Gas Condensate Dew Point

Copyright 2001, Society of Petroleum Engineers Inc.

This paper was prepared for presentation at the 2001 SPE Middle East Oil Show held inBahrain, 17–20 March 2001.

This paper was selected for presentation by an SPE Program Committee following review ofinformation contained in an abstract submitted by the author(s). Contents of the paper, aspresented, have not been reviewed by the Society of Petroleum Engineers and are subject tocorrection by the author(s). The material, as presented, does not necessarily reflect anyposition of the Society of Petroleum Engineers, its officers, or members. Papers presented atSPE meetings are subject to publication review by Editorial Committees of the Society ofPetroleum Engineers. Electronic reproduction, distribution, or storage of any part of this paperfor commercial purposes without the written consent of the Society of Petroleum Engineers isprohibited. Permission to reproduce in print is restricted to an abstract of not more than 300words; illustrations may not be copied. The abstract must contain conspicuousacknowledgment 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.

AbstractThe paper presents a new empirical correlation to predict thedewpoint pressure of gas-condensate fluids from readilyavailable field data. The new correlation relates the dewpointpressure of a gas-condensate fluid directly to its reservoirtemperature, pseudoreduced pressure and temperature, primaryseparator gas-oil ratio, the primary separator pressure andtemperature, and relative densities of separator gas andheptanes-plus fraction. The correlation was developed basedon field and laboratory PVT analysis data of several gas-condensate fluid samples representing different gas reservoirsin the Middle East. Additional data sets, not included in thedevelopment of this correlation, were used to validate the newmodel’s accuracy. Based on the error statistical analysisresults, the new model outperforms the existing correlations.

IntroductionThe dewpoint of a gas-condensate fluid occurs when a gasmixture containing heavy hydrocarbons is depressured until aliquid is formed. The dewpoint is defined by a substantialamount of gas phase that exists in equilibrium with aninfinitesimal amount of liquid phase. The determination ofgas-condensate dewpoint pressure is essential for fluidcharacterization, gas reservoir performance calculations, andfor the design of production systems.

Traditionally, the dewpoint pressure of a gas-condensatefluid is experimentally determined in the laboratory in aprocess called constant mass expansion (CME) test using avisual window-type PVT cell. The laboratory measurement ofthe dewpoint pressure provides the most accurate and reliabledetermination. However, due to the following economical andtechnical reasons, quite often this information cannot beobtained from laboratory measurements: (1) the laboratory

analysis can be expensive and cumbersome, (2) inability toobtain a representative sample, (3) sample volume isinsufficient to obtain complete analysis, and (4) laboratoryanalyses are in error. In this case, the dewpoint pressure valuescan be predicted from empirically derived correlations.

The dewpoint pressure correlations, proposed in theliterature, are considered very limited. In addition, thesecorrelations were developed based on gas-condensate fluidsamples obtained from certain reservoirs of specific regions inthe world. Due to varying compositions of gas-condensatefluids from reservoirs of different regions, different empiricalcorrelations may not provide good predictions of dewpointpressures when they are applied to gas-condensate fluidsbehaving differently from the fluids based on which they weredeveloped. Most of these empirical correlations are stronglyrelating the dewpoint pressure to the gas-condensate fluidcomposition. Therefore, there is a great interest to evaluate theaccuracy of these empirical correlations in predicting thedewpoint pressures of the Middle East gas-condensate fluids.

This paper presents a new empirical correlation forpredicting dewpoint pressure of gas-condensate systemsexclusively for the Middle East gas-condensate fluids usingmultiple linear/nonlinear regression procedures. In addition,the accuracy of some empirically derived dewpoint pressurecorrelations was evaluated to determine their applicability forthe gas-condensate fluids prevailing in the Middle East.

Review of LiteratureIn 1947, Sage and Olds 1 studied experimentally the behaviorof five paired samples of oil and gas obtained from wells inSan Joacuin fields in California. Their investigations resultedin developing a rough correlation relating the retrogradedewpoint pressure to the gas-oil ratio, temperature and stock-tank API oil gravity. The results of this correlation werepresented in tabulated and graphical forms. This correlation isapplicable only for gas-oil ratio of 15,000-40,000 scf/STB, fortemperature of 100-220oF, and for API oil gravity of 52o-64o.

