SPE 71703
CAPILLARY PRESSURE DATA FROM NMR LOGS AND ITS IMPLICATIONS ON
FIELD ECONOMICS
M. Altunbay, Baker Atlas, R. Martain, Baker Atlas, M. Robinson, Baker Atlas
Copyright 2001, Society of Petroleum Engineers Inc.
This paper was prepared for presentation at the 2001 SPE Annual Technical Conference and
Exhibition held in New Orleans, Louisiana, 30 September–3 October 2001.
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
The analogous behaviors of capillary pressure and 1/T2 decay
versus saturation data provide a methodology for deriving
synthetic capillary pressure information directly from NMR
logs even if core data are not available. Estimations of
irreducible water saturation4 from NMR log-based synthetic
capillary pressure profiles have numerous benefits. One
benefit is the enhanced accuracy of permeability estimations
by using synthetic bulk volume irreducible in the bound water
model. Availability of synthetic bulk volume irreducible data
from capillary pressure modeling eliminates the requirements
for determining the T2cutoff time from core analysis data or
assuming generally accepted constants for it (33 ms for
clastics and 92 ms for carbonates). In addition to the bound
water model, an alternative permeability profile can also be
generated from Purcell's7 capillary pressure-permeability
relation. Another benefit derived from a synthetic capillary
pressure profile is the determination of Oil Water or Gas
Water Contact (OWC, GWC) by simply converting the
capillary pressure information to height above free water level.
This information can also be compared with the Free Water
Level (FWL) from formation test tools. The relative positions
of FWL and OWC assist in inferring wettability14 and the
degree of wettability magnitude for the subject reservoir.
Relative permeability estimations from synthetically created
capillary pressure data provide an avenue to make the first-
approximation profiles of the effective permeability values to
the phases contained in the pore space. Consequently,
effective permeability profiles in conjunction with other
petrophysical and fluid properties and the pressure information
lead to the anticipated production rates and overall evaluation
of the economic viability of a well(s) and field. Applications
of the technique with inherent assumptions, limitations, and
remedies for the limitations are reviewed and explained
through a case study presented in this paper.
Introduction
One of the concerns in dealing with fluid flow through porous
media is the non-chemical interaction between the fluids and
porous media. This interaction results from capillary forces
between the wetting and non-wetting fluids in highly curved
pore structures. A wetting fluid preferentially fills a pore
structure. To remove the wetting fluid from the pore space
requires work to be performed. As an alternative, work must
also be performed to force a non-wetting fluid into a pore
space. The required work for either intrusion or expulsion of
fluids in or from pores provides a measure of the capillary
characteristics for porous media. The resistance to the
intrusion or expulsion of a non-wetting phase into a pore
structure varies with interfacial tension between the wetting
and non-wetting phases, contact angle between the invading
fluid and the pore throat and inversely with the pore diameter.
Intrusion or expulsion of a non-wetting fluid into or from
pores differs from each other in the amount of work they
require. Non-uniform pore shapes and the difference between
the contact angles of an advancing or receding liquid cause
this hysteresis. Therefore, the imbibition and drainage curves
of capillary pressure versus saturation never overlap.
Considering these aforementioned capillary phenomenon
characteristics, outlining the dependencies of capillary
characteristics for porous media before comparing laboratory
produced results against their theoretical approximations is
imperative. Pore geometry, rock wettability, interfacial
tension, saturation history and the difference in fluid densities
affect the magnitude of capillary pressure. Hence, the
comparison of any synthetic capillary pressure data should be
made against the core data obtained from a rock with similar
type of pore geometry and the proper saturation treatment of
the sample.
We have devised a methodology to convert wireline NMR
data into Pc versus Saturation profile for each level of wireline
log data with1 or without the availability of core capillary
2 M. ALTUNBAY, R. MARTAIN, M. ROBINSON SPE 71703

