SPE 75686

Correlations To Determine Retrograde Dew Pressure and C7+ Percentage of Gas

Condensate Reservoirs on Basis of Production Test Data of Eastern
Venezuelan Fields
ILsis Marruffo*, PDVSA, Jose Maita, Jesus Him*, Gonzalo Rojas*, Universidad de Oriente. *SPE
This paper was prepared for presentation at the SPE Gas Technology Symposium held in
Calgary, Alberta, Canada, 30 April2 May 2002.
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
presented, have not been reviewed by the Society of Petroleum Engineers and are subject to
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
SPE meetings are subject to publication review by Editorial Committees of the Society of
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
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
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.

Abstract
In this paper are presented new correlations to determine the
dew point pressure and C7+ content of gas condensate
reservoirs, based on production fields parameters usually
available. To develop these correlations it were collected 146
PVT analyses of Venezuela (Anaco) fields: San Joaquin: 42 ;
Santa Rosa: 26 ; El Roble: 16 ; Guario: 4 ; Santa Ana: 3 ; El
Toco: 3 ; Zapatos: 15 ; Mata-R: 15 ; Aguasay: 13 ; La Ceibita:
5 ; Carisito: 4.
The PVT results were validated through consistence tests:
mathematical recombination and mass balance. After these
steps an statistical analysis of the variables was done.
For this, statistical methods were used and they were focused
on multiple no linear regressions.
The correlations's accuracy was the following: dew point
pressure showed 5.74 % of average error with a correlation
coefficient of 0.82 based on 80 PVTs data, the first C7+
content correlation showed 9.59 % of average error with a
correlation coefficient of 0.98 based on 81 PVTs data, the
second C7+ content correlation showed 7.85 % of average
error with a correlation coefficient of 0.99 based on 90
PVTs data.
The new retrograde dew point correlation obtained improves
the Nemeth and Kennedy`s (10) that forecast dew points with
9.69 % of average error when it is applied on 54 Anacos PVT
analyses which had all the compositional information that it
requires; with the new dew point correlation the average error
was 5 % for the same data. These new correlations allow to
improve the characterization and exploitation plans of gas
condensate reservoirs without PVT information.

Introduction
Anaco area has been one of the most important venezuelan
petroleum operational districts and it has a high number of gas
condensate reservoirs. Due to the importance of gas
exploitation in this area, it is fundamental the development of
new technologies that allow their efficient characterization and
exploitation. There are two basic parameters of the gas
condensate reservoirs: the dew pressure and the heavy
components content (C7+), since retrograde condensation
occurs when the pressure reservoir falls below it. They are
rather available for this kind of reservoirs, and in order to get
them it was necessary to develop new correlations to estimate
them based on field parameters easily and usually available.
These parameters are: condensate API gravity, gas/condensate
production ratio and in some cases separator gas gravity and
reservoir temperature.
The selected PVT analyses passed the consistent tests of
mathematical recombination and the mass balance of the
constant volume depletion experiment (CVD), this allowed to
verify that those samples were in thermodynamic equilibrium
at each pressure depletion level. Finally, a nonlinear regression
software was used to fit appropriate models to the data.
Data Description
The collection of PVT data, included the PVT reports of the
Greater Oficina Area as well as the Greater Anaco Area. The
total PVT data included 146 analyses. After the process of
selection and validation, it was obtained a consistent PVT
data base of 114 tests. After a dispersion analyses, the final
data base to develop the new correlations was reduced to
106 PVTs.
Correlations Development
The equations were developed through a commercial
nonlinear regression software to fit appropriate models. The
ones for C7+ content were developed using a matrix to connect
all the independent parameters or variables. To develop the
dew point correlations, it was used the multiple non linear

I. MARRUFFO, J. MAITA, J. HIM, AND G. ROJAS

regressions section of the program , applying statistical tools

like residual analyses and cross-plots.
It is important to note that the more relevant parameters to
achieve this equations were the gas heavy components content
(%C7+) and the gas/condensate production ratio (GCR).
The following are the developed correlations:
First C7+ correlation:
%C7+ = (GCR).....................................................................(1)

GCR

70680

0.8207

%C7+=

....................................................(2)

Second C7+ correlation :

%C7+ = (GCR, SSG).......................................................(3)

%C7+ = 10260* (GCR * SSG )

0.8499

..............................(4)

Dew point correlation:

Pd = (GCR, %C7+, API, Tr)........................................(5)
K 4*Tr K 5 K 6*C + K 7
GCR K 2

Pd = K1 * K 3 * K 8 * API

+
C 7

(6)

Where:
K1 = 346.7764689,K2 = 0.0974139,K3 =-0.294782419.