In 1952, Organick and Golding 2 presented a correlation topredict saturation pressures, which could be a dewpoint or abubble point pressure, for gas-condensate and volatile oilreservoir fluids. Saturation pressure is related directly to thechemical composition of the mixtures with the aid of two-generalized composition characteristics: (1) the molal averageboiling point ( B ) in oR, and (2) the modified average

SPE 68230

A New Correlation for Gas-condensate Dewpoint Pressure PredictionA.A. Humoud, SPE, Saudi Aramco, and M.A. Al-Marhoun, SPE, King Fahd U. of Petroleum and Minerals

2 A.A. HUMOUD, M.A. AL-MARHOUN SPE 68230

equivalent molecular weight (Wm). These parameters can becalculated from the composition of the gas mixture. Thecorrelation was given in the form of 14 working charts, and oneach chart the saturation pressure is plotted againsttemperature. Each chart is for a specific value of Wm and givesa set of curves representing different values of B .

In 1967, Nemeth and Kennedy 3 developed a correlation inthe form of an equation, which relates the dewpoint pressureof a gas-condensate fluid to its chemical composition,temperature and characteristics of C7+. The final form of theequation contains eleven constants; See the Appendix. Thedewpoint pressure and temperature ranges varied from 1,270-10,790 psi, and 40-320oF respectively. The average absoluteerror for the 579 experimental data points used to develop thiscorrelation was found to be 7.4%.

In 1996, Potsch and Braeuer 4 presented a graphicalmethod for determining the dewpoint pressure as a backup forthe laboratory visual reading of dewpoint pressure during aCME test. The key idea of this method is to plot the number ofmoles, calculated as a function of single-phase compressibilityfactor (Z-factor), versus pressure. Above dewpoint pressure,the plot yields a straight line, and below dewpoint pressure theplot shows a curve. The point of intersection marks thedewpoint pressure.

Data AcquisitionSeventy-four (74) data sets, representing different gas-condensate fluids of the Middle East, were acquired and madeavailable for this research. Each data set included fieldproduction data, fluid compositional data, and the results ofCME test. The field production data included the reservoirpressure and temperature, the operating pressure andtemperature of the primary separator, the primary separatorgas-oil ratio, the separator gas specific gravity and theheptanes-plus specific gravity. The results of compositionalanalysis were utilized to estimate the pseudocritical propertiesof the gas-condensate fluids. All the dewpoint pressure valuesused in this study were experimentally determined from theCME tests performed on the gas-condensate fluids.

Table 1 lists the ranges of main parameters of the MiddleEast gas-condensate fluids. Wide ranges of dewpoint pressure,temperature and gas-oil ratio were covered. Lean and richgases with high concentration of acid gas were also covered inthis study. Development of CorrelationMultiple least-square linear/nonlinear regression was utilizedto develop this new dewpoint pressure correlation. Statisticalregression programs were developed to build the models,investigate the behavior of many regression models andevaluate each combination of the dependent variable with theindependent variables.

The first step in developing the correlation was to selectthe parameters that are anticipated to influence the dewpointpressure behavior. Two types of independent variables wereused to develop the new model. The first type of independent

variables was based on the pressure and temperatureconditions, while the second type of independent variables wasrelated to the gas-condensate fluid composition.

