Day_4

PVT & EOS Modelling

Using PVTsim Software
Eng.Mohamed Abdel Ghani
Senior Reservoir Engineer
AGENDA
• Example for Gas condensate CVD QC
• PVT Data Correction
• Error Evaluation
• Correction Technique
• Introduction for PVT modelling
• Equation Of State Theory
• Equation Of State Models
• Reservoir Fluid Characterization Using Software (C7+ Characterization):
• Splitting
• Lumping
• Regression control & Matching
PVT QC for Gas Condensate
Reservoirs Samples
PVT QC for gas condensate reservoirs
• PVT QC for gas condensate reservoirs incorporates some steps:
• Sample validation:
1. Validity check as oil reservoirs
2. Contamination check as oil reservoirs
3. Composition check as oil reservoirs
• CVD check:
1. Comparing to CCE liquid drop out
2. Material balance check (forward & Backward)
3. K values versus Pressure

4
 PVT QC for gas condensate reservoirs
• PVT QC for gas condensate reservoirs incorporates some steps:
• CVD check:
1. Comparing to CCE liquid drop out

VALID NOT VALID

5
 PVT QC for gas condensate reservoirs
• CVD check:
2. Material balance check (forward & Backward)
• Forward Material balance check: Subtracting the
evolved gas compositions from the initial
composition and compare the calculated residual
liquid composition versus the lab measured
composition.
• Backward Material balance check: adding the
evolved gas compositions to the lab measured
residual liquid composition to get initial composition
then compare the initial composition versus the lab
measured composition
 PVT QC for gas condensate reservoirs
• PVT QC for gas condensate reservoirs incorporates some steps:
• CVD check:
3. K values versus Pressure

7
 PVT QC for gas condensate reservoirs
• PVT QC for gas condensate reservoirs incorporates some steps:

8
PVT Data Correction
 PVT Data Correction
• Laboratory PVT data can be not representative due to:
1. In partially depleted reservoirs or in fields that originally existed at the bubble-point
pressure, it is difficult to obtain a fluid sample, which usually represents the original oil
in the reservoir at the time of discovery.
2. Also, in collecting fluid samples from oil wells, the possibility exists of obtaining
samples with a saturation pressure that might be lower than the actual saturation
pressure of the reservoir.
3. Human errors.
• In these cases, it is necessary to correct or adjust the laboratory PVT measured data to
reflect the actual saturation pressure.

PVT Experiments 10
 PVT Data Correction
• How to know the correct saturation pressure: 1
1. Allocation initial reservoir pressure and temperature on P-T
diagram.
2. Monitor production performance.
3. Select suitable correlation by having the basic analysis for the
crude: API,GOR, Gas gravity.

2
 PVT Data Correction
• The proposed correction procedure for adjusting the following laboratory test data is
described in the subsequent sections:
1. Constant-composition expansion (CCE) test
2. Differential expansion (DE) test
3. Oil viscosity test
4. Separator tests

PVT Experiments 12
 PVT Data Correction
• Constant-composition expansion (CCE) test correction:
• Below Pb:
1. Calculate the Y-function,
2. Plot the values of the Y-function versus pressure on a Cartesian scale and draw the best
straight line to get a,b coeffients.
3. Recalculate the relative volume Vrel values using the “new” saturation pressure

• Above Pb:
1. Plot the “old” relative volume values above the “old” saturation pressure versus
pressure on a regular scale and draw the best straight line through these points.
2. Calculate the slope of the Line S
3. Draw a straight line that passes through the point (Vrel = 1, psat new) and parallel to
the line of Step 1.
13
 PVT Data Correction
• Differential expansion (DE) test correction:
• The laboratory-measured Bod data must be corrected to account for the new bubble-
point pressure pb new. The proposed procedure is summarized in the following steps:
• Relative oil volume Bod versus pressure:
1. Plot the Bod data versus gauge pressure on a regular scale.
2. Draw the best straight line through the middle pressure range of 30%–90% pb.
3. Extend the straight line to the new bubble-point press
4. Transfer any curvature at the end of the original curve, i.e., ΔBo1 at pb old, to the new
bubble-point pressure by placing ΔBol above or below the straight line at pb new.
5. Repeat the above process and draw a curve that connects the generated Bod points
with original curve at the point of intersection with the straight line.

