.372; b = - 6141 K s = concentration of the inhibitor in weight percent in the aque-
Ice: a = 14.484; b = - 5351 K ous phase
M = molecular weight of the inhibitor divided by the molecular
where p~t is in atm. In Equation 15-17, Va is the molar volume of water weight of water
(Va = 19.6 cm" for ice, and 18.0 cm" for liquid water). Equation 15-15 may
now be modified to: Values of the constant, K, for some commonly used inhibitors are given in
Table 15-5.

," - ,G - RT In [~wYwP/(l"':'"'p ~;)l (15-19) Table 15-5

Values tor Various Inhibitors ot the Constant, K, Entering the Formula
Proposed by Hammerschmidt (Equation 15-21) tor Calculating Inhibitor
(
By combining Equation 15-9 and Equation 15-19, the following expression Ettect
may be derived for the difference between the chemical potential of water
in the {3-state, and in a gaseous state: Inhibitor K

Methanol 2335

('";/) - +WYW+':' exp ~;:)l Ethanol

Ethylene glycol
2335
2700
Diethylene glycol 4000
+ ~,t(T P0' = O) _ J T ~H dT + P ~ ~
(15-20) Triethylene glycol 5400
RT To RT2 RT
Source: Hammerschmidt (1969).

The fugacity coefficient <Pwof water in the gas phase may be calculated us-
ing the SRK-equation (fugacity expression given in Equation 5-29).
By combining Equations 15-2 and 15-20, it is possible to evaluate whether
hydrate formation in the gas phase is thermodynamically favorable or not,
i.e., whether (,tH- ,tG)has a positive or a negative value.
A more advanced procedure in the calculation of the inhibitor effect is to
use an activity coefficient model to calculate the influence of the inhibitor
on the chemical potential of water in the o-state. When inhibitors are con-
sidered, an additional term must be included in Equation 15-9, which then
becomes:

~,t(T, P) _ In (x ) = ~,t(To, P = O) _ J T ~H dT + P S,! (15-22)

RT w'Yw RT r, RT2 RT
HYDRATE INHIBITORS
where x and I'w are, respectively, the molar concentration, and the activity
The hydrate-formation curve may be shifted towards lower temperatures coefficient of water in the water rich phase. ~,tstill stands for the difference
(the dashed-dotted Iines in Figure 15-1) by addition of a hydrate inhibitor. in chemical potential between the hydrate state and the pure water sta te.
Commonly used hydrate inhibitors are methanol, ethanol, glycols, sodium The activity coefficient and the fugacity of water are interrelated as follows:
chloride, and calcium chloride. The formula of Hammerschmidt (1969)
presents a simple way to calculate the decrease, ~ T, in temperature during I'w = fw(T, P, xI. X2, ... , xw)/(f;:'(T, Pix; (15-23)
the formation of a gas hydrate. It has the following form: where fw = fugacity of water at the actual conditions
Xl, X2,etc. = molar inhibitor concentrations in the water rich phase
Ks f;:' = fugacity of pure liquid water at the same temperature
~T = M(lOO _ s) (15-21)
and pressure
 230 Properties oi Oils and Natural Gases Gas Hydrates 231

The aetivity coefficent of water may be ealeulated using theJJNIQUAC ae-
tivity eoeffieient model (Abrams and Prausnitz, 1975). This approaeh has
been used by Anderson and Prausnitz (1986) and by Munek et al. (1988). In
the latter work the modified UNIQUAC-model of Sander et al. (1986) is
used to model the inhibition eHeet of eleetrolytes (e.g., NaCI and CaCI2).
COMPARISON WITH EXPERIMENTAL
having been inhibited by ethylene glyeol (EG). Also shown in Figures 15-4
and 15-5 are results ealculated using the model of Munek et al. Figures 15-3
through 15-5 show that the hydrate formation temperatures where no inhib-
itor is present are predieted to within O.5C. The aeeuraey is apparently
Table 15-6
Molar Composition of the Seven Component Mixture for Hydrate
OBSERVATIONS Formation Results Shown in Figure 15-3

Figure 15-3 shows experimental results for the deeomposition eonditions Component Mole %

of hydrates formed from a seven-eotnponent gas mixture (Ng and Robinson, N2 5.26
1984) whose molar eomposition is given in Table 15-6. Also shown are calcu- CO2 13.37
lated results using the model of Munek et al. (1988). Figures 15-4 and 15-5
C, 73.90
show experimental hydrate results for a gas eondensate mixture, and an oil
mixture, respeetively. The molar eompositions are given in Table 15-7. Also C2 3.85

shown in Figure 15-4 are experimental observations of methanol's influenee C3 2.02

as an inhibitor on the hydrate formation from the gas eondensate mixture. n-C. 0.80
Figure 15-5 shows the results of the hydrate formation from the oil mixture
nc, 0.80
Source: Ng and Robinson (1984).