Several models were mathematically formulated withdifferent sets of parameters, expressed in different forms oflinear, logarithmic and power relationships. The relationshipsof these parameters, on the individual and combined basis, tothe dewpoint pressure were investigated. Based on ofcoefficient of determination (r2) and the t-test results, the mostimportant parameters were selected and the least importantones were excluded from the correlation. The best model thatfits the seventy-four (74) experimental data points was foundto be;

)γ,γ,T,P,T,P,R,(TP C7gSPprprSPSPSPRd += ƒ …………...(1)

where

+C7γ = heptanes-plus specific gravity (water = 1.0)

gSPγ = primary separator gas specific gravity (air = 1.0)

Ppr = pseudoreduced pressurePSP = primary separator pressure (psia)RSP = primary separator gas-oil ratio (scf/SP bbl)Tpr = pseudoreduced temperatureTSP = primary separator temperature (oR)TR = reservoir temperature (oR)

The following equation has been found to be the best formthat minimizes the deviation from measured data:

)TPln()Rln()Tln()ln(P SPSP3m2R10d ⋅β+β+β+β=

+γ

β+β+β+7C

6

pr

5

pr

4

PT………………………....(2)

where

β0 = 43.777183β1 = -3.594131β2 = -0.247436β3 = -0.053527β4 = -4.291404β5 = -3.698703β6 = -4.590091

The mass gas-oil ratio (Rm) is defined as:

+γγ⋅

=7C

gSPSPm

RR ………………………………………..(3)

The pseudoreduced pressure and temperature are defined as:

SPE 68230 A NEW CORRELATION FOR GAS-CONDENSATE DEWPOINT PRESSURE PREDICTION 3

pc

Rpr P

PP = ………………………………………………..(4)

pc

Rpr T

TT = ……………………………………………….(5)

where

PR = reservoir pressure (psia)

For known gas mixture composition, the pseudocriticalpressure (Ppc) in psia and the pseudocritical temperature (Tpc)in oR are estimated using Kay’s mixing rules defined as: 5

∑=

⋅=n

1iciipc PyP ……………………………………...…..(6)

∑=

⋅=n

1iciipc TyT ………………...………………………..(7)

where

Pci = critical pressure of component Ci

Tci = critical temperature of component Ci

yi = mole fraction component i in the gas mixture

Correlation Error Analysis. The new correlation, given inEq. 2, was developed with a correlation coefficient (r) of0.9479, which indicates that about 95% of the data variation inthe dewpoint pressure (dependent variable) can be explainedby the model. The average absolute relative error was 4.33%and the error standard deviation was 3.34%.

The error distribution of this new correlation is presentedin Fig. 1 as average absolute relative error versus datafrequency. It indicates that approximately 65% of thepredicted dewpoint pressures fall within 5% average absoluteerror, and about 95% of the data points are within 10% error.The model’s accuracy for all the data points used to developthis correlation is 15%.

Sensitivity of New Model. The influence of the individualindependent variables on the dewpoint pressure was tested. Inthis sensitivity test, the variation in the dewpoint pressure isobserved by varying an independent variable over the practicalrange while holding the other variables constant at minimum,average and maximum values.

Figure 2 shows that dewpoint pressure is very sensitive tothe reservoir temperature when other variables are held atminimum values. It is also observed that the sensitivity of themodel is reduced as the values of other variables increase. Theinfluence of the reservoir temperature becomes much lesssignificant when other variables are at their maximum values.

Figure 3 illustrates that the model is more sensitive to lowvalues of gas-oil ratio than to high values. It is also noticedthat the influence of the gas-oil ratio on the model isdecreasing with decreasing the values of other variables.

New Correlation for Pseudocritical Properties. Thepseudocritical pressure and temperature of a gas-condensatefluid can be estimated using some empirical relationshipsbased on the reservoir gas gravity, such as Standing’scorrelation presented in the Appendix, if the composition isnot available.

For accuracy purposes, a new correlation has beendeveloped in this research for estimation of pseudocriticalproperties of the Middle East gas-condensate fluids. Usinglinear least-squares regression, the new relationships asfunction of reservoir gas specific gravity ( gRγ ) are:

gRpc 3.555.694P γ−= ……………………………....…..(8)

gRpc 7.2135.208T γ+= …....…………………………....(9)

Using the correlation given in equations 8 and 9, theaverage absolute errors were found to be 1.8% and 1.0% inpredicting the pseudocritical temperature and pressurerespectively compared to the Kay’s mixing method (equations6 and 7). Based on the estimated pseudocritical propertiesusing equations 8 and 9, the new model in Eq. 2 predicted thedewpoint pressures with an average absolute error of 6.4%, ascompared to 13% using Standing’s correlation.

