SPE 68231

The Compositional Characterization of Gas Condensate Fluids - A Review Featuring

the Impact of the Analysis Data Quality on the Accuracy of Equation of State Based
PVT Predictions
T.J. Broad, SPE, Oilphase Division of Schlumberger-Middle East, N. Varotsis, SPE, Consultant-Oilphase Division of
Schlumberger, N. Pasadakis, Technical University of Crete

Copyright 2001, Society of Petroleum Engineers Inc.

This paper reviews the current analytical techniques and
This paper was prepared for presentation at the 2001 SPE Middle East Oil Show held in discusses their advantages and disadvantages. The
Bahrain, 17–20 March 2001.
compositional data obtained on gas condensate reservoir fluids
This paper was selected for presentation by an SPE Program Committee following review of
information contained in an abstract submitted by the author(s). Contents of the paper, as
by gas chromatography (GC), gas chromatography-mass
presented, have not been reviewed by the Society of Petroleum Engineers and are subject to spectrometry (GC-MS) and fractional distillation are
correction by the author(s). The material, as presented, does not necessarily reflect any
position of the Society of Petroleum Engineers, its officers, or members. Papers presented at compared and assessed. The effects that the differences in the
SPE meetings are subject to publication review by Editorial Committees of the Society of derived molecular compositions bear on the EoS based
Petroleum Engineers. Electronic reproduction, distribution, or storage of any part of this paper
for commercial purposes without the written consent of the Society of Petroleum Engineers is predictions of the physical properties of petroleum fluids are
prohibited. Permission to reproduce in print is restricted to an abstract of not more than 300
words; illustrations may not be copied. The abstract must contain conspicuous illustrated using test cases studied in the PVT lab. These
acknowledgment of where and by whom the paper was presented. Write Librarian, SPE, P.O.
Box 833836, Richardson, TX 75083-3836, U.S.A., fax 01-972-952-9435.
effects are particularly pronounced for the estimation of the
retrograde liquid deposit curve, which constitutes a crucial set
of data for efficient reservoir management.
Abstract
The compositional analysis of gas condensate reservoir fluids Introduction
is an integral part of their routine PVT laboratory analysis. Petroleum reservoir fluids are composed of many thousands of
Naturally occurring hydrocarbon mixtures contain thousands different components, hydrocarbons and hetero-compounds
of components, the presence and concentration of which covering a wide range of boiling points and molecular
determine their physical properties and phase behavior during weights. Gas condensates contain fewer hydrocarbons of
production. Accurate compositional data is the prerequisite for heavier molecular weight than oils whilst the heaviest
generating reliable Equation-of-State (EoS) based predictions hydrocarbons in a low shrinkage oil can have a molecular
of the PVT and physical properties of the reservoir fluids over weight in thousands.
a wide range of conditions as well as for quality assurance of There are different ways to classify the constituents of the
the crude. reservoir fluids. Normally the defined components consist of
methane through normal pentane, the main non-hydrocarbon
Unfortunately, no standard methods are available in the constituents (permanent gases) and the heavier than C6
petroleum industry for measuring compositions. Different components expressed as Single Carbon Number Fractions.
analytical techniques are employed for obtaining the Methane through propane exhibit unique molecular structures;
molecular composition of tank oil such as direct injection butane can exist as two isomers and pentane as three isomers.
analysis, simulated distillation by gas chromatography, For hexanes and heavier, the number of isomers rises
separation of the sample by fractional distillation, exponentially. The hydrocarbons heavier than C6 are grouped
identification of specific compounds by mass spectrometry, as normal, iso, cyclo and aromatic (nica). For many years, the
component group distributions by High Pressure Liquid industry has been concerned with the very heavy asphaltenic
Chromatography (HPLC), etc. In addition, several correlations constituents of crude oil. This paper does not address the
are utilised for defining the physical properties of the single specific issues associated with the classification and
carbon number fractions (SCNF) and of the heavy end characterization of these very heavy ends.
fraction. This diversification inevitably leads to differences in Equations of State have been used in the industry for many
the compositions reported by various laboratories which can years. Normally, a PVT study is performed on the reservoir
significantly affect the quality of the predictions provided by fluid to define the fluid properties at reservoir conditions and
the EoS based simulators. at surface separation conditions. The PVT study also includes
the compositional analysis of the reservoir fluid in terms of
2 T. J. BROAD, N. VAROTSIS, N. PASADAKIS SPE 68231