pressure information. The reasonable accuracy of the results In capillary phenomenon, the pressure imbalance between two
grants taking the next step to derive relative permeability phases is mostly controlled by pore throat size. On the
based on capillary pressure, then the effective permeability contrary, S/V term in the T2 decay equation yields the pore
and fractional flow data, and, consequently, leads to a body size. To equate capillary pressure and T2-decay
preliminary look at the well economics. equations would be troubling if not for the fact that commonly
there is a relation between pore-body and pore-throat size, an
Theoretical background: occurrence or a fractal quality more common for clastics than
The governing equation for the T2 decay phenomena for spin- carbonates. As pores deviate from geometrically definable
spin relaxation can be written as follows: shapes into irregular voids with stretched extensions, the
relation between pore-body and pore-throat size deteriorates
1 1 S 1 and becomes more difficult to quantify. Consequently, the
= + ρ + γ G 2 TE 2 D ability to compute properties depending on pore-throat
T2 T2bulk V 12
attributes from pore-body attributes becomes less reliable.

Bulk Surface Relaxation due If we assume a geometrically definable shape for pores
Relaxation Relaxation to diffusion (spherical or tube like), however, we can write the surface to
volume ratio in terms of pore size:
Historically, to obtain capillary pressure from NMR
data5,6,2,17,18,19,20 started from the observance of the similarity S Fs
between the inverted magnetization decay of the polarized =
pore-filling fluids and the pore-size distribution obtained from V rb
capillary pressure data for the same sample. NMR T2 decay in for a spherical pore:
porous media is commonly simplified and interpreted in terms S 4 ⋅π ⋅ r 2 3
of pore size distribution (where pore size distribution is used = =
synonymously with pore-body size). In its simplified form, the V 4 r
⋅π ⋅ r 3
T2 relaxation time is proportional to the ratio of the pore 3
volume to pore surface area and is inversely proportional to
surface relaxivity. It should be noted that this relationship is The shape factor (Fs) is 3 for round pores and 2 for cylindrical
valid when pores are water wet, there is no diffusive coupling, tubes. Both equations (Pc and T2-decay) define two different
the T2 distribution is not shifted by the presence of phenomena and can be linked by their common variable pore-
hydrocarbons and the T2 of the bulk fluid is large enough to throat-size (rpt) for the capillary pressure and the pore-body-
drop the term (1/T2bulk) from the preceding equation. size from the T2-decay with an assumption of correlatable
body to throat size. If we substitute the rpt in the capillary
Construction of capillary pressure curves from NMR data1 is pressure equation with the corresponding T2 term from the T2-
based on fundamental definitions of capillary pressure and the decay equation, capillary pressure can be expressed within the
T2-decay phenomena; hence the complexities caused by the terms of T2 decay phenomenon for spin-spin relaxation.
presence of hydrocarbons and their respective placement in
various size pores must be addressed after the introduction of
the linkage between these two. When porous media contain 1 ρ r pt
= Fs T2
more than a single fluid, the pressure for each fluid differs Pc 2 σ cos(θ ) rb
even if the fluids are in a state of pressure equilibrium. The
mutual attraction between porous media and fluid (adhesion By rearranging the above equation and regrouping the surface
tension) differ for each couple, which is the reason for the relaxivity and pore throat to pore body ratio into a variable
pressure imbalance, defined as capillary pressure. This
called C:
concept can be expressed in terms of the following equation:
ρ r pt
C= Fs
2 σ cos (θ ) 2σ cos(θ ) rb
Pc =
rpt
We can simplify the original equation into:
Capillary pressure may be determined from laboratory tests, or
correlations, or calculated from known fluid distributions and  1 
fluid properties. In this study, we will elaborate on the   = C T2
methodology to convert log NMR data into Pc versus  Pc 
saturation profile with or without the availability of core data.
SPE 71703 CAPILLARY PRESSURE DATA FROM NMR LOGS AND ITS IMPLICATIONS ON FIELD ECONOMICS 3

Algebraic manipulation of the above relation between the In most cases, to identify the reasons for the deviation and to
capillary pressure and the T2-decay yields: rectify the scaling coefficient is impossible due to lack of data
or simply lack of rock samples. Of note is the scaling
1 coefficient C does not reflect the effects of magnetic
Log (C ) = Log   − Log (T2 ) susceptibility on the magnetic field. Nevertheless, the C
 Pc  computed as a variable for each saturation level (at each time
segment of the inversion) suffices for the reasonably accurate
This is the coefficient that scales the Sw versus T2 curve to transformation of NMR data into capillary pressure
match as close as possible to Sw versus 1/Pc curve. The information provided that the correction of T2-spectra for non-
coefficient serves as a merit function in the absence of wetting hydrocarbons2 is performed (Volokitin et al. SCA
hydrocarbons. 9924).