14
 PVT Data Correction
• Differential expansion (DE) test correction:
• Solution gas-oil ratio:
• The correction procedure for the gas-oil ratio Rsd data is identical to that of Bod except

15
 PVT Data Correction
• Differential expansion (DE) test
correction:
• Correcting Oil Viscosity Data
• Below Pb: 1/μo

1. Defining the fluidity as the reciprocal of the

oil viscosity, i.e., 1/μo,
2. Plot fluidity versus pressure on a Cartesian
scale
3. Draw the best straight line through the
points and extend it to the new saturation
pressure pb new.
4. New oil viscosity values above pb are read μo
from the straight line.

16
 PVT Data Correction
• Differential expansion (DE) test
correction:
• Correcting Oil Viscosity Data
• Above Pb:
1. Plot the viscosity values for all points above
the old saturation pressure
2. Through the point on the extended viscosity
curve at pb new, draw a straight line (Line B)
parallel to A.

17
 PVT Data Correction
• Correcting the Separator Tests Data
• Solution gas-oil ratio:
• The total gas-oil ratio Rsfb is changed in the same proportion as the differential ratio was
changed.

• Formation volume factor:

• The separator oil formation volume factor Bofb is adjusted in the same proportion as the
differential liberation values:

• Note : No corrections are needed for the stock-tank gas-oil ratio and the stock-tank API
gravity.
Introduction for PVT modelling
Modelling Definition
• Modelling is when we have a phenomena that can
be measured by numbers then we relate the
phenomena parameters by an equation or
correlation.
• Two types of mathematical relations:
1. Equation is a result of physical experiment results like
PV=nRT
2. Correlation is just trying to tie some parameters in a
mathematical relation without any physical
experimental base like Correlation of Z factor with GOR= A*Qo+Rs
critical pseudo pressure and temperature.
PVT Modelling Meaning
• PVT modelling also is to try to get a relation between PVT properties
like FVF (Bo) or Oil viscosity and Oil API or Dissolved gas amount
(GOR) by changing pressure and temperature.
• Example
PVT Modelling Importance
• Why is PVT modelling important??
• In industry we have two cases in which we need PVT modelling:
• Blind case : As in many cases we don’t have a PVT sample to get experimental data so we
use the well test available data like API or gas gravity or composition to get reservoir fluid
properties like FVF and bubble point (ex; Green field with no PVT data)
• Extrapolating case (History matching): we can have experimental data in certain range
of pressure and temperature and we need to extrapolate these results to higher or lower
range of pressure and temperature. Also if we have a certain experiment (ex;CCE) and
we need the data for another not available experiment (ex; DV).
• Finally in all cases, when we go to reservoir simulation we have two options
to input PVT data:
1. Lab experimental data as a table so the simulator will have a higher time in simulation
due to interpolating between values.
2. PVT model “Mathematical relation” so the simulator simply will use the mathematical
relation so it will take less time in simulation “substitute and get result”
PVT Modelling Methods and Criteria for Selection
• Actually there are two methods for PVT Modelling:
1. Black Oil / Modified Black Oil Modelling method “Correlations”
2. Compositional modelling method
1.Black oil modelling:
• This method simply relates the output parameters like FVF or Rs by another
input parameters like API or gas specific gravity, pressure and temperature
regardless of the composition of the reservoir fluid.
• It doesn’t see the composition
• Black oil is just a name but it is used also in modelling dry gas properties.
• Example:
PVT Modelling Methods and Criteria for Selection
2.Copositional modelling:
• This method depends on the properties (Pc, Tc, Accenetric Component Mole%
factor, Critical volume, etc….) of each component in the N2 0.38
reservoir fluid like methane or propane to predict the CO2 1.37
C1 43.45
reservoir properties like FVF (Bo) or the solution gas (Rs). C2 7.84
• The equation on which this modelling depends is called the C3 5.3
equation of state (EOS). IC4 1.14
nC4 2.93
• In case we don’t have any experimental data except IC5 1.16
composition, we can use default values for components nC5 1.6
properties. C6 2.23
• Generally, components properties are used as default except C7+ 32.6
Total 100
C7+ that will need to be adjusted until having match to
available experimental data.
Introduction for EOS and PVT modelling
Component Mol % Mol wt Liquid Density lb/ft³ Crit T °F Crit P psia Acentric factor - Normal Tb °F Wt av Mol wt Crit V ft³/lb-mol
N2 0.546 28.014 -232.51 492.32 0.04 -320.35 28.014 1.44
CO2 2.826 44.01 87.89 1069.87 0.225 -109.3 44.01 1.51
C1 55.566 16.043 -116.59 667.2 0.008 -258.79 16.043 1.59
C2 8.594 30.07 90.05 708.35 9.80E-02 -127.39 30.07 2.37
C3 5.745 44.097 205.97 615.76 0.152 -43.69 44.097 3.25
iC4 1.009 58.124 274.91 529.06 0.176 10.85 58.124 4.21
nC4 2.435 58.124 305.69 551.1 0.193 31.19 58.124 4.08
iC5 0.895 72.151 369.05 490.85 0.227 82.13 72.151 4.9
nC5 1.24 72.151 385.61 489.38 0.251 96.89 72.151 4.87
C6 1.581 86.178 41.4522 453.65 430.59 0.296 155.75 86.178 5.93
C7 2.552 91.5 46.0718 491.242 493.69 0.4548 197.51 91.5 7.31
C8 2.747 101.2 47.7574 525.907 463.13 0.4832 242.15 101.2 7.41
C9 1.699 119.1 48.7562 574.858 394.13 0.5345 288.05 119.1 8.34
C10-C12 3.542 146.343 50.1169 640.869 330.68 0.6126 368.397 147.166 9.94
C13-C14 1.788 182.087 51.5613 713.289 282.79 0.7071 458.087 182.395 12.13
C15-C16 1.434 213.559 52.5215 769.832 256.52 0.7873 525.558 213.858 14.25
C17-C18 1.15 243.614 53.3637 819.538 240.01 0.86 583.12 243.815 16.34
C19-C21 1.312 275.308 54.2783 869.466 228.94 0.9331 639.197 275.778 18.63
C22-C25 1.193 322.923 55.3835 938.637 217.12 1.0336 714.418 323.599 22.21
C26-C30 0.912 384.68 56.5878 1021.744 207.76 1.1467 799.617 385.716 27.03
C31-C38 0.726 470.995 57.9802 1129.78 200.71 1.2681 893.182 473.096 34.09
C39-C80 0.507 656.242 60.3275 1354.515 196.43 1.32 1062.817 675.531 51.33
PVT Modelling Methods and Criteria for Selection
• Reservoir Fluid Typing for modelling method selection criteria:
1. Black oil model: Generally we use black oil modelling in case of reservoir
fluid without high compositional variations during depletion or reservoir
production time life like in case of Heavy oil and ordinary black oil & dry gas
reservoir.
2. Modified Black oil model: Volatile oil and wet gas reservoirs
3. Compositional (EOS) modelling: Near critical fluids (Oil or gas) and gas
condensate reservoirs.
How do we enter PVT data for simulation
REMEMBER!!
software? Black oil means:
Black oil
reservoirs or Dry
Gas