mole percent. Subsequently, the EoS based model is tuned to For converting the weight percent data obtained from gas
match the reported experimental data. Once tuned, the EoS chromatography into molar composition the molecular weights
can be used to predict the fluid properties over a wide range of of all the eluted components must be known.
pressure and temperature conditions and as an integral part of Unfortunately, for the analysis of the equilibrium liquid
compositional reservoir simulators. phase, no generally accepted standard procedure is available5.
The primary input data to the EoS is the composition of the As the liquid contains many thousands of different
reservoir fluid in terms of mole percent and the physical components, a precise analysis of this phase is extremely
properties of the heavier SCNFs together with those of the complex. For simplification, the results are normally reported
plus fraction. The physical properties normally required are in terms of grouped Single Carbon Number Fractions or
molecular weight and density. It naturally follows that if the pseudo-componens including several SCNFs and a plus
input data set is incomplete or inaccurate, the subsequent fraction.
tuning of the EoS will be problematical. The most common way nowadays for analysing liquid
Dependent upon the analytical techniques used to petroleum samples is to use a gas chromatograph equipped
determine the fluid composition, the molecular weights of the with an FID and a non-polar capillary column6,7,8. Despite the
SCNFs and the plus fraction can influence the overall high resolution of the available columns, many of the heavier
composition. In other words, different molar compositions will hydrocarbons still remain unresolved. Great care must be
be obtained if different molecular weights are used. taken when integrating the resultant chromatogram to ensure
The main tuning parameters in the EOS are the critical that such phenomena as baseline drift, chromatographic
temperatures and pressures and the acentric factors of the baseline and unresolved peaks are correctly accounted
SCNFs and the plus fraction. These parameters are derived for9,10,11. Internal standard techniques can be used for the
from various correlations based upon the molecular weight determination of the total chromatogram area12. A
and density. Therefore, it is a prerequisite that these properties disanvantage inherent to the method is that two
are defined during the course of the compositional chromatographic runs are required per sample due to the fact
measurements. that it is practically impossible to find an internal standard not
overlapping with the constituents of the oil.
A Review of Analytical Techniques A problem, which does lead to significant quantitative
It is not possible to measure the complete composition of a errors, is the discrimination that the sample components,
reservoir fluid as described above by gas chromatographic undergo when introduced into a split injection system in the
techniques because of their complex nature. Several analytical chromatograph. Unfortunately, sample discrimination occurs
techniques have been proposed for the compositional in split injection systems when trying to analyse samples with
characterization of the “live” fluid with a single direct a wide boiling point range. An alternative to the split injection
injection1,2,3,4. They require rather complex analytical setup system is to dissolve the liquid samples in a solvent, however
and experimental procedures and their main domain of this can result in some hydrocarbons not being dissolved
application is for synthetic fluids or for very light gas completely in the latter.
condesates. The fact that they cannot ensure the complete An additional problem concerning the gas
quantitative determination of the heavy fraction as it might be chromatographic analysis of liquid hydrocarbon mixtures is
retained in the various elements of the chromatographic the limit in the maximum carbon number SCNF that can be
instrument constitutes one of their disadvantages. fully eluted from the analytical column and detected. The
The analytical protocol for reservoir fluids generally instrumentation that is available today (columns, injection
accepted today, includes the flash of the live fluid into systems and detectors) together with the thermostability of the
equilibrium gas and liquid phases at atmospheric conditions. petroleum constituents does not allow for the accurate
The overall composition of the fluid is then obtained by determination of components higher than C40. Relative
analyzing these two phases and subsequently mathematically response factors have to be assigned for those hydrocarbons
recombining their composition, to obtain the composition of heavier than C20 and lighter than the plus fraction.
the original single-phase fluid. Additionally, backflush techniques can be used to avoid
The analysis of the equilibrium gaseous phase is relatively contamination of the analytical columns by the heavy
straightforward as the majority of the components can be components13.
easily identified and analysed by gas chromatography The physical properties (molecular weight and density) of
(methane through normal pentane and permanent gases). the SCNFs which are used as input data for the EOS cannot be
Multidimensional chromatography using non-polar (capillary determined directly from the results of the chromatographic
or packed) columns and molecular sieves and multiple analysis14,15. The molecular weight and the density of each
detection systems are used. The most commonly used SCNF is usually obtained either from the data proposed by
chromatographic detectors are the Thermal Conductivity one Katz et al16 or from the normal alkanes properties. However,
(TCD) for permanent gases and the Flame Ionisation one each crude oil is unique and the actual physical properties of
(FID) for hydrocarbons. Whereas the TCD is a concentration its SCNFs are dependent upon the nica distribution therein. In
sensitive detector, the FID output is relative to the eluted mass. particular, for two of the gas condensate systems reported in
 THE COMPOSITIONAL CHARACTERIZATION OF GAS CONDENSATE FLUIDS -- A REVIEW FEATURING THE IMPACT
SPE 68231 OF THE ANALYSIS DATA QUALITY ON THE ACCURACY OF EQUATION OF STATE BASED PVT PREDICTIONS 3