1
We use the above methodology to generate scaling coefficient
T2, ms
1/Pc, psi-1
tables for any subject formation by using a porosity-range-
0.9
C qualifier. The porosity-range-qualifier checks NMR log
0.8
C*T2
derived porosity to select the proper C table for that range of
clastics or carbonates for processing the wireline NMR data to
0.7
derive capillary pressure versus saturation profile for the given
0.6 depth level or interval. The accuracy of this procedure
0.5
becomes better when core calibration of the C coefficient is
possible.
Sw
0.4

0.3
In the absence of core data, we devised a methodology to
convert the NMR log data to capillary pressure based on the
0.2
same concept for deriving the proper scaling coefficient.
0.1 Derivation of the scaling coefficient from hydrocarbon-
corrected T2-spectra relies on the analogy between the time
0
0.0001 0.001 0.01 0.1 1 10 100 1000 10000 dependent desaturation of a core plug at a constant capillary
C, 1/Pc, psi-1, C*T2, T2, ms pressure and the time dependency of the partial porosity

Any deviation between the C*T2 and 1/Pc versus Sw curves (as
observed in the very low Sw values) maybe because of the
inapplicability of the selected shape factor or any variability in
the surface relaxivity or possibly both. The following graph
shows the variability of C for different samples emphasizing
the requirement of computing C for each saturation point of
each sample.

1
Smpl1
Smpl2
Smpl3
0.1 Smpl4

0.01
buildup in an NMR experiment. First, the transformation of T2
C versus φi into T2 versus Sw is obtained.
0.001
n =t n

∑ φi
i =t1
0.0001 S wi = 1 −
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
φe
Σ ( φ i) / φ t
4 M. ALTUNBAY, R. MARTAIN, M. ROBINSON SPE 71703

Then, the rate of change in porosity build up is computed. The The transformation is completed with the plot of Pc versus Sw.
rate of change in cumulative partial porosity is reflective of n =t n
the changing pore sizes and also analogous to desaturation rate
at an increasing capillary pressure (increased rpm) in a ∑ φi Ri
i =t1
centrifuge Pc test. S wi = 1 − Pc i =
φe C ⋅ T2i
1

0.9
The theoretical approach of the conversion is based on
0.8 assumptions and inferences. Perfectly ordered spherical grains
0.7
in theory:
0.6
Sw
0.5

0.4

0.3

0.2

0.1

0
0.1 1 10 T2, ms 100 1000

dS wi
mi =
dT2i
The slope change is computed from the initiation time of the
NMR experiment to the end of each time increment. The ratio
of effective pore volume to cumulative slope change for each are replaced with irregular and randomly packed grains of the
increment yields a multiplier for each time slice. depositional environment and diagenetic changes.

φe
Ri = n dS w i
Log (∑ )
i =1 dT2 i

In the absence of core data for calibration of “C” –the merit

function- or -the scaling coefficient-; the derived model
equation based on historical core capillary pressure and NMR
test data is used for generating the C. The model equation for
C is derived for sandstone and carbonates covering a wide
range of shape factors and porosity ranges.

35

carbonate
30
sandstone
A scaling coefficient with a selected shape factor compensates
25
to a certain extent for the deviation from the theoretical
20 starting point. Hence, the accuracy of the transformation relies
C on the data quality (S/N ratio, adequate polarization, good
15
borehole conditions etc), formation characteristics and the
10
proper selection of the shape factor. Of further note is that the
comparison of the NMR log-derived capillary pressure should
5 be made with the laboratory capillary pressure attained at the
0
insitu conditions especially if the formation is unconsolidated.
In addition, the capillary pressure data obtained from the
0 5 10 15 φ 20 25 30 35
transformation reflects the capillary characteristics of the
volume of the rock that is only 1 to 3 in. into the formation –
SPE 71703 CAPILLARY PRESSURE DATA FROM NMR LOGS AND ITS IMPLICATIONS ON FIELD ECONOMICS 5

flushed zone- and only 1 mm thick. Any formation damage The sources for uncertainty in the laboratory test data cannot
will be reflected in the results while the core sample may not be summarized by the error in volumetric readings only, but
have the similar alteration because of cleaning of the samples are sufficient to warn us about the pitfalls of imperfect
prior to the laboratory test. comparisons.