LAB BLACK OIL

Experiements RESERVOIR 3
Using EOS
1 2 tuning then
export the
SAMPLE
results as tables
Enter The
Choose a correlation
Properties as a
that best matching
Table
the actual lab data
The Simulation will
The Simulation will - In case you have only the
use the table and composition
interpolate between use the correlation
- The Simulation will use the
table values directly (Saving table and interpolate between
interpolation time) table values
How do we enter PVT data for simulation
REMEMBER!!
software? Black oil means:
Volatile oil
reservoirs or
Wet Gas

MODIFIED
LAB BLACK OIL
Experiements RESERVOIR 3
Using EOS
1 2 tuning then
export the
SAMPLE
results as tables
Enter The
Choose a correlation
Properties as a
that best matching
Table
the actual lab data
The Simulation will
The Simulation will - In case you have only the
us the table and composition
interpolate between use the correlation
- The Simulation will us the
table values directly (Saving table and interpolate between
interpolation time) table values
How do we enter PVT data for simulation
software?

Very Volatile
LAB Oil or Gas
Experiements Condensate

Only one
SAMPLE Solution

Using EOS
tuning then
export the
model
How do we enter PVT data for simulation
software?

Using
Only one Solution
Correlations

No
SAMPLE
PVT Modelling Methods and Criteria for Selection
• Before Modelling:
• Data Preparation:
1. For black oil modelling:
• All the available experimental data like API, GOR
• Remember, we just select the correlations that best match our data
• Any predicted data doesn’t depend on matching (not consistent
process)
2. For compositional modelling:
• We need the composition and all the experimental data
• We adjust the components properties (especially C7+) until having
match
• The predicted data is dependent on the component properties
adjusted by matching (Consistent process)
Introduction for EOS and PVT modelling
Component Mol % Mol wt Liquid Density lb/ft³ Crit T °F Crit P psia Acentric factor - Normal Tb °F Wt av Mol wt Crit V ft³/lb-mol
N2 0.546 28.014 -232.51 492.32 0.04 -320.35 28.014 1.44
CO2 2.826 44.01 87.89 1069.87 0.225 -109.3 44.01 1.51
C1 55.566 16.043 -116.59 667.2 0.008 -258.79 16.043 1.59
C2 8.594 30.07 90.05 708.35 9.80E-02 -127.39 30.07 2.37
C3 5.745 44.097 205.97 615.76 0.152 -43.69 44.097 3.25
iC4 1.009 58.124 274.91 529.06 0.176 10.85 58.124 4.21
nC4 2.435 58.124 305.69 551.1 0.193 31.19 58.124 4.08
iC5 0.895 72.151 369.05 490.85 0.227 82.13 72.151 4.9
nC5 1.24 72.151 385.61 489.38 0.251 96.89 72.151 4.87
C6 1.581 86.178 41.4522 453.65 430.59 0.296 155.75 86.178 5.93
C7 2.552 91.5 46.0718 491.242 493.69 0.4548 197.51 91.5 7.31
C8 2.747 101.2 47.7574 525.907 463.13 0.4832 242.15 101.2 7.41
C9 1.699 119.1 48.7562 574.858 394.13 0.5345 288.05 119.1 8.34
C10-C12 3.542 146.343 50.1169 640.869 330.68 0.6126 368.397 147.166 9.94
C13-C14 1.788 182.087 51.5613 713.289 282.79 0.7071 458.087 182.395 12.13
C15-C16 1.434 213.559 52.5215 769.832 256.52 0.7873 525.558 213.858 14.25
C17-C18 1.15 243.614 53.3637 819.538 240.01 0.86 583.12 243.815 16.34
C19-C21 1.312 275.308 54.2783 869.466 228.94 0.9331 639.197 275.778 18.63
C22-C25 1.193 322.923 55.3835 938.637 217.12 1.0336 714.418 323.599 22.21