this paper, although the molecular weights of their SCNFs are component groups has been used in addition to the common
very similar to the Katz data, their densities however lie compositional information (SCNF) in EoS prediction19. The
midway between the Katz and n-paraffin data. reported EoS prediction results indicate that the additional
The molecular weight and density of the plus fraction can information concerning the chemical nature of the fluids
be determined by material balance from the mass composition increases its accuracy.
of the fluid obtained from chromatography, the assumed
physical properties of the SCNFs and the respective properties Gas Chromatography-Mass Spectrometry
of the total sample. Gross errors can be obtained in the plus The combination of gas chromatography with mass
fraction properties depending upon how close the assumed spectroscopy (GC-MS) offers a powerful tool for the
properties of the light fraction are to the actual values. These compositional characterization of petroleum fluids, as the
errors are magnified as the hydrocarbon number of the plus mass spectrometer is a detector, which allows for the
fraction is increased. identification of the individual hydrocarbons. Once the
hydrocarbon has been identified, its molecular weight and
Distillation density can be assigned from a database and assuming ideal
True Boiling Point (TBP) Distillation has been a standard mixing laws, the molecular weight and density of each SCNF
analytical tool in the industry for many years. Despite can be defined with accuracy. An additional benefit of the
advances in technology, the technique is still rather crude and mass spectroscopy is that a nica breakdown can be attributed
governed by column efficiencies, boil up rates, carry-overs to each SCNF according to its constituent types of
etc. The technique is very time consuming but does address all hydrocarbons.
the shortcomings of the gas chromatographic methods for It has been reported that special care must be taken for the
obtaining the overall composition of the flashed equilibrium proper quantitation from the GC-MS20. The instrument has to
liquid. be checked on a regular basis by carefully tuning its analytical
The liquid is subjected to a TBP Distillation with overhead performance.
cuts being taken to correspond with the boiling point of the Ideally, the equilibrium flashed liquid should be subjected
respective n-alkanes. The distillation continues to the plus to a single cut distillation up to the desired plus fraction which
fraction. Molecular weight and density measurements are is normally dodecanes plus (C12+). The physical properties of
performed on each of the captured cuts together with the plus the plus fraction need to be determined. The distillate should
fraction. be analyzed by GC-MS to provide a detailed composition of
Nevertheless, there are several disadvantages associated the liquid. Subsequent processing of the analytical data leads
with the TBP distillation. Firstly, it is very time consuming to a semi-detailed composition in which the molecular
compared to a gas chromatographic analysis and therefore weights, densities and nica breakdowns of the SCNFs, are
very costly. Secondly, due to low distillation column determined together with the physical properties of the plus
efficiencies, it is not possible to achieve a perfect break in fraction.
between cuts and there is some carry over and overlap This technique, coupled with an accurate analysis of the
between cuts. The quantitative results have to be corrected for equilibrium gas, provides a complete and accurate suite of
the carryover between the different fractions3. Additionally, compositional data, which optimises the tuning process of an
larger sample volumes are required for distillation compared EoS and the resultant predictive capabilities.
to the chromatographic techniques. Notwithstanding these
disadvantages, the TBP distillation does provide most of the Case Studies
data required for effective input into the EoS. In this work, the effect that different compositional techniques
have on subsequent EoS predictions are demonstrated for three
Separation into component groups real gas condensate fluids. The gas condensates, which are
Solvent extraction, open column chromatography and High presented were collected by single-phase sampling techniques,
Pressure Liquid Chromatography (HPLC) are the main whereby the fluids were kept at monophasic conditions from
analytical techniques involved in the identification and the moment the samples were collected at the sampling depth
quantitation of component groups existing in an oil to the time the fluids were introduced into the analytical cells
sample17,18. Using the above techniques, the oil mixture can be in the laboratory. For all fluids, duplicate samples were
separated into saturates, aromatics resins and asphaltenes obtained and detailed quality checks indicated that each
fractions (SARA). captured fluid was representative of the reservoir fluid. Some
The characterization of crude oils in terms of component basic properties of the fluids are presented in Table 1.
group content, provides valuable information since it describes During the analytical program, it was not possible to
the real chemical nature of the reservoir fluid. Among these perform distillation on the flashed liquids for fluids A and B
analytical methods, the HPLC is the method of choice for due to the low volumes recovered from the single-phase
routine characterization of hydrocarbon mixtures, as it requires samplers. For fluid C, distillation was performed on the
short analysis time and relative simple instrumentation and flashed liquid from a separator liquid sample to obtain the
procedures. In a recent work, the separation scheme into physical properties of the plus fraction.
4 T. J. BROAD, N. VAROTSIS, N. PASADAKIS SPE 68231