50~ 35~ 01 wt
200.0
Experimental data l. MEOH 300.0
point
-- ea 1cu 1ated hydrate
curves
250.0 Experimental data point
--.--.Phase envelope
150.0 __ Calculated hydrate curves

...,
E
200.0

<O
E
...,
<O

~ 100.0

'-
=> .;; '50.0
~
~ 1---'-- -- --- 1-

<, ~V>
e,
'"
_" \ 1-
e,
100.0

50.0

o) 50.0

o.o~-=-=~
230.0 240.0 250. o I
260.o e
270. o
I
280. o
1
290.0
I
300.0
0.0 t===;;;
260.0
---===
270.0
~I
280.0
1
290.0
I
300.0
Temperature, [Kelvinl Temperature, [Kelvinl

Figure 15-3. ealculated hydrate curve for a seven component gas mixture. Figure 15-4. Experimental (Ng et al., 1987) and calculated (Munck et al., 1988)
Methanol (MEOH) is used as inhibitor (Munck et al., 1988). Experimental data results for the hydrate formation conditions for the gas condensate mixture of Ta-
points. (Ng and Robinson, 1984). ble 15-7. Methanol (MEOH) is used as inhibitor.
232 Properties oj Oils and Natural Gases Gas Hydrates 233

300.0
O.U EG
maintained after addition of considerable amounts of methanol to the water
.: Experimental hydrate point, phase. The hydrate formation temperatures from a gas condensate mixture
.: Experimental bubble point in the presence of EG is predicted to within 1C, when the water phase con-
250.0
Calculated hydrate curve tains 20 weight percent EG.
Calculated bubble point curve
200.0
REFERENCES
..,E
~
'" 150.0
Abrams, D. S. and Prausnitz, J. M., "Statistical Thermodynamics of Liquid
'"
s,
::l Mixtures: A New Expression for the Excess Gibbs Energy of Partly or
'"'" Completely Miscible System," AIChE l., 21, 1975, pp. 116-128.
k'" 100.0
Anderson, F..E. and Prausnitz, J. M., "Inhibition of Gas Hydrates by Meth-
anol," AIChE t., 32, 1986, pp. 1329-1333.
50.0 Erickson, D. D., "Development of a Natural Gas Hydrate Prediction Com-
puter Prograrn," M.Sc. Thesis, Colorado School of Mines, 1983.
Temperature, , Hammerschmidt, E. G., "Possble Technical Control of Hydrate Formation
0.0 [Kelvinl in Natural Gas Pipelnes," Brennstoff-Chemie, 50, 1969, pp. 117-123.
270.0 275.0 280.0 285.0 290.0 295.0 300.0

Ng, H.-J. and Robinson, D. B., AIChE National Meeting, Atlanta, Geor-
Figure 15-5. Experimental (Ng et al., 1987) and calculated (Munck et al., 1988)
gia, 1984.
results for the hydrate formation conditions for the oil mixture of Table 15-7. Ethyl-
ene glycol (EG) is used as inhibitor. Ng, H.-J., Chen, C.-J., and Saeterstad, T., "Hydrate Formation and Inhibi-
tion in Gas Condensate and Hydrocarbon Liquid Systems," paper pre-
sented at the AIChE Meeting, Houston, Texas, April 1987.
Table 15-7 Parrish, W. R. and Prausnitz, J. M., "Dssociation Pressures of Gas Hydrates
Compositions of the Gas Condensate Mixture for Hydrate Results Shown
Formed by Gas Mxtures," I&EC Process Des. Dev., 11, 1972, pp. 26-35.
In Figure 15-4 and of the 011 Mixture for Hydrate Results Shown in Figure
Sander, B., Fredenslund, Aa., and Rasmussen, F., "Calculation of Vapor-
15-5
Liquid Equilibria in Mixed Solvent/Salt Systems Using an Extended UN-
Gas IQUAC Equation," Chem. Eng. Sci., 1986, pp. 1171-1183.
Condensate Munck, J., Skjold-jergensen, S., and Rasmussen, F., "Computations of the
Component Mole % 011 Mole % Formation of Gas Hydrates," Chem. Eng. Sci., in press 1988.
N2 0.64 0.16 van der Waals, J. H. and Platteuw, J. C., "Clathrate Solutions," Adv. Chem.
CO2
Phys., 2, 1959, pp. 1-57.
3.11 2.10