C26-C30 0.912 384.68 56.5878 1021.744 207.76 1.1467 799.617 385.716 27.03
C31-C38 0.726 470.995 57.9802 1129.78 200.71 1.2681 893.182 473.096 34.09
C39-C80 0.507 656.242 60.3275 1354.515 196.43 1.32 1062.817 675.531 51.33
 1.Equation of state theory
• An equation simulates each component behaviour in the composition
under various operational conditions.
• Very volatile oil (API>40) or gas condensate fluids require compositional
modelling as the available black oil correlations can lead errorneous
results.
• Why we need EOS??
1. Check if PVT analysis reasonable or not.
2. QC PVT tests
3. Generate PVT data for the ranges outside the lab measurements
4. Extract matched PVT composition to be used in compositional model(when
composition change with production like gas condensate reservoir).
5. Extract new PVT properties for different temperature or GOR.
6. The matched model working like virtual lab.
 1.Equation of state theory
History of EOS
• 1662: Boyle’s Law -PV = constant at a fixed T • 1901: Lewis -Fugacity
• 1787: Charles Law -ΔV is proportional to ΔT at • 1940: Benedict-Webb-Rubin -Eight constant
constant P EOS
• 1801: Dalton’s Law -P = sum of the partial pressure • 1949: Redlich-Kwong -Two parameter cubic
• 1802: Cagniard de la Tour -Discovery of critical EOS
state
• 1972: Soave -modification of Redlich-Kwong
• 1834: Clapeyron -Combined Boyles Law Temperature dependent attraction parameter.
• and Charles Law into PV=RT • 1976: Peng-Robinson -Two-parameter EOS
• 1873: van der Waals -First EOS and idea of • 1978: Peng-Robinson -improved the “α” term
corresponding states in the model for heavier components
• 1880: Amagat’s Law -Volume of mixtures of gases • 1980: Schmidt-Wenzel -Two-parameter EOS
= sum of pure components volumes
does a better job in predicting liquid densities
• 1885:Thiesenand 1901 OnnesViral EOS -viral
equation in density, viral equation in pressure • 1982: Peneloux et al.-A consistent volume
correction for cubic EOS
 2.Equation of state models
1. Ideal Gas Law
2. Real Gas Law
3. Cubic EOS’s:
A. Van der Waals (1873)
B. Redlich-Kwong (1948)
C. Soave -modification of Redlich-Kwong (1972)
D. Peng-Robinson (1976/1978)
 2.Equation of state models
1. Ideal Gas Law:
1. PV=RT
2. Volume of the gas molecules is small with respect to the volume occupied
3. No intermolecular forces (attractive/repulsive)
4. All collisions are perfectly elastic
 2.Equation of state models
2. Real Gas Law:
1. PV=ZRT
2. All non-idealities are lumped into a single term, Z
3. Z=f(P,V,T, composition), function of properties of the constituents
 2.Equation of state models
3. Cubic EOS’s:
A. Van der Waals (1873):
1. P=RT/(V-b) -a/V^2
2. “b” is the correction for finite volume of gas
molecules (sometimes referred to as the excluded
volume or repulsion parameter)
3. “a” is the correction for molecular attraction
4. “a” and “b” are functions of the subject
component’s properties (Pc, Tc, Vc).
5. Called “2-Parameter Cubic EOS”
 2.Equation of state models
3. Cubic EOS’s:
A. Van der Waals (1873):
How To determine the “a” & “b”
 He observed that the critical isotherm has
a horizontal slope and an inflection point
at the critical point, as in the shown graph
 This can be expressed mathematically by