The flashed liquid obtained from fluids A and B were densities the calculated values of the C20+ fraction are not
analyzed by gas chromatography. Components up to C36 were acceptable.
detected. The calculated compositions are presented as C20+ in Subsequently, the compositions of the STOs were
Table 2. The molecular weight and density of the plus fraction calculated in C12+ form using SCNFs molecular weights and
was determined based on experimentally derermined values on densities from Katz, n-alkanes and from the GC-MS analysis
the total sample and the C19- composition. of the two fluids. The utilized values of the physical properties
The two STOs produced from A and B samples were of the SCNFs are presented in Table 4. The obtained
additionally analyzed by GC-MS according to the compositions are presented in Table 5.
methodology presented by Varotsis et al20. The eluted The flashed liquid produced from Fluid C was subjected to
hydrocarbon peaks in the C11- fraction were classified into fractional distillation and it was separated into a light C11- and
alkanes, iso-alkanes, cyclo-alkanes and aromatics (nica) based a heavy C12+ fraction. The molecular weights and densities of
on their mass spectra. A typical distribution per carbon the total sample and the C12+ residue were determined
number of these groups in the STO of fluid B is presented in experimentally. From the compositional analysis of the
Figure 1. distillate by CG-MS the MW and density of SCNFs contained
Sample C was initially distilled into a light C11- fraction in the C11- were determined and the C12+ properties were
and a heavy end. Molecular weight and density were measured calculated (Table 6). The heavy fraction properties obtained
for both fractions. The C11- part of the sample was analyzed by by the two methods appear to be in good agreement.
GC-MS (Table 3).
EoS simulations were performed using the PVTSIM EOS Predictions
software. The Soave Redlich Kwong EOS was used together The fluid compositions obtained as described above were used
with the Peneloux volume correction factor. For the with PVTSIM to predict some of the pertinent properties of
preliminary simulations no regression was performed at all, so the reservoir fluid, which were subsequently compared with
that only the composition of the fluid effected the simulated the experimentally determined data. For the molecular weight
output data. and density of the SCNFs and the plus fraction, appropriate
values were used depending upon the different compositional
Effect of the physical properties on the molar composition methods applied as presented in Table 4.
The gas chromatographic analyses of fluids A and B were The obtained results from the simulation of the three fluids
extended out through C36 and were grouped back to a C20+ are presented in Tables 7-9. The simulated retrograde liquid
fraction. This method is reasonably accurate for light deposit curves from the Constant Mass study are presented in
condensates as the major part of the fluid is lighter than C20+. Figures 2-4 together with the experimentally determined
However, the method does not provide any accurate data on values.
the properties of the C20+ fraction. These properties have to be The compositional data for fluids A and B presented in
calculated using the measured molecular weight and density of Table 2 indicate that the actual values chosen for the physical
the total sample together with a set of assumed properties for properties of the SCNFs bear significant impact on the
the SCNFs of the analyzed light part of the sample as follows: properties of the plus fraction. As molecular weights and
molecular weight calculation: it is assumed that there are densities influence in their turn the calculated values for
100 weight units of fluid and the number of moles is critical pressure, critical temperature and acentric factor, it is
calculated by 100/MWfluid. The number of moles of the lighter anticipated that inaccurate values of the above will be directly
hydrocarbons through Cn-1- is calculated by ΣWT%i / MWi. reflected in the quality of the EoS predictions particularly
The number of moles of the Cn+ residue is then obtained as the when detailed compositional inputs are utilized.
difference from the total moles of fluid and its molecular To illustrate the effect of the C12+ properties on the
weight can then be calculated from the above equation. simulation results, the phase behavior of fluid C was simulated
density calculation : It is assumed that there are 100 using different values for the physical properties of the plus
weight units of fluid and the volume is then calculated by 100/ fraction, The calculated dew point pressures, Z factor and max
ρfluid. The volume of lighter hydrocarbons through Cn-1 is RLD% are presented in Table 10 and the variation in dew
calculated by Σwt%i/ρi. The volume of Cn+ is then calculated point pressure with the molecular weight of the C12+ residue is
by substracting these two volumes and its density can then be presented in Figure 5.
derived subsequently. .
The molecular compositions of the flashed liquids from
gas condensates A and B are presented in Table 2. Both the Conclusions
Katz as well as the n-alkanes data have been used for the The lab measured molar compositions of flashed petroleum
derivation of the compositions. Whereas the choice of the liquids can be the affected by the experimental technique
assumed physical properties of the SCNF has little impact on adopted for the analysis. More specifically, unrealistic values
the overall fluid compositions for these fluids, they have a for the molecular weight and density of the heavy end can be
very significant effect on the properties of the C20+ fraction. It calculated according to the physical properties values assumed
can be noticed that using the n-alkanes molecular weights and for the SCNFs. This problem is more pronounced when
 THE COMPOSITIONAL CHARACTERIZATION OF GAS CONDENSATE FLUIDS -- A REVIEW FEATURING THE IMPACT
SPE 68231 OF THE ANALYSIS DATA QUALITY ON THE ACCURACY OF EQUATION OF STATE BASED PVT PREDICTIONS 5