C, 73.03 26.19

C2 8.04 8.27

C3 4.28 7.50
te, 0.73 1.83

n-C. 1.50 4.05

ic, 0.54 1.85 I

n-C5 0.60 2.45

C6+ 7.53 45.60 I

MW 32.4 90.2
Source: Ng el al.(1987).
 Simulation oj Miscible Gas lnjection 235

weight components will be in contact with original reservoir oil. With the
increasing content of intermediate molecular weight components in the gas
phase, miscibility may be achieved even though this was not possible with
the original injection gas.
Chapter 16 In case of a condensing gas drive, the original reservoir fluid and the in-
jection gas are not miscible, but miscibility may be achieved at a later time,
Simulation of Miscible Gas Injection at the injection well. The reservoir fluid takes up intermediate molecular
weight components from the injection gas. Since the liquid phase has a
lower mobility than the gas phase, liquid which has taken up intermediate
molecular weight components will be sought and contacted by the injection
gas. Further transference of intermediate molecular weight components
PROCESS DESCRIPTION from the gas to the liquid phase may then occur. After some time, miscibility
may be obtained between the liquid phase and the injection gas.
Reinjecting a gas produced from an oil field into the reservoir is a com- Ternary diagrams like the one shown in Figure 16-1 are often used to il-
mon practice. While previously this was mainly regarded as a convenient lustrate the mixing processes at a given temperature and pressure. The corn-
way of getting rid of the gas, much interest is now being devoted to gas in- position of a petroleum mixture is represented using three groups of compo-
jection as an enhanced oil recovery (EOR) process. Naturally, reinjection of
gas reduces the pressure drop associated with producing oil from a field over
a period of time. However, the injected gas also influences the oillgas phase
equilibrium compositions in the reservoir. The gas may selectively take up
components from the oil phase (vaporizing gas drive), the oil may take up
components from the gas phase (condensing gas drive), or the gas and the oil
may exhibit first contact miscibility. This means that only one phase is
Critical Tie-Line
achieved, no matter in what proportion the oil and the gas are mixed. A
miscible gas displacement of the oil is advantageous, because valuable heavy
components will at the end be contained in a phase with a fairly high mobil- Critical Mixture
ity. The mobility is inversely proportional to the viscosity, and the viscosity is
reduced when the oil takes up gaseous components. The status of- miscible
displacement has been extensively reviewed by Stalkup (1984).
Miscibility may be achieved as a result of:
l. First contact miscibility
2. A vaporizing gas drive
3. A condensing gas drive
First contact miscibility is found when the original reservoir fluid and the
injection gas are miscible. This means that all possible mixtures of the two
""cOmpOSitiOn Ffath
compositions are single-phase fluids at the specified conditions. Between A ancl B

With a vaporizing gas drive, miscibility is achieved at some distance from

the injection well. At the injection well, both a gas and a liquid phase are
C7
Molel e7+ C2-C6
present at equilibrium. Compared to the injection gas, the equilibrium gas
phase contains more intermediate molecular weight components, which
have been extracted from the oil phase. The gas phase has a higher mobility Figure 16-1. Ternary diagram. Mixture A and Mixture B are miscible by first con-
than the oil phase. Hence, gas which has taken up intermediate molecular tact.

234
236 Properties 01 Oils and Natural Gases Simulation 01 Miscible Gas Injecton 237

C, + COz + NZ
nents, C + CO2 + N2, C2-C6, and C7+. Each corner of the diagram
represents 100 % of a given group of components. Any concentration be-
tween O and 100 % is represented at a proportionate distance between the
bottom of the triangle and the opposite corner. Point A in Figure 16-1 corre-
sponds to an oil mixture, and point B to a gas mixture. Also indicated in New Gas Compositions
Figure 16-1 is the location of the two-phase area. The dashed lines are the
tie-lines, connecting gas and liquid phase compositions that are in thermo-
dynamic equilibrium. The tangent to the two-phase area in the critical
point is called the critical tie-line. The locations of the points corresponding
Critica! He-Une
to the reservoir Huid composition and the injection gas, relative to the crit-
cal tie-Iine, determine whether miscibility can be achieved.
One may distinguish between the following cases:
l. The injection gas and the reservoir oil compositions are both located
on the two-phase side of the critical tie-line: Miscibilty cannot be
achieved.
..
2. The composition of the original reservoir Huid is located on the two-
phase side of the critical tie-Iine, and the composition of the injection
gas is on the other side: Condensing gas drive is possible, and miscibl-
ity (multi-contact) can be achieved.
3. The cornpositon of the injection gas is located on the two-phase side of
the critical te-line, and the composition of the original oil is on the 011

other side: Vaporizing gas drive is possible, and miscibility (multi-con- e7+
C z - C6

tact) can be achieved.

4. The composition of the original oil and the injection gas are both 10- Figure 16-2. The principies of a vaporizing gas drive iIIustrated in a ternary dia-
cated on the sngle-phase side of the critical tie-lne: The oil and the gramo
gas exhibit first contact miscibility.