K4 =-0.047833243,K5 =0.281255219,K6 =0.00068358.

K7 =1.906328237,K8 =8.417626216.
This equation has the same form of Gulf of Mxico correlation
for bubble point (Ref. 5)
TABLE 1-DESCRIPTION OF THE DATA USED FOR THE CORRELATIONS
Number of PVT

PVT Properties

Minimum

Maximum

C 7 + First
Correlation

81

Gas Condensate
Ratio (SCF/STB)

3084

253130

C7 + Second
Correlation

90

Gas Condensate
Ratio (SCF/STB)

2577

253130

Specific gas
separator gravity
(SSG).

0,655

0,904

Gas Condensate
Ratio (SCF/STB)

2000

200.000

0,37
39
160

15,15
61
352

Dew point
Correlation

80

Percentage of C
API
Temperature F

SPE 75686

Results: Discussion and Evaluation

During the development of all kind of correlations it is always
necessary to validate the models using statistical techniques
like residual analyses and cross-plots, to get confidence in
those correlations.
The correlations accuracy was the following: dew point
pressure showed 5.74 % of average error with a correlation
coefficient of 0.82 based on 80 PVTs data; the first C7+
content correlation showed 9.59 % of average error with a
correlation coefficient of 0.98 based on 81 PVTs data; the
second C7+ content correlation showed 7.85 % of average error
with a correlation coefficient of 0.99 based on 90 PVTs data.
Figure 1 shows that the GCR and the C7+ are inversely
proportional. Figure 2 shows the Cross-plot for this
correlation. (%C7+ calculated vs. %C7+ observed).
Figure 3 shows the behavior of the variation %C7+ vs.
GCR*SSG.
Figure 4 shows the cross-plot for this correlation. It is
important clarify that although this correlation is better that the
previous one, any of them may be used because the difference
between their average errors is not so considerable.
Dew Point Correlation
After finishing the elimination of dispersed points through
residuals analysis, 80 PVT analyses were used to obtain the
dew point correlation. The independent variables were: %C7+
(C7+ percentage , obtained from equations 2 and 4 ), GCR (
Gas Condensate Ratio , SCF/STB), API ( API Gravity) and
Tr
( Reservoir Temperature, F). For the data used (80
PVTs) it was obtained a correlation coefficient of 0.82 and an
average error of 5.74%.
Figures 5 and 6 show the final residuals and the cross-plot for
the dew point correlation , respectively. Reviewing the
petroleum literature it was found the Nemeth and Kennedy
( NK ) correlation (10) for gas condensate dew point. This
correlation requires the knowledge of the gas condensate
composition which is very difficult to have without a PVT
analysis. The NK correlation was applied to 54 PVT`s that had
the information required by it yielding an average error of
9.69% which is higher, for this 54 PVTs, than the average
error of 5% obtained with the new correlation presented in this
paper (eq.6). Besides of calculating the dew point pressure
with less error than the NKs , the new correlation only
requires information from the production tests.
Figures 7 and 8 show the cross-plots of the NK and the new
correlations, respectively , and Figure 9 shows a comparison
between both. It is observed that the new correlation shows
less dispersion ( the points fall closer to the 45 line). Also
Table 2 illustrates the results obtained with both correlations
applied to 54 PVT analyses.