extended analysis is provided. Fractional distillation of the 5. Dandekar A.Y. et al.: “Compositional Analysis of North Sea
STO provides direct measurements of these properties. The Oils”, Petroleum Science and Technology, 18, 2000, pp.975-
liquid analysis by a combination of GC-MS and distillation 998.
offers the advantage of estimating the actual MW and SG of 6. Wang Z., Fingas M., Li K.: “Fractionation of a Light Crude Oil
and Idendification and Quantitation of Aliphatic, Aromatic, and
the SCNFs contained in the C- cut and provides for accurate Biomarker Compounds by GC-FID and GC-MS, Part I”, Journal
properties of the C12+ fraction. of Chromatographic Science, 32, 1994, pp. 361-366.
The molecular weights and densities of the heavier SCNFs 7. Beens J., Brinkman U., “The Role of Gas Chromatography in
and, in particular, of the plus fraction were found to bear a the Compositional Analyses in the Petroleum Industry”, Trends
significant effect on the performance of EoS predictions. in Analysical Chemistry, 19, 2000, pp. 260-275
For the purposes of EOS simulations, there would appear 8. Ronningsen H.P, et al. : “Characterization of North Sea
to be a trade off in the carbon number of the plus fraction and Petroleum Fractions: Hydrocarbon Group Types, Density, and
the accuracy with which its physical properties can be Molecular Weight”, Energy & Fuels, 3, 1989, pp. 744-755.
assessed. A reasonable compromise appears to be C12+. 9. Neer L.a., Deo M.D.: “Simulated Distillation of Oils with Wide
Carbon Number Distribution”, Journal of Chromatographic
Further work concering EoS modeling should allow the Science, 33,1995, pp. 133-138.
use of additional compositional information including specific 10. Park J.H., Hussam A., Couasnon P., Carr P.: “The Precision of
component groups split. Such developments are expected to be Area and Height Measurements with Flame Ionization Detectors
highly beneficial to the quality of the PVT predictions. in Temperature-Programmed Capillary Gas Chromatography”,
Microchemical Journal, 35, 1987, pp. 232-239.
Nomenclature 11. Katritzky A.R., Ignatchenko E.S., Barcock R.A., Lobanov V.S.:
Pd = Dew point pressure “Prediction of Gas Chromatographic Retention Times and
T = Reservoir temperature Response Factors Using a General Quantitative Structure-
ρ = Oil density Property Relationship Treatment”, Anal. Chem. 66, 1994, 1799-
1807.
SG = Specific Gravity 12. Yancey J.A., Kosman J.J., Cavalier R.C., Irion J.W.
MW = Fluid Molecular weight “Importance of Quality Control in a Gas Chromatographic
CMS = Constant Mass Study Method to Characterize Crude Oils” Journal of High Resolution
RLD = Retrograde Liquid Deposit Chromatography, 17, 1994, pp.463-468.
STO = Stock Tank Oil 13. Varotsis N., Pasadakis N.: “An Integrated Method for the Gas
Pd = Dew Point Chromatographic Analysis of Natural Gases and Liquid
Z = Gas Deviation Factor Petroleum Mixtures” Chimika Chronika, 25, 1995, 287-300.
14. Durand J.P., Fafet A., Barreau A.: “Direct and Automatic
Subscripts Capillary GC Analysis for Molecular Weight Determination and
Distribution in Crude Oils and Condensates up to C20”, Journal
i = i-th component of High Resolution Chromatography, 12, 1989,pp.230-233.
15. Osjord E.H., Ronningsen H.P., Tau L.: “Distribution of Weight,
Acknowledgements Density, and Molecular Weight in Crude Oil Derived from
The authors would like to thank the management of Oilphase, Computerized Capillary GC Analysis” Journal of High
Division of Schlumberger EPS, for their permission to publish Resolution Chromatography, 8, 1985, pp.683-690.
this paper. 16. Katz D., Firoozabadi A.: “Predicting Phase Behavior of
Condensate/Crude-oil Systems Using Methane Interaction
References Coefficients”, JPT, Nov. 1978, pp.1949-1655.
17. Ali M.A., Nofal W.A.:“Application of High Performance Liquid
Chromatography for Hydrocarbon Group Type Analysis of
1. Khan M.S., Hatamian H.: “Improved method of compositional
Crude Oils”, Fuel Science and Technology Int’l, 12, 1994,
analysis of liquid at high pressure condition in PVT study of gas
pp.21-33.
condensates”, paper SPE 21429 presented at the SPE Middle
18. Akhlaq M.S.: “Rapid Group-Type Analysis of Crude Oils Using
East Oil Show, Bahrain, November 16-19, 1991.
High-Performance Liquid Chromatography and Gas
2. Burke N.E. et al.: “Extended Analysis of Live Reservoir Oils by
Chromatography”, Journal of Chromatography, 644, 1993,
Gas Chromatography”, SPE 21003, presented in the SPE Intl.
pp.253-258.
Symposium on Oilfield and Geothermal Chemistry, Anaheim,
19. Leontaritis K.J. “PARA-Based (Paraffin-Aromatic-Resin-
February 20-22, 1991.
Asphaltene) Reservoir Oil Characterizations”, SPE 37252
3. Kriel W.A. et al.: “Improved Gas Chromatographic Analysis of
presented in the SPE Intl. Symposium on Oilfield Chemistry,
Reservoir Gas and Condensate Samples”, SPE 25190 presented
Houston, Texas, February 16-21, 1997.
in the SPE Intl. Symposium on Oilfield Chemistry, New
20. Varotsis N., Guieze P.: “Reservoir Fluid Characterisation Using
Orleans, March 2-5, 1993.
Gas Chromatography-Mass Spectrometry”, Petroleum Science
4. Danesh A., Todd A.C.,: “A Novel Sampling Method for
& Engineering 15, 1996, pp. 81-89.
Compositional Analysis of High Pressure Fluids”, Fluid Phase
Equilibria, 57, 1990, pp. 161-170.
6 T. J. BROAD, N. VAROTSIS, N. PASADAKIS SPE 68231