Compositions A and B in Figure 16-1 are both located on the single-phase
side of the critical tie-line and are, therefore, miscible by first contacto Misci-
bility between the oil and the injection gas of Figure 16-2 may be achieved
as a result of a vaporizing process. When the original oil is mixed with an
appropriate amount of injection gas, two phases are formed. The equilb-
rium phases can be represented as two points on the borderline between the
single and two-phase areas in the ternary diagram. The two points are con-
nected by a tie-Ine in Figure 16-2. The new gas composition is brought in
contact with original reservoir oil, and a new gas phase is formed, which
has a composition more like the composition at the critcal point. At some
stage, miscibility may be obtained between the original oil and the gas. The
mechanisms of a condensing drive are illustrated in Figure 16-3. The oil
takes up intermediate molecular weight components from the gas phase,
and the composition of the oil phase approaches that of the mixture at the
critical point. After some time, miscibility may be achieved between the oil
and the injection gas.
Ternary diagrams represent the phase equilibrium picture at a fixed tem-
perature and pressure. Figure 16-4 illustrates qualitatively the development
in the two-phase area as a function of pressure. The size of the two-phase
area decreases with increasing pressure, meaning that the chances of achiev-
ing miscibility are better at a higher pressure. The lowest pressure where, at
a fixed temperature, miscibility may be achieved between a given reservoir
Huid and a given injection gas is called the minimum miscibility pressure
(MMP). The lowest pressure where the reservoir oil and the injection gas are
miscible by first contact is called the first contact minimum miscibility pres-
sure (FCMP).
In reality, ternary diagrams can only give an approximate picture of the
processes that take place when gas is being injected into an oil reservoir. The
reason is that the reservoir Huid, and in many cases, also the injected gas,
are multi-component mixtures. The two-phase area shown in a ternary di a-
gram is based on one specific distribution of pure components within the
groups CI + CO2 + N2, C2-C6, and C7 +. The distribution of components in
238 Properties oi Oils and Natural Gases Smulaton oj Mscible Gas lnieciion 239

Cl + cO2 + N2

Cr-l t c a l

He-Une

Jncreasing Pres sure

Composltions

Oi 1

C7+ C2 - C6 C7+ C2-C6

Figure 16-3. The principies of a condensing gas drive illustrated in a ternary dia- Figure 16-4. Development in the two-phase area of a petroleum mixture as a
gramo function of pressure.

the oil mixture may differ markedly from that in the injection gas. When an
oil is mixed with a gas of a different distribution, the distribution of compo-
nents within one group, e.g., C2-CS, will change. This behavior is further
explained in the section dealing with Cell-to-Cell Simulation.
CALCULATION OF THE MINIMUM MISCIBILITY PRESSURE
(MMP)
With a total of only three real components present in the reservoir Huid
and the injection gas, it is possible to give a precise definition of MMP, using
a ternary diagram. The first contact miscibility pressure (FCMP) is the low-
est pressure where a straight line between the reservoir fluid and the injec-
tion gas nowhere crosses the two-phase area. With a vaporizing gas drive in
the three component case, the critical tie-line, at MMP, will pass through
the original oil composition. Similarly, with a condensing gas drive at MMP,
the critical te-line will pass through the injection gas composition. This def-
inition may, in principle, be extended to the multi-component case, but it is
then no longer possible to represent the results accurately in a ternary dia-
gram, due to the exchange of real components among the groups of compo-
nents representing the corners of the triangle.
Turek et al. (1986) suggested a procedure for calculating the MMP of a
multi-component mixture. The principle is sketched in Figure 16-5 for a va-
'1 porizing mechanism. Liquid of composition L is initially mixed with gas of
composition G. Two phases are formed. The tie-line passes through the
point MI. As a first approximation it is assumed that all the tie-lines are par-
allel. An estimate of the tie-line passing through the originalliquid composi-
tion L can, therefore, easily be established. From what is known, the critical
tie-line at MMP passes through the point L. The line connecting the points L
and M2 must, therefore, at MMP be a good estimate of the location of the
critical tie-line. In reality the tie-lines are not exactly parallel, and an itera-
tive procedure is therefore needed to determine the relative locations of the
,]{
240 Properties of Oils and Natural Gases Simulation of Miscible Gas lniection 241

tie-line through L and the critical te-line. If they are coincident, the se-
lected pressure equals MMP. If the critical te-line is located to the right of
the tie-line through L, the selected pressure is lower than MMP, and vice
versa if the critical tie-line lies on the left side of the tie-Iine through L.
Jensen and Michelsen (1987) have shown that the generalization of the
ternary-dagram approach to the multi-eomponent case may in some cases
give unexpected results. If a te-lne passes through one of the original corn-
positions, the following relation must be fulfilled:
component number N, .e., ZN equals zero. The critical point can be re-
garded as a te-line of zero length. If at MMP the critical tie-line passes
through the composition of the original oil, Equations 16-1 and 16-2 must be
fulfilled for Xand y equal to the mole fraction of component i in the critical
mixture. Since ZNequals zero, one of the following equations must also be
fulfilled:
XN= YN= O (16-3)