Comparison of CorrelationsThe data sets used to develop the new correlation were utilizedto evaluate the accuracy of two existing dewpoint pressurecorrelations: (1) the Organick and Golding correlation, and (2)the Nemeth and Kennedy correlation. Both statistical andgraphical means were used in this comparative evaluation.Statistical Error Analysis. Table 2 lists three statistical errorparameters that are used to evaluate the two existingcorrelations in comparison with the newly developedcorrelation. The parameters include the average absoluterelative error (Ea), the maximum relative error (Emax), and thestandard deviation error (s).

Table 2 shows that using the Organick and Goldingcorrelation to predict the dewpoint pressures of the MiddleEast gas-condensate fluids resulted in an average absoluteerror of 32.6%, and a maximum relative error of 53.4%. Thisindicates that the accuracy of this correlation is unsatisfactory.

Table 2 shows also that the accuracy of the Nemeth andKennedy correlation is better than the Organick and Goldingcorrelation. The average absolute error in the dewpointpressure predictions was 11.64%, and the maximum relativeerror was 37.4%.

The newly developed correlation predicted the dewpointpressures with an average absolute error of 4.33%, and amaximum relative error of 15.1%. Therefore, the newcorrelation outperforms the existing correlations.


Graphical Error Analysis. Two graphical analysistechniques were considered in this study for comparativeevaluation of the existing correlations with the newcorrelation. These techniques included the crossplot and theparameter grouping analysis. Crossplots. The crossplots of estimated versus experimentalvalues of the dewpoint pressures are shown in Fig. 4 through6. For the purpose of accuracy analysis, both zero-error line orthe 45o line and the + 10% error variance are drawn on theseplots.

The crossplot of Organick and Golding, presented in Fig.4, shows that the majority of data points are widely scatteredand highly deviated from the 45o line. The crossplot alsoreveals that this correlation always underestimates thedewpoint pressures compared to the experimental values.

The crossplot of Nemeth and Kennedy correlation,presented in Fig. 5, shows that the correlation dewpointpressure predictions were within + 10% error variance fordewpoint pressures less than 5500 psia. However, thecorrelation had the tendency to underestimate the dewpointpressures at pressures higher than 5500 psia with increasingthe deviation as the dewpoint pressure increases.

The crossplot of the newly developed correlation,presented in Fig. 6, shows clearly that the majority of the datapoints fall within +10% error variance. It is also observedfrom this crossplot that the data points are well scatteredclosely around the 45o line. Overall, The newly developedcorrelation showed a better prediction accuracy and a bettererror scatter in the data points than the other two existingcorrelations.

Parameter Grouping Analysis. The degree of errorrandomness or error distribution with the change in anyparameter was tested by plotting the average absolute errorversus groups of a variable.

Figure 7, is a plot of average absolute error grouped by thedewpoint pressure. The plot shows that both existingcorrelations are influenced by the dewpoint pressure withincreasing the average absolute error as the dewpoint pressureincreases. However, the new correlation is showing a randomaverage absolute error over different ranges of dewpointpressure.

Figure 8, is a plot of average absolute error grouped by thereservoir temperature. The Organick and Golding correlationshows a random error distribution with the change in reservoirtemperature. On the other hand, the Nemeth and Kennedycorrelation is strongly influenced by the change in reservoirtemperature with decreasing the average absolute error as thereservoir temperature increases. The correlation showed aconstant Ea at temperatures greater than 250oF. The plot showsalso that the error in the new correlation is not influenced bythe change in reservoir temperature. The least error occurred attemperature range of 200-250oF.

Validation of New CorrelationIn order to examine the applicability and reliability of thenewly developed correlation, it was validated using 20 data

sets for the Middle East gas-condensate fluids that were notused in the development of the new correlation. The newcorrelation was also compared against the two existingcorrelations.