Table 1. Basic Properties of the Fluids Table 3. Molecular Compositions of STOs calculated
using different assumed physical properties for the
Property Fluid A Fluid B Fluid C SCNFs. (Fluid C)
Res. Temperature F 279 264 212
(K) (410) (402) (373)
Component Katz n-Alkan GC-MS
Dew Point psia 6084 6116 4985
C1 0.00 0.00 0.00
(MPa) (41.95) (42.17) (34.37)
C2 0.00 0.00 0.00
Max RLD % 8.4 30.8 6.3
C3 0.26 0.26 0.26
Z at Dew Point 1.110 1.107 0.972
i-C4 0.15 0.15 0.15
n-C4 0.63 0.63 0.63
Table 2 – Molecular Compositions of STOs i-C5 0.72 0.72 0.72
calculated using different assumed physical n-C5 1.48 1.48 1.48
C6 5.42 5.27 5.34
properties for the SCNFs. (Fluids A & B)
C7 11.52 11.19 11.91
C8 14.79 13.76 14.49
Fluid A Fluid B C9 14.43 13.60 14.38
Component Katz n-Alkan Katz n-Alkan C10 12.62 11.88 12.43
% mole C11 9.59 9.02 9.25
C1 0.12 0.12 C12+ 28.39 32.05 28.96
C2 0.31 0.31 0.21 0.21 Total 100.00 100.00 100.00
C3 0.43 0.43 0.55 0.55
i C4 0.32 0.32 0.67 0.67
n C4 0.65 0.65 1.32 1.32
i C5 0.76 0.76 1.58 1.58
n C5 1.13 1.13 2.27 2.27
C6 4.10 4.01 6.65 6.48
C7 8.52 8.18 10.20 9.77
C8 13.43 12.61 13.09 12.27
C9 12.25 11.58 10.54 9.94
C10 10.92 10.31 9.38 8.83
C11 8.31 7.83 7.56 7.12
C12 6.06 5.73 4.78 4.52
C13 5.82 5.54 4.59 4.36
C14 4.88 4.68 3.83 3.67
C15 4.05 3.93 3.44 3.34
C16 3.03 2.98 2.67 2.62
C17 2.41 2.38 2.27 2.24
C18 2.35 2.32 2.23 2.20
C19 1.92 1.88 2.03 1.99
C20+ 8.34 12.43 10.02 13.94
Total 100.00 100.00 100.00 100.00