Z = ay + (1 - a)x i = 1, 2, 3, ... , NCOMP (16-1) or

where Z = mole fraction of component i in the original reservoir oil a = - lI(KN - 1) (16-4)
for a vaporizing gas drive, and the mole fraction of corn-
ponent i in the injection gas for a condensing gas drive By combining Equations 16-2 and 16-4, the following relation is ob-
y, X= mole fractions of component i in the gas and liquid tained:

phases in equilibrium
a = a parameter that is usually less than O or greater than 1 x = z/(l - (K - l)/(KN - 1)) (16-5)
NCOMP = number of components
The oil contains methane (M), which is more volatile than CO2 (N) and,
Equation 16-1 can be rewritten to: therefore, has a higher equilibrium ratio:

x, = z/(l + a(K - 1)) i = 1, 2, 3, ... , NCOMP (16-2) KM>KN (16-6)

where the equilibrium factor K = y/x. Jensen and Michelsen considered a XMcalculated from Equation 16-5 will be negative. This is obviously wrong,
vaporizing drive where the original reservoir fluid contains C but no CO2 and Equation 16-4 must therefore be rejectedl This means that Equation
The injection gas, on the other hand, consists of pure CO2. CO2 is given 16-3 must be fulfilled at MMP, i.e., at MMP no CO2 is present at the point in
the reservoir where miscibility is achieved.
Based on considerations quite analogous to those given, it can be shown
G that a condensing gas drive at MMP does not contain any components not
present in the injection gas. With an injection gas consisting of a mixture of
Vapor Cl-C5, the miscible fluid at MMP will contain no components heavier than
C5, and this is obviously not the desired resulto
Jensen and Michelsen have shown results for other multi-cornponent mix-
Figure 16-5. A schematic lllustra- Vapor tures in which the ternary diagram analogue gives results which are compa-
tion of a single step in the MMP- rable to what can be obtained with other procedures for determining the
+
calculation procedure of Turek et minimum miscibility pressure.
al. (1986). Liquid

CELL- TO-CELL SIMULATION

Li quid
Gas injection into a petroleum reservoir can be simulated using a cell-to-
cell model. The calculation procedures were described by Metcalfe et al.
L (1972). The program simulates cells of equal volumes in a series, as shown in

~
242 Properties o] Oils and Natural Gases Simulation o] Miscible Gas lniection 243

CELL 1 CELL 2 CELL N Figure 16-6. The temperature and the pressure are the same in each cell,
04ll cu.t w[LL STROrI4
GAS.lNQ GASIANO
and the volume is kept constant. Initially, all the cells contain the same
I I
BATCH 1 fluido A specified amount of gas is added to cell 1. It is assumed that perfect
mixing occurs, and that thermodynamic equilibrium is reached. This means
INJECTlOH GAS
'- that the conditions in the cell can be described by a (P,T) flash calculation
(as discussed in Chapter 6). After the injected gas and the cell fluid mix, the
gas-plus-liquid volumes will be larger than the assumed cell volume. The
C;.t.SIANQOlll Wl:lL ST"[A"
excess volume is transferred to cell 2. Metcalfe et al. have, as illustrated in
~ll
I I Figure 16-7, used three different criteria for defining the excess volume:
BATCH 2
1. Stagnant Oil-AlI the gas formed in cell 1 is transferred to cell 2,
INJE"C,TlOH GAS
'- whereas all the oil remains in cell 1.
2. Moving Excess=-Al] the gas formed in cell I is transferred to cell 2. If
I the volume of the remaining oil phase exceeds that of the original cell,
I
I
the excess volume of oil is also transferred to cell 2.

J
I
I 3. Phase Mobility Criterion- The cell volume remains constant through-
I
I out the calculation, and the excess volume is transferred to cell 2. If
I two phases are present, gas and liquid are moved according to their
I wEll STRQ14
uA!> ...u Vil. I

Ir-
relative phase mobilities. These are calculated from permeability data
BATCH NN
and from the phase viscosities.
INJECTlOH GAS
L-
--- Similarly, the excess volume from ce1l2 is transferred to cell3, etc. When
one batch calculation has been completed, a new injection into cell 1 can
take place, and the cell-to-cell calculation continued. The composition of
Figure 16-6. Principies of cell-to-cell calculation. the injected gas may be changed Irom batch to batch.
Computer simulation of miscible gas injection may thus be carried out by
A STAGNANT Oll B. MOVING EXCESS Oll a series of flash and material balance calculations, accounting for the mass
GASTO
transfer between cells.
NEXT CEll In case of only one phase in a cell, it can be difficult to decide whether
GAS GAS IANO Oll) ~AS
miscibility has been achieved or noto Ternary diagrams are commonly used
h Oll for this purpose, but as mentioned in the Process Description section, ter-
8;~ Oll nary diagrams can only give an approximate picture of the thermodynamic
FINAL CELl ORIGINAL CEll FINAL CEll
state of the reservoir fluido This is exemplified in Figure 16-8 (Pedersen et
ORIOINAl CEll
CONOITION CONomON CONOITION CONomON al., 1986). The ternary diagram shows the two-phase area of a typical North
C. Oll ANO OAS MOVED BY PHASE MOBIUTIES Sea oil. When initially determining the location of the two-phase area, the
GAS ANO Oll 10 NEXT CELl distributions of components within the groups Cl + N2 + CO2, C2-C6, and
I
C7 + were assumed to be constant. Also shown in Figure 16-8 are the te-lines
OAS (ANO Oll) ~AS between the gas and liquid compositions in cells 1 through 5 after injection
Oll of 6.6 moles of a hydrocarbon gas mixture (74 mole % Cl, 13 mole % C3,
GAS ~ and 13 mole % n-Cg). The results indicate a two-phase area outside the two-
~ Oll
phase area, constructed on the basis of the initial component distributions
ORIGINAL CEll
CONOITION within the groups C + N2 + CO2, C2-C6, and C7 +. This example shows
that ternary diagrams should not be used for mult-cornponent mixtures to
Figure 16-7. Different criteria for moving excess fluid from one cell to the next
determine the minimum miscibility pressure. Pedersen et al. (1986) have
one.
l~
244 Propertes oj Oils and Natural Gases
l Simulation oi Miscible Gas Iniection 245