Table 3 summarizes the results of error analyses of thethree correlations. The table shows that the Organick andGolding correlation has the highest average absolute andmaximum relative errors. The Nemeth and Kennedycorrelation shows a better accuracy than the Organick andGolding correlation.

The accuracy of the new correlation was tested using twodifferent methods for estimation of pseudocritical pressure andtemperature, based on the available fluid composition (Eqs. 6and 7) and the gas specific gravity correlation (Eqs. 8 and 9)for unknown fluid composition. The predictions of the newcorrelation for the dewpoint pressure were found within lessthan 4% average absolute error, and with a maximum relativeerror of less than 10% as shown in Table 3.

Conclusions1. A new empirical correlation to predict the gas-

condensate dewpoint pressure has been developed.2. The developmental approach of this new correlation

is based on some readily available gas-condensateproperties and parameters. Most of these parameterswere not considered in the existing correlations.

3. Comparative evaluation of existing correlations wasmade using statistical and graphical error analyses. Itshowed that the new correlation outperforms theexisting correlations.

4. The empirical correlation developed in this study wasvalidated by data, which were not used in thedevelopment of this correlation, and the modelprovided a better accuracy than the existingcorrelations.

5. The new correlation is considered more relevant andapplicable to the Middle East gas-condensateproperties and conditions than the existingcorrelations since the new model has been developedbased on a large number of data sets for the MiddleEast gas-condensate fluid samples.

6. New correlations to estimate pseudocritical propertiesas a function of gas relative density were developedbased on Middle East gas-condensate data. Higheraccuracies were obtained when these correlations areused instead of Standing’s ones.

Nomenclature

Ea = average absolute percent relative errorEmax = maximum absolute percent relative errorγapi = stock tank oil gravity, oAPI

+γ 7C = specific gravity of heptanes-plus fraction

(water = 1.0)

gsγ = average specific gravity of surface separator


gas (air = 1.0)γgR = reservoir gas specific gravity (air = 1.0)

γgSP = specific gravity of gas from primary

separator (air = 1.0)

MC7+ = the molecular weight of heptanes plusPd = dewpoint pressure, psiaPpc = pseudocritical pressure, psiaPci = critical pressure of component ciPpr = pseudoreduced pressure of the gas mixturePSP = primary separator pressure (psig)Rm = mass gas-oil ratio, defined by Eq. 3RSP = producing gas-oil ratio from primary

separator (scf/STB)r2 = coefficient of determinationr = correlation coefficients = standard deviationT = temperature, oR

Tci = critical temperature of component iTpc = pseudocritical temperature, oRTpr = pseudoreduced temperature of the gas

mixtureTSP = primary separator temperature (oR)TR = reservoir temperature (oR)yi = mole fraction of component i in the gas

mixtureyCi = mole fraction of component Ci in gas mixture

Subscript

SP primary separatorR reservoirg gasC1,C2,… methane, ethane,…C7+ property of heptanes plus fractionm mass

d dewpoint pc pseudocritical pr pseudoreduced

References1. Sage, B.H. and Olds, R.H.: “Volumetric Behavior of

Oil and Gas from Several San Joaquin Valley Fields,”Trans., AIME (1947) Vol. 170, 156-173.

2. Organick, E.I. and Golding, B.H.: “Prediction ofSaturation Pressures for Condensate-Gas and Volatile-Oil Mixtures,” Trans., AIME (1952) Vol. 195, 135-148.

3. Nemeth, L.K. and Kennedy, H.T.: “A Correlation ofDewpoint Pressure With Fluid Composition andTemperature,” paper SPE 1477 presented at SPE 41stAnnual Fall Meeting held in Dallas, Tex., 1966.

4. Potsch, K.T. and Braeuer, L., “A Novel GraphicalMethod for Determining Dewpoint Pressures of GasCondensates,” Paper SPE 36919, presented at the 1996SPE European Conference held in Italy, October 22-24,1996.

5. Sutton, R.P., “Compressibility Factors of High-Molecular-Weight Reservoir Gases,” Paper SPE 14265,presented at the 1985 SPE Annual TechnicalConference and Exhibition, Las Vegas, Sept. 22-25,1985.