Calculated Physical Properties of the C20+ fraction

Mol Wt 310.5 208.1 344.7 247.8

Density 0.817 1.263 0.870 1.184

Table 4. Molecular weight and density of the SCNFs

SCNF Katz Data n-Alkanes GC-MS GC-MS GC-MS

Data (Fluid A) (Fluid B) (Fluid C)

MW Density MW Density MW Density MW Density MW Density

C7 96.0 0.727 100.2 0.688 95.7 0.716 94.8 0.719 92.85 0.721
C8 107.0 0.749 114.2 0.707 109.5 0.732 110.1 0.729 109.20 0.729
C9 121.0 0.768 128.3 0.722 122.0 0.753 123.1 0.748 121.46 0.753
C10 134.0 0.782 142.3 0.734 137.4 0.755 139.6 0.744 136.00 0.76
C11 147.0 0.793 156.3 0.744 152.0 0.762 153.9 0.755 152.35 0.759
C12+ 224.0 0.827 207.1 0.872 218.4 0.847 234.2 0.871 212.60 0.828
 THE COMPOSITIONAL CHARACTERIZATION OF GAS CONDENSATE FLUIDS -- A REVIEW FEATURING THE IMPACT
SPE 68231 OF THE ANALYSIS DATA QUALITY ON THE ACCURACY OF EQUATION OF STATE BASED PVT PREDICTIONS 7