suggested that the definition of the miscibility should be based on the loca- o Crltlcal Polnt

tion of the critical point (Te, Pe) of the reservoir Huid, relative to the reser- Reservoir Condltions
voir conditions (Tres, Pres). Miscibility is achieved when the reservoir Huid is --- Crltical Locus

in the single-phase region, and Tres, Pres is located on the dew-point side of
the phase envelope. This is illustrated in Figure 16-9, which shows the phase
envelopes of the Huids in cell1 through cell5, after injection of an apprecia-
ble amount of a hydrocarbon gas (of the same composition as that given ~
.c

above) into a North Sea oil. In cell 1, Te<Tres, i.e., miscibility has been e
~
achieved according to the definition of Pedersen et al. This definition does
not completely agree with the one given in the beginning of the chapter.
.t~
This is because two phases may be formed in cell1 at a later stage. In most Cell 5
cases the newly formed phase, however, constitutes only a very small frac- 400
Cell 4
tion of the total volume, and the definition of miscibility holds from a prac- 300
tical point of view. This agrees with the observations of Zick et al. (1986).
200
Jensen and Michelsen suggested a somewhat different kind of cell calcula-
100
tion. The starting point, for a vaporizing drive, is a container filled with
injection gas. Original reservoir oil is continuously added to the container
and equilibrated with the gas. The oil that is not dissolved in the gas phase is Temperature (el

Figure 16-9. Phase envelopes of mixtures of a reservoir oil and an injection gas,
C1 + CO2 + N2 present in cells 1-5 of a cell-to-cell simulation immediately after miscibility has
been achieved in cell 1. Source: Pedersen et al. (1986).

Table 16-1
Results for Mlnimum Miscibillty Pressures Obtained from Slim-tube
Experiments vs. Those from Simulation Model of Jensen and Michelsen
(1987)

Reference
to Exp. InJection MIscibility Measured Calculated
Results Gas Mechanlsm MMP (atm) MMP (atm) % Dev.

Natural
Glasa (1986) Vaporizing 474 445 -6
Gas
Hagoort and Natural
Condensing 125 121 -3
Dumore (1985) Gas
Firoozabadi
Natural
and Aziz Vaporizing 396 449 14
Gas
(1986)
Firoozabadi
Natural
and Aziz Vaporizing 409 496 21
Gas
(1986)
Firoozabadi
and Aziz Methane Vaporizing 430 474 10
(1986)
Figure 16-8. Two-phase area of original reservoir fluid and the tie-line in cells 1-
5 after injection of 6.6 mole gas per initial cell mole. % Dev. = 100 x (calculated - experimental)/experimental.
246 Properties oj Ols and Natural Gases Simulation o] Mscible Gas Iniection 247

continuously let out. If, at some later stage, no further oil is formed in the
container, miscibility has been achieved. To simulate a condensing drive the
container is initially filled with original reservoir oil. Injection gas is added,
and the gas which is not dissolved in the oil phase is continuously let out. If,
at some later stage, all the added gas goes into solution in the container
fluid, miscibility has been achieved. As compared with the situation in the
reservoir, the container fluid reflects the situation right at the gas front.
COMPARISON OF CALCVLATED AND EXPERIMENTALLY
DETERMINED MMP's
Slim-tube experiments, as described in Chapter 3, are the most commonly
used experimental technique for carrying out experimental gas displace-
ment studies.
Some results of slim-tube experiments are found in the literature. It is not
obvious what the relation is between an MMP determined by a slirn-tube
experiment, and the results of an MMP calculation. In Table 16-1, experi-
mental MMP results (from a slim-tube experiment) are compared with those
simulated using the container model of Jensen and Michelsen as described in
the section on Cell-to-Cell Simulation. The results typically agree within
10 % .
Pedersen, K. S., Fjellerup, J., Thomassen, P., and Fredenslund, Aa., "Stud-
ies of Gas Injection into Oil Reservoirs by a Cell-to-Cell Simulation
Model," paper presented at the 61st Annual Conference of the Society of
Petroleum Engineers, New Orleans, LA, October 5-8, 1986.
Stalkup, F. L, Miscible Displacement, Monograph Volume 8, H. L. Doherty
Series, Society of Petroleum Engineers, 1984.
Turek, E. A., Luks, K. D., and Baker, L. E., "Calculation of Minimum Mis-
cibility Pressure," paper presented at the 5th Symposium on Enhanced Oil
Recovery of the Society of Petroleum Engineers, Tulsa, OK, April 20-23,
1986.
Zick, A. A., "A Combined CondensinglVaporizing Mechanism in the Dis-
placement of Oil by Enriched Gases," paper presented at the 61st Annual
Conference of the Society of Petroleum Engineers, New Orleans, LA, Oc-
tober 5-8, 1986.