6. Ahmed, T.: Hydrocarbon Phase Behavior, Vol. 7, GulfPublishing Company: Houston, 1989.

SI Metric Conversion Factors141.5/(131.5+ γapi) = γo (unit-less)bbl x 1.589 873 E + 01 = m3

ft3 x 2.831 685 E – 02 = m3

(oF + 40)/1.8 – 40 =

oC

psi x 6.894 757 E + 00 = kPa

Appendix – CorrelationsNemeth and Kenndy dewpoint pressure correlation: 3

Where

A1 = -2.0623054 A2 = 6.6259728 A3 = -4.4670559 x 10-3 A4 = 1.0448346 x 10-4

A5 = 3.2673714 x 10-2 A6 = -3.6453277 x 10-3

A7 = 7.4299951 x 10-5 A8 = -1.1381195 x 10-1

A9 = 6.2476497 x 10-4 A10 = -1.0716866 x 10-6

A11 = 1.0746622 x 10

Standing pseudocritical properties correlation: 6

2gRgRpc 1.117.51706P γ−γ−=

2gRgRpc 5.71330187T γ−γ+=

( )

113

7C

7C10

2

7C

7C9

7C

7C837C7C7

27C7C6

7C7C541C

1C37C2

N1C4C3C6CSH2CO1d

A]001.0

M[A]

001.0

M[A

001.0

MA)My(A)My(A

)My(ATA002.0y

yAA

]y2.0y4.0yy2yyy[APn 22

++γ

++γ

+

+γ++

++++

+γ+

++++++=

+

+

+

+

+

+++++

+++


Table 1: Ranges of Gas-condensate Fluid Data

Parameter Minimum Maximum

Pd (psia) 2700 7465

TR (oF) 100 310

PSP (psia) 60 1215

RSP (scf/SP bbl) 3,400 150,000

TSP (oF) 70 190

APIC7+ 42 56

γgR 0.7089 1.466

γgSP 0.66 0.82

CO2 (mole %) 0.12 3.93

H2S (mole %) 0.0 9.32

C1 (mole %) 57.7 83.9

C7+ (mole %) 0.53 13.0

Table 2: Statistical Accuracy of Dewpoint Pressure Correlations for Model Development Data

Correlation Ea

(%)Emax

(%)s

(%)

Organick and Golding 32.56 53.38 13.50

Nemeth and Kenndey 11.64 37.40 9.83

This study 4.33 15.10 3.34

Correlation

Ea

(%)

Emax

(%)

s

(%)

Organick and Golding

35.43

44.83

9.95

Nemeth and Kenndey

11.29 20.83 5.97

This study (1) 3.72 9.41 2.37

This study (2) 3.35 9.05 2.49

Table 3: Statistical Accuracy of Dewpoint Pressure Correlations for Model Validation Data

Notes: (1) The pseudocritical properties were estimated based on the gas-condensate fluid composition using Kay’s method (Eqs. 6 and 7).

(2) The pseudocritical properties were estimated using new correlation (Eqs. 8 and 9).


Figure 2: Sensitivity of new model to reservoir temperature

Dew

po

int

Pre

ssu

re (

psi

a)Reservoir Temperature (oF)

2000

3000

4000

5000

6000

7000

8000

50 100 150 200 250 300 350

Minimum Values

Maximum Values

Average Values

Figure 2: Sensitivity of new model to reservoir temperature

Dew

po

int

Pre

ssu

re (

psi

a)Reservoir Temperature (oF)

2000

3000

4000

5000

6000

7000

8000

50 100 150 200 250 300 350

Minimum Values

Maximum Values

Average Values

Figure 3: Sensitivity of new model to gas-oil ratio

Dew

po

int

Pre

ssu

re (

psi

a)

Gas Oil-Ratio (scf/SP bbl)

1000

3000

5000

7000

9000

11000

4,000 14,000 24,000 34,000 44,000 54,000

Minimum Values

Maximum Values

Average Values

Figure 3: Sensitivity of new model to gas-oil ratio

Dew

po

int

Pre

ssu

re (

psi

a)