Table 5. Composition of the STOs of Fluids A and B (C12+ form)

Fluid A Fluid B
Component Katz n-Alkan GC-MS Katz n-Alkan GC-MS.
C1 0.00 0.00 0.00 0.10 0.10 0.10
C2 0.31 0.31 0.31 0.21 0.21 0.21
C3 0.43 0.43 0.43 0.55 0.55 0.55
i-C4 0.33 0.33 0.33 0.67 0.67 0.67
n-C4 0.64 0.64 0.64 1.31 1.31 1.31
i-C5 0.76 0.76 0.76 1.58 1.58 1.58
n-C5 1.13 1.13 1.13 2.27 2.27 2.27
C6 4.12 4.02 4.06 6.65 6.48 6.56
C7 8.55 8.19 8.57 10.20 9.77 10.33
C8 13.48 12.63 13.17 13.09 12.26 12.72
C9 12.29 11.59 12.19 10.54 9.94 10.36
C10 10.96 10.32 10.69 9.38 8.83 9.00
C11 8.33 7.83 8.06 7.56 7.11 7.22
C12+ 38.67 41.82 39.66 35.87 38.89 37.09
Total 100.00 100.00 100.00 100.00 100.00 100.00

Table 6. Fluid C: Physical Properties of the C12+ fraction

using different compositional techniques

Calculated
Lab measured Katz n-Alkanes GC-MS
MW SG MW SG MW SG MW SG
C12+ 240 0.837 217 0.797 192 0.869 213 0.828

Table 7. EoS Predictions for Fluid A Table 10. Effect of C12+ cut properties on the EoS
results for Fluid C
Reservoir EoS predictions C12+ C12+ Dew Point Z @ Pd Max RLD %
Parameter MW Density (MPa) (CMS)
Lab Katz n-Alkan GC-MS
Temp, K 410 410 410 410 190.0 0.829 27.75 0.923 6.13
Dew Point, MPa 41.95 32.58 31.32 40.16 200.0 0.829 29.66 0.943 6.54
Max RLD % 8.4 8.2 7.8 8.7 212.6 0.829 32.74 0.977 7.13
Z at Pd 1.110 1.028 1.016 1.026
220.0 0.829 34.56 0.998 7.46
Table 8. EoS Predictions for Fluid B 230.0 0.829 37.40 1.033 7.92
212.6 0.810 31.06 0.959 7.05
212.6 0.820 31.91 0.968 7.08
Reservoir EoS predictions
212.6 0.829 32.74 0.977 7.13
Parameter
Lab Katz n-Alkan GC-MS 212.6 0.840 33.87 0.990 7.20
Temp, K 402 402 402 402 212.6 0.850 35.05 1.003 7.27
Dew Point, MPa 42.17 38.93 37.73 43.61
Max RLD % 30.8 35.9 33.4 32.3
Z at Pd 1.107 1.119 1.110 1.115

Table 9. EoS Predictions for Fluid C

Reservoir EoS predictions

Parameter
Lab GC-MS
Temp, K 212 212
Dew Point, MPa 34.37 32.74
Max RLD % 6.3 7.1
Z at Pd 0.972 0.977
8 T. J. BROAD, N. VAROTSIS, N. PASADAKIS SPE 68231

40

30

% RLD
Experimental
20 Katz
0 Alkanes
10
C4

GC-MS
C5
C6
C7

Arom-
C8

Cyclo-
CN
C9

Iso-
C10
C11 0 10 20 30 40
Normal Pressure (MPa)

Figure 3. RLD curves for fluid B

Figure 1. Distribution of the nica components in Fluid A derived
by GC-MS

8
10
8 6

% RLD
% RLD

6 Experimental 4 Experimental
4 Katz GC-MS
Alkanes
2
2 GC-MS 0
0
0 10 20 30 40
0 10 20 30 40
Pressure (MPa)
Pressure (MPa)
Figure 4. RLD curves for Fluid C
Figure 2. RLD curves for fluid A