REFERENCES

Firoozabadi, A. and Aziz, K., "Analyss and Correlation of Nitrogen and
Lean-Gas Miscibility Pressure," SPE Reservoir Engineering, Nov. 1986,
pp. 575-582.
Clase, 0., "Blandbar Fortrengning rsrapport 1986," report (in Norwe-
gian) from IKV, Trondheim, Norway, 1986.
Hagoort, J. and Dumore, J. M., "Deterrnination of Minimum Miscibility
Pressure with an Equation of State Program," 3rd European Meeting on
Improved Oil Recovery, Rome, Italy, 1985.
Jensen, F. and Michelsen, M. L., "Calculation of First Contact and Multiple
Contact Minimum Miscibility Pressure," paper presented at the 4th Euro-
pean Symposium on Enhanced Oil Recovery, Hamburg, Germany, Oct.
27-29, 1987.
Metcalfe, R. S., Fussel, D. D., and Shelton, J. L., "A Multicell Equilibrium
Separation Model for the Study of Multiple Contact Miscibility in Rich-
Gas Drives," paper presented at the SPE-AIME 47th Annual Meeting,
San Antonio, Texas, Oct. 8-ll, 1972.
 Index
A C
Acentric factor (Pitzer's)
correlations for, 115-116
tuning of, 15~158
values of, 83
Alani-Kennedy equation, 90-91
calculation results, 147-151
API density calculation, 94-96
calculation results, 147-151
Aromatics, 4-5
B
Binary interaction coefficients
Soave-Redlich- Kwong, 84
tuning of, 155,159-160
Black oils
bubble-point determination, 43
composition, 7,71-72,74-76
C7 + -characterization, 114-128
phase envelopes, 6
PVT measurements, 40-54
PVT simulations, 130-140
thermodynamic properties,
141-150
Ba,44
calculation results, 165
definition, 44
measurement of, 46
Bubble points
calculation results, 141-144
measurement of, 43
249
Composition data, 69-77
Composition determination, 10-37
extension beyond C20 + , 35-37
gas chromatography, 10-17
liquid phase analysis, 12-22
recombination, 32-35
sampling, 39
TBP distillation, 17
Compressibility factor
calculation results, 144-148
experimental results, 131,133,
measurement of, 44,48
from Soave-Redlich-Kwong, 85,88
Constant mass expansion
experiment, 56-57
simulation, 130-131
Constant volume depletion
experiment, 58-60
simulation, 135-140,156-160
Corresponding states theory
principie, 175
thermal conductivity, 188-194
viscosity, 175-181
Critical properties
correlations for Te and Pe, 114-116
tuning of, 155-158
values of Te and Pe, 83
C7 + -Characterization, 114-128
binary interaction coefficients, 122
continuous thermodynamics, 123
density distribution, 121-122
250 Properties 01 Oils and Natural Gases
grouping of pseudocomponents,
124-125
molar distribution, 116-120
molecular weight distribution, 121
TBP fractions, 114-116
TBP residue, 116-120
D Molecular weight
Density
Alani-Kennedyequation,
90-91,147-151
API method, 94-96,147-151
calculation of, 89-96
calculation results, 144-151
experimental determination, 25
internal consistency, 30
Peneloux,89-90,144-150
Standing-Katz,91-94,147-151
volume translation concept, 89-90
Dew points
calculation results, 141-144
experimental determination, 54-56
Differential depletion (liberation)
experiment,43-48
simulation, 131-135
E Gas hydrates, 220-2~
Enthalpy
calculation results, 151-153
experimental data, 151-153
ideal gas, 87
residual, 87-88
from Soave-Redlich-Kwong, 87-88
Entropy
ideal gas, 87
residual, 87-88
Irom Soave-Redlich-Kwong, 87-88
Equations of state, 79-96
cubic, 81-91
Peneloux, 89-90
Soave-Redlich- Kwong (SRK),
82-88
types of, 79-82
lndex 251
F calculation of, 238-246 minimum miscibility pressure,
experimental results, 246 49-54
Flash calculation, 99-112
measurement of, 49-54 real gas deviation factor, 44
multiphase, 104-105
Miscible gas injection separator gas/oil ratio, 40
(P,H), 105-108
condensing drive, 234-238 separator test, 48-49
(P,T), 100-105
simulation of, 238-246 shrinkage factor, 44,46
(P,T) with water, 108-109
vaporizing drive, 234-237 slim tube test, 49-54
results, 144
solution gas/oil ratio, 44,46
simplified (P,T) (kij = O), 109-112
experimental determination, 25-29 swelling test, 49
stability anal ysis, 100-102
internal consistency, 30 volatile fluids, 60-63
two phase, 100-104
as tuning parameter, 166-169 volume factors, 50
Flash separation, 48-49
PVT simulations, 130-140
calculation results, 144
Naphthenes,4-5 constant mass expansion, 130-131
Fugacity coefficients
constant volume depletion,
Soave-Redlich- Kwong, 86,88
P-R 135-140
differential depletion, 131-135
G Paraffins, 4-5
differentialliberation, 131-135
Peneloux density, 89-90
Gas chrornatographic analyses, 10-17 phase envelopes, 141-144
calculation results, 144-150
Gas condensates separator test, 140
Phase envelopes, 6
composition, 7,70,73,76-77
calculation pracedure, 112,113
C7 + -characterization, 114-128 Reservoir fluids
calculation results, 141-144
dew point determination, 54-56 composition, 7
of different reservoir fluids, 6
phase envelopes, 6 constituents, 4-6
Phase compositions r
PVT measurements, 54-60 phase envelopes, 6
from Soave-Redlich- Kwong, 86
PVT simulations, 130-140 PNA-distribution, 8
measurements of (see
thermodynamic properties,
Composition Determination)
141-150
PNA distribution
estimation of, 120-121
s
computation, 223-228
experimental determination of, Separator test
experimental results, 230-233
10-17 experiment, 48-49
inhibitors, 228-238
PVT measurements simulation, 140
types, 220-222.
black oils, 40-54 Shrinkage factor, 44
Gas/oil ratio (GOR)
Bo,44,46 definition, 44
separator (GOR), 33,40
bubble points, 43 measurement of, 46
solution (GOR), 44,46
compressibility factors, 44 Slim tu be test, 49-54 .
Gas reinjection, 49-54,234-246
constant mass expansion, 56-57 simulation of, 241-246
constant volume depletion, 58-60 Soave-Redlich- Kwong (SRK)
H-N
dew points, 54-56 equation, 82-88
Heat capacity differential depletion, 43-48 binary interaction coefficients, 84
ideal gas, 87,126-128 differentialliberation, 43-48 compressibility factors frorn, 85,88
Hydrates (See Gas hydrates) flash separation, 48-49 enthalpies from, 87-88
formation volume factors, 50 entrapies from, 87-88
Minimum miscibility pressure gas condensates, 54-60 fugacity coefficients from, 86,88
(MMP) gas reinjection, 49-54 phase compositions from, 86
 252 Properties o} Oils and Natural Gases
J J