CORRELATIONS TO DETERMINE RETROGRADE DEW PRESSURE AND C7 PERCENTAGE OF GAS

CONDENSATE RESERVOIRS ON BASIS OF PRODUCTION TEST DATA OF EASTERN VENEZUELAN FIELDS

SPE 75686

Table 2.Results of Dew Point Correlations Applied to 54 PVTs

CORRELATIONOF NEMETH
ANDKENNEDY

18

NEW CORRELATION

16

14

3*10 6
6
27,9*10
0.891
0.95

399.75

201

9.69

C 7 = 10260*(GCR*SSG)
R

12

-0,8499

= 0,9874; R=0,99

10

%C7

33,9*10 6
30,3*10 6
0.12
0.34

SCR
SCT
R2
R
AVERAGE
OF

RESIDUALS

AVERAGE
ERROR

0
0

20000

40000

60000

80000 100000 120000 140000 160000 180000 200000

14

GCRXSSG

Fig.3-Second Correlation of C7+ ( 90 PVTs )

12

%C 7 + = (GCR/70680) -0,8207

16

R 2 = 0,9604; R = 0,98

10

14

R 2 =0.9727,R=0.98

8
%C7+

12

C7 calculated

10

0
0

50000

100000

150000

200000

250000

300000

GCR(SCF/STB)

Fig 1-First Correlation of C7+ ( 81 PVTs)

0
0

C7+ observed

10

12

14

16

Fig.4-Crossplot for the Second Correlation of C7+

(90 PVTs )

14

12

R 2 =0.9725,R=0.98

%C7+ calculated

10

0
0

10

12

14

%C7 observed

Fig.2-Crossplot for the First Correlation of C7+.

Fig.5-Residuals Analysis for the Dew Point Correlation

I. MARRUFFO, J. MAITA, J. HIM, AND G. ROJAS

6000

6000

5000

5000

R =0.6735,R=0.82
4000

Pd.calculated

Pd.Calculated(lpcm)

SPE 75686

3000
2000
1000

4000

3000

2000

1000

0
0

1000

2000

3000

4000

Pd. Observed (lpcm)

5000

6000

0
0

Fig.6-Crossplot for the Dew Point Correlation(80 PVTs)

2000

3000

Pd.observed

Serie1 Correlation of Nemeth and Kennedy

4000

5000

6000

Serie2 New Correlation of Pd

Fig.9-Comparison of NK and New Correlations

6000

Conclusions
1- The key parameters for the determination of the Dew point
Pressure are: the C7+ content and the gas condensate ratio
(GCR)

5000

R 2 =0.2099,R=0.45
4000

Pd.Calculated

1000

2-The C7+ content and the GCR, of eastern venezuelan gas

condensates are inversely proportional.

3000

2000

1000

0
0

1000

2000

3000

Pd.Observed

4000

5000

6000

Fig.7-Crossplot for the NK Correlation ( 54 PVTs )

4- It were obtained two new correlation to estimate the C7+

content on gas condensate reservoirs, at dew point conditions ,
on basis of production test data., usually available; one of
them does not require the knowledge of the gas condensate
composition only the GCR and the other one requires
additionally the SSG.

6000

5000
2

Pd. Calculated

3-It was obtained a new correlation to determine the Dew

point Pressure of gas condensates reservoirs on basis of
production test data., usually available. It does not require the
knowledge of the gas condensate composition which is very
difficult to get without a PVT analysis.

R =0.731,R=0.85

5-A resume of the new correlations is the following:

Correlation
Variables
PVT Analyses Average
To Calculate
Correlated
No. Error ( %)

4000

3000

C7+(first)
C7+(second)
Dew Point

2000

1000
1000

2000

3000

4000

5000

Pd. Observed

Fig.8-Crossplot for the New Dew Point Correlation

(54 PVTs)

6000

GCR
GCR,SSG
GCR,%C7+,API,Tr

81
90
80

9.595
7.85
5.74

6-The correlation obtained in this work to determine the gas

condensate Dew point Pressure improves in four percentiles
the Nemeth and Kennedys (one which has been used
extensively by the petroleum industry), when it was applied to
Eastern Venezuelan Anacos fields.