 Differentiating the equation with respect

to the volume at the critical point will
results in
 2.Equation of state models
3. Cubic EOS’s:
A. Van der Waals (1873):
• Solving the last two equation together will yield
a and b equations
• Combining the equations with Van der Waals
EOS equation at the critical point will give a and
b values
• Law of Corresponding States (van der Waals
EOS):Two substances should have similar
properties at corresponding conditions with
respect to their critical Temperature (Tc) and
Pressure (Pc)

40
 2.Equation of state models
3. Cubic EOS’s:
A. Van der Waals (1873) application:

PV=ZRT

41
 2.Equation of state models
3. Cubic EOS’s:
A. Van der Waals (1873):
• Gives qualitatively good estimation for the hydrocarbon properties
• All EOS has been developed to pure components first, and then
extended to mixtures through the use of mixing rules
• The mixing rules are simply means of calculating mixture parameters
equivalent to those of pure substances

Phase Equilibria 42
 2.Equation of state models
Do You Remember mixing rules??
 2.Equation of state models
3. Cubic EOS’s:
B. Redlich-Kwong (RK) (1948) EOS:
1. Prediction of the volumetric and physical properties of the vapor phase could be
improved by simple adjustment for the van der Waals attractive pressure term “a /V2 “
2. Temperature correction is added to the denominator of the second term
3. a and b are functions of the subject components properties (Pc, Tc)

The new attractive term

 2.Equation of state models
3. Cubic EOS’s:
B. Redlich-Kwong (1948) EOS application:

In Case of Mixtures

Phase Equilibria 45
 2.Equation of state models
3. Cubic EOS’s:
C. Soave’s modification of Redlich-Kwong (SRK) (1972):
1. Replace the term (a/T0.5) with a temperature –dependent term aa(T)
2. A *a (T) is dimensionless factor
 Acentric Factor
• Pitzer (1955): introduce w as a correlating parameter to characterize
the centricity or nonsphericity of a molecule.