Standing-Katz density, 91-94 molecular weight of plus fraction,

calculation results, 147-151 166-169
Surface tension Tc,Pc and w, 155-158
calculation of, 196-206
Viscosity
definition, 196
calculation of, 172-185
experimental results, 201-206
corresponding states theory,
measurement of, 66-68
175-181
relation to viscosity, 199-200
definition, 172
Swelling test, 49
experimental results, 181-185
measurement of, 63-66
Volatile fluids
T-Z composition, 7,72
C7 + -characterization, 114-128
TBP-distillation, 17 phase envelopes, 6
Thermal conductivity PVT measurement, 60-63 ..
calculation of, 187-194 PVT simulations, 130-140
corresponding states theory, thermodynamic properties,
188-194 141-150
, definition, 187 ~Volume factor, 50
experimental results, 194
Transport properties (See Surface Wax, 208-218
Tension, Thermal conductivity, calculation of wax formation,
and Viscosity) 208c214,217-218
Tuning, 155-170 composition, 208
binary interaction coefficients, experimental results, 217-218
155,159-166 inhibitors, 214-216

~~"""' __ """''''''.Ir.''>~...,,..
__ ~,.kLW--''''''''-'l.,~,---,,,,,,,,",,:,~:,,"'I"~~~-~~

'It..L ..r~T!ITn ' -,';r.:1l";!l r. 'l!

.'p
A',U' "'U, U t-hlJ'-H l,,'"f r: tU" "

'~
r~ I nI
'-,I ~t~.~~~<~
I n T '. (' ;,
'~'_1<' I
~

, w 'NVENT~_~~~.':J. 339c ._._,.... .,...~

~
MATe:?I,'. r
r~:-I~'i~~-~-~'(;~
24-8'f- --'1
~ , __ _~._ ,_. _ "_k" o,, _" ~_,. '.

\ .. ~-~...."..'",:..~,,-.,.", ... .,..,\"' .... -'.:--,.