SPE 75686

CORRELATIONS TO DETERMINE RETROGRADE DEW PRESSURE AND C7 PERCENTAGE OF GAS

CONDENSATE RESERVOIRS ON BASIS OF PRODUCTION TEST DATA OF EASTERN VENEZUELAN FIELDS

Nomenclature
% C7+ Percentage of heptane plus
Pd
Dew point pressure, psia
GCR
Gas Condensate Ratio, SCF/STB
SSG
Specific gas separator gravity ( air=1)
API
API condensate gravity
Tr
Reservoir temperature,F
R
Correlation Coefficient among variables
R2
Determination Coefficient among variables
% Error
Average error, %
Yi
Observed or experimental value( available data)
y
Estimated or calculated value by a correlation
ei
residual
SCR
Sum of regression squares
SCT
Sum of total squares
SCE
Sum of squares for error ( nor explained by the
regression)
Ki
Regression coefficients of the correlations
Ei
Relative error percentage
Ea
Absolute average error percentage
N
Number of values corresponding to certain data
Acknowledgments
The authors want to thank to PDVSA-Puerto La Cruz
Exploration and Production to allow us to develop this work,
and to Dr. Jaime Almeida (PDVSA-Intevep) for his help in
results discussion.
References
1. Marruffo, I; Cabrera, J; Contreras N., Fonseca ;: Estudio
Integrado Carisito-Aguasay Central. PDVSA. Junio.
(2000).
Inventario de PVTs,
Tomos de
PVTsValidacin de Informes PVT. Campo Carisito,
Aguasay C y Zapatos. CDG-Produccin
Puerto
la Cruz.
2.

Rojas, G.: Anlisis PVT de Gas Condensado.

PDVSA, CIED, Universidad Corporativa. Puerto la Cruz.
Septiembre. (1999).

3.

Almeida, J.: Recomendaciones para la Evaluacin de

Informes PVT y Muestreo de Fluidos en Yacimientos
de Gas Condensado. INTEVEP, S.A. Los Teques.
(1989).

4.

McCain, W.: The Properties of Petroleum Fluids.

Segunda Edicin. PennWell Publishing Company. Tulsa,
Oklahoma.(1989).

5.

Petrosky, G y Farshad, F.: Pressure Volume

Temperature Correlations for Gulf of Mexico Crude
Oils. SPE, Chevron U.S.A. Production. (1998).

6.

Kazmier, L. y Daz, A.: Estadstica aplicada a la

administracin y a la economa. Segunda edicin.
McGraw Hill. Serie Schaum. Madrid. (1989).

7.

Neter, J. y Wasserman, W.: Fundamentos de

estadstica. Nueva edicin. C.E.C.S.A. Compaa
Editoriales Continental, S.A. Mexico. (1973).

8.

Hamke, E. y Reitcsh, G.: Estadsticas para negocios.

Segunda Edicin. McGraw-Hill. Madrid, Espaa. (1997).

9.

Visauta, B.: Anlisis estadstico con SPSS para

Windows. McGraw Hill. Madrid. Espaa. (1997).

10. Nemeth,L.K. y Kennedy, H.T.: A Correlation of

Dewpoint Pressure with Fluid Composition and
Temperature. SPE Paper Number 1477. Junio.
Texas. (1979).
Appendix-Statistical Parameters
The following statistical equations were used for the
correlation development:
1-Determination coefficient(R2) is defined by the ratio of the
sum of regression squares and the sum of the total squares.
SCR
SCE
R2 = -------- = 1 - -------- ..................(A-1)
SCT
SCT
0 R2 1
2-The correlation coefficient ( R) is given by the square root
of the determination coefficient (R2)1/2 . This criteria allows to
measure the level of association that exists between two
variables
:
Dependent
independent
or
independent independent.
R=

exp lained var iation

SCR
......................(A-2)
=
total var iation
SCT

3-The analysis of residuals is based on the difference between

the observed and adjusted values of a population or data.
ei= Yi Yi ...............................................................(A-3)
On the whole , the analysis of residuals means to make an
analysis of residual graphs
4-The relative error percentage is the difference between
observed and calculated or estimated values divided by the
observed value and multiply by 100. That is

Yi Yi
* 100 ......................................(A-4)
Ei =
Yi

I. MARRUFFO, J. MAITA, J. HIM, AND G. ROJAS

5-The absolute average error percentage is given by the sum of

all relative error percentages divided by the number of values
used in the calculation of the average error.

Ea =

1 n
Ei ................................................................(A-5)
n i =1

6-The cross-plots are the graphic representation of the

calculated values by the correlation vs. the experimental or
observed values. When both values are equals ( Zero error),
all the points fall on a 45 line and R2=1 . Points far from the
45 line indicate low accuracy of the correlation.
SI Metric Conversion Factors
1 pound mole 379.63 standard cubic feet
1 barrel 5.615 cubic feet
Water density at 60F and 14.7 Asia = 62.4 pound/cubic feet.

SPE 75686