• Vapor Pressure (Pv):

The pressure exerted by a vapor in equilibrium with its liquid
• Equilibrium :
The number of molecules leaving one phase and returning to the
same phase is equal
Phase Equilibria 47
 Vapor Pressure
• The relationship of Pv and T is
straight line on Cox chart.
• Lee and Kesler (1975) proposed the
following generalized vapor
pressure equation:
 2.Equation of state models
3. Cubic EOS’s:
C. Soave’s modification of Redlich-Kwong (SRK) (1972) application:

In Case of Mixtures, Soave adopted the following mixing rules:

Phase Equilibria 49
 Binary interaction Coefficient
• An empirically determined correction factor
• Designed to characterize any binary system formed by components i and j
in the hydrocarbon mixture.
• Used to model the intermolecular interaction through empirical
adjustment of the (aα) term
• The Kij between HC increases as the relative difference between their Mwt
increases
• If components have the same Mwt kij =0
• Graboski and Daubert (1978) and Soave (1972) suggested that no binary
interaction coefficients are required for hydrocarbon systems
• However, with nonhydrocarbons present, Kij can greatly improve the
predictions of SRK
 2.Equation of state models
3. Cubic EOS’s:
• In the two-phase region, the largest positive root corresponds to the
compressibility factor of the vapor phase, Zv, while the smallest positive root
corresponds to that of the liquid phase, ZL.
 2.Equation of state models
3. Cubic EOS’s:
C. Soave’s modification of Redlich-
Kwong (SRK) (1972) modifications:
4. (1982 ) Peneloux, Rauzy, and Freze • To get ci
introduce the ci “volume correction
factor” to improve the volumetric
predictions of the SRK equation • Where ZRA = Rackett compressibility
• Ci only affect the liquid and gas factor
volumes “GOR”
• The procedure for the calculation is • IF ZRA is not available Peneloux
called “Volume translation method”
proposed

• Yamada and Gunn proposed this

correlation
52
 2.Equation of state models
3. Cubic EOS’s:
D. Peng and Robinson (1976a):
• They illustrated the need for an improvement in the ability of the equation of
state to predict liquid densities and other fluid properties, particularly in the
vicinity of the critical region.
• Peng and Robinson (PR) proposed the following expression:
 2.Equation of state models
3. Cubic EOS’s:
C. Peng and Robinson (1976a) application:

Used if
w>0.49 in
PR 78

In Case of Mixtures, same Soave’s method is adopted:

Phase Equilibria 54
 2.Equation of state models
3. Cubic EOS’s:
C. Peng and Robinson (1976a) application:
Jhaveri and Youngren (1984) pointed out that, when applying the Peng-Robinson
equation of state to reservoir fluids, the error associated with the equation in the prediction
of gas-phase Z-factors ranged from 3 to 5% and the error in the liquid density predictions
ranged from 6 to 12%.

Phase Equilibria 55
 2.Equation of state models
3. Cubic EOS’s:
C. Peng and Robinson (1976a) application:
Following the procedure proposed by Peneloux and coworkers (see the SRK EOS), Jhaveri and
Youngren introduced the volume correction parameter, ci, to the PR EOS.

where Si is a dimensionless parameter, called the shift parameter, and bi is the Peng-Robinson
covolume,

Phase Equilibria 56
 2.Equation of state models
3. Cubic EOS’s general formula:

Phase Equilibria 57
Introduction for EOS and PVT modelling
Component Mol % Mol wt Liquid Density lb/ft³ Crit T °F Crit P psia Acentric factor - Normal Tb °F Wt av Mol wt Crit V ft³/lb-mol
N2 0.546 28.014 -232.51 492.32 0.04 -320.35 28.014 1.44
CO2 2.826 44.01 87.89 1069.87 0.225 -109.3 44.01 1.51
C1 55.566 16.043 -116.59 667.2 0.008 -258.79 16.043 1.59
C2 8.594 30.07 90.05 708.35 9.80E-02 -127.39 30.07 2.37
C3 5.745 44.097 205.97 615.76 0.152 -43.69 44.097 3.25
iC4 1.009 58.124 274.91 529.06 0.176 10.85 58.124 4.21
nC4 2.435 58.124 305.69 551.1 0.193 31.19 58.124 4.08
iC5 0.895 72.151 369.05 490.85 0.227 82.13 72.151 4.9
nC5 1.24 72.151 385.61 489.38 0.251 96.89 72.151 4.87
C6 1.581 86.178 41.4522 453.65 430.59 0.296 155.75 86.178 5.93
C7 2.552 91.5 46.0718 491.242 493.69 0.4548 197.51 91.5 7.31
C8 2.747 101.2 47.7574 525.907 463.13 0.4832 242.15 101.2 7.41
C9 1.699 119.1 48.7562 574.858 394.13 0.5345 288.05 119.1 8.34
C10-C12 3.542 146.343 50.1169 640.869 330.68 0.6126 368.397 147.166 9.94
C13-C14 1.788 182.087 51.5613 713.289 282.79 0.7071 458.087 182.395 12.13
C15-C16 1.434 213.559 52.5215 769.832 256.52 0.7873 525.558 213.858 14.25
C17-C18 1.15 243.614 53.3637 819.538 240.01 0.86 583.12 243.815 16.34
C19-C21 1.312 275.308 54.2783 869.466 228.94 0.9331 639.197 275.778 18.63
C22-C25 1.193 322.923 55.3835 938.637 217.12 1.0336 714.418 323.599 22.21
C26-C30 0.912 384.68 56.5878 1021.744 207.76 1.1467 799.617 385.716 27.03
C31-C38 0.726 470.995 57.9802 1129.78 200.71 1.2681 893.182 473.096 34.09
C39-C80 0.507 656.242 60.3275 1354.515 196.43 1.32 1062.817 675.531 51.33
 8.C7+ characterization by regression.
1. The parameters of the EOS are often adjusted to honor the available data. Such
adjustments to regress on the available data is often referred as “Tuning” in
the industry.
2. For a consistent set of lab data, it is often necessary to modify the EOS
parameters (the parameters which impact a and b).
3. Most of the proposed methods in the literature suggest modification of the
plus fraction properties. In general, the attributes (Tc, Pc, w) of the plus fraction
constituents, or the coefficients of the a and b (ΩA and ΩB, respectively) are
modified during the tuning process.
4. For systems containing significant amount of non-HC components (some EOR
fluid systems) non-HC-C7+ kij’s may also become a key set parameters.
5. After proper “tuning”, the subject fluid is considered to be “Characterized”.
Why do we have C7+ not before that like C5+
or C4+ or lighter?
• In the laboratory the samples are flashed to standard conditions
before making any analyses.
• Flashing the oil results in a gas and a liquid sample that are analyzed
separately.
• The gas will always be analyzed by a gas chromatographic (GC)
analysis. Two alternative types of fluid analyses are used for the
liquid. These are a gas chromatographic (GC) analysis and a true
boiling point (TBP) analysis.
• None of these analyses will identify all the chemical species contained
in the fluid but will separate the C7+ fraction into boiling point cuts.
 OR
Oil
TBP GC

Reservoir
Fluid Compositional
Flash
Sample Analysis

GAS
GC
Splitting & Lumping
 Spiliting for the Plus fluids (Lumping option)
• It is difficult to characterize the plus fraction properly in terms of their
critical properties and Acentric factors
• Erroneous predictions and conclusions can be resulted if If the plus
fraction used as a single component
• Several authors indicated that theses errors substantially decreased
with Splitting
• So, How to adequately split the + fraction.??????
Splitting Schemes
• Is a procedure to divide the +fraction into
HC groups and identified physical
properties
• Mwt or Number of carbon number was
used to extend the molar distribution
• The splitting schemes was based on the
observation that:
• The lighter systems (eg. Condensates)
exhibits exponential molar distribution
• Heavier systems show left-skewed
distributions

Reservoir Fluid Characterization 64

Lumping Schemes
• Normally EOS model has large number of components
• For PVT modeling purposes it is OK
• But for Compositional modeling Too large for practical purposes due to large
cost and time consuming
• So, small no. of Components is needed for compositional modeling
• To decrease the no of components Lumping is needed

Reservoir Fluid Characterization 65

Lumping Schemes
• Lumping means:
• Grouping of pure components together and represent them by Single Carbon
Number (SCN)
• Two problem are associated with Lumping:
• How to select the groups?
• What mixing rules should be used for determining the physical properties?

Reservoir Fluid Characterization 66

Equation of state sensitivity

Phase Equilibria 67
Matching Tolerances for Gas and gas Condensates

Physical property Accepted range

Dew point Pressure 2-5%
Liquid Drop out ± 10%
Stock tank liquid ± 2%
Density

Phase Equilibria 68
Matching Tolerances for Crude oils
Physical property Accepted range
Bubble point Pressure ± 2%
GOR (field)
1st Stage ± 10%
2nd and 3rd Stages ± 30%

GOR (Lab) ± 5%
Stock tank liquid Density ± 2%

Phase Equilibria 69
Regresion weights setting in PVTsim