Gas Oil-Ratio (scf/SP bbl)

1000

3000

5000

7000

9000

11000

4,000 14,000 24,000 34,000 44,000 54,000

Minimum Values

Maximum Values

Average Values

Es

tim

ate

d D

ew

po

int

Pre

ss

ure

(p

sia

)

Experimental Dewpoint Pressure (psia)

Figure 4: Crossplot of Organick and Golding correlation

2000

3000

4000

5000

6000

7000

8000

2000 3000 4000 5000 6000 7000 8000

+10%

Es

tim

ate

d D

ew

po

int

Pre

ss

ure

(p

sia

)


Figure 4: Crossplot of Organick and Golding correlation

2000

3000

4000

5000

6000

7000

8000

2000 3000 4000 5000 6000 7000 8000

+10%

Figure 1: New model error distribution

Dat

a F

req

uen

cy

(%)

Average Absolute Error (%)

0

10

20

30

40

50

60

70

80

90

100

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

Figure 1: New model error distribution

Dat

a F

req

uen

cy

(%)

Average Absolute Error (%)

0

10

20

30

40

50

60

70

80

90

100

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

8 A.A. HUMOUD, M.A. AL-MARHOUN SPE 68230A

vera

ge

Ab

solu

te E

rro

r (%

)

Dewpoint Pressure (psia)

0

5

10

15

20

25

30

35

40

45

50

4000 5000 6000 7000 8000

This Study

Nemeth & Kennedy

Organick & Golding

< 4000(6)*

4000-5000(16)

5000-6000(31)

6000-7000(15)

> 7000(6)

Figure 7: Accuracy of Correlations for ranges of dewpoint pressures (* number of data points)

Ave

rag

e A

bso

lute

Err

or

(%)

Dewpoint Pressure (psia)

0

5

10

15

20

25

30

35

40

45

50

4000 5000 6000 7000 8000

This Study

Nemeth & Kennedy

Organick & Golding

< 4000(6)*

4000-5000(16)

5000-6000(31)

6000-7000(15)

> 7000(6)

Figure 7: Accuracy of Correlations for ranges of dewpoint pressures (* number of data points)

Av

erag

e A

bso

lute

Err

or

(%)

Reservoir Temperature (oF)

< 150(8)*

150 -200(17)

> 300(10)

200 -250(8)

250 -300(31)

0

5

10

15

20

25

30

35

40

45

100 150 200 250 300

This Study

Nemeth and Kennedy


Figure 8: Accuracy of Correlations for ranges of reservoir temperature (* number of data points)

Av

erag

e A

bso

lute

Err

or

(%)

Reservoir Temperature (oF)

< 150(8)*

150 -200(17)

> 300(10)

200 -250(8)

250 -300(31)

0

5

10

15

20

25

30

35

40

45

100 150 200 250 300

This Study

Nemeth and Kennedy


Figure 8: Accuracy of Correlations for ranges of reservoir temperature (* number of data points)

Est

imat

ed D

ewp

oin

t P

ress

ure

(p

sia)


Figure 5: Crossplot of Nemeth and Kennedy correlation

2000

3000

4000

5000

6000

7000

8000

2000 3000 4000 5000 6000 7000 8000

+10%

Est

imat

ed D

ewp

oin

t P

ress

ure

(p

sia)


Figure 5: Crossplot of Nemeth and Kennedy correlation

2000

3000

4000

5000

6000

7000

8000

2000 3000 4000 5000 6000 7000 8000

+10%

Est

imat

ed D

ewp

oin

t P

ress

ure

(p

sia)


Figure 6: Crossplot of new correlation

2000

3000

4000

5000

6000

7000

8000

2000 3000 4000 5000 6000 7000 8000

+10%

Est

imat

ed D

ewp

oin

t P

ress

ure

(p

sia)


Figure 6: Crossplot of new correlation

2000

3000

4000

5000

6000

7000

8000

2000 3000 4000 5000 6000 7000 8000

+10%

Documents

Gas Condensate Dew Point