Sensitivity issue The following collaborative testing of the derivation of

Statistical treatment of the laboratory-derived capillary
pressure data indicates that the random causes may create
fluctuating deviations in both sign and magnitude. We
selected volumetric readings during a centrifuge test as one of
the main sources of error. Insufficient stabilization time at a
constant capillary pressure (constant rpm of a centrifuge) or
simple reading errors of the test tube at the minimum
desaturation volumes may cause this error. Hence, we
introduce 10% error in the volumetric readings of a centrifuge
air/brine test and used the Hassler-Brunner method22 with the
parameter estimation technique to compute the irreducible
water saturations. Error propagation has caused the irreducible
water saturation values to have an of ±77% uncertainty level.
50
45
40
35
Pc Lab data
Pc Lab data w/ 10% error in
Volumetric readings
synthetic capillary pressure from NMR data and the centrifuge
testing of core samples are performed to validate the
methodology and evaluate the error margins for both
techniques. First, air /brine centrifuge test follows the proper
test protocol, then the core NMR experiment is run with the
sample at 100% saturation. We used the Hassler-Brunner
algorithm to calculate the end-face and average wetting phase
saturations. Additionally, we used the following equation21
with the parameter estimation technique to compute the Swir
and Pd.
P P P
S w = S wir + (1 − S wir ) d + (1 − S wir ) d Ln ( c )
Pc Pc Pd
Then, the scaling coefficient is computed for each bin in the
T2-spectra by using the porosity-dependent empirical equation
from our database and the T2 distribution is converted to the
3

30 2.5
Pc,psi

25
2
20
φi
15 1.5

10

1
5

0
0.5
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Sw

0
A systematic study of the random errors in the volumetric 0.1 1 10 T2, ms 100 1000 10000

readings of the capillary pressure desaturation measurements

yields the following uncertainties for displacement pressure
capillary pressure versus saturation with a shape factor of 3
(Pd) and irreducible wetting phase saturation Swir.
(spherical pores).
Centrifuge Capillary Pressure Test

100

10
% error in Swirr, Pd

0.1

Error in Swirr
Error in Pd, psi
0.01
0 2 4 6 8 10 12
% error in volumetric readings
6 M. ALTUNBAY, R. MARTAIN, M. ROBINSON SPE 71703

The comparison of Swir and the curve shapes from NMR-based The following graph shows the uncertainty in fractional flow
and centrifuge-based capillary pressure curves indicate an and OWC level determination if the scaling coefficient is ± 50,
approximate deviation of 0.03 to 0.04 saturation units. As 30 and 20% in error.
evident from the above graph, the Hassler-Brunner treatment
of laboratory core data contained a 0.02 to 0.025 saturation
point deviation on its own from the expected trend due to the
characteristic of the model equation. Any other model NMR Pc with 50%
equation would have yielded similar deviations from the error in C
laboratory-derived data points because of its statistical nature. NMR Pc with 30%
error in C
NMR-derived capillary pressure error margins remain at NMR Pc with 20%
acceptable levels if the expectations for OWC (oil-water- error in C
contact) accuracy are comparable to the acquired NMR log
resolution.

We studied the potential error margins in synthetic capillary kr

pressure curves for highly elevated levels of error in the
scaling coefficient C and the shape factor. The propagation
rate of error due to the error in C is much smaller than the
propagation rate of error in laboratory-acquired and processed
data. This results from the lesser number of components in
the RMS (root mean square) equation for the synthetic Pc
derivation then the number of components in the RMS-error
for the Hassler-Brunner solution of Pc based on the laboratory
test.
Applications
Reasonably accurate estimations of capillary pressure versus
saturation profile for every level of wireline log data signals
new avenues for using NMR logs in the domain of reservoir
and production engineering in addition to petrophysical
interpretations. We will elaborate on the immediate
byproducts of the methodology.

1. Bulk volume irreducible (BVI) information independent

of T2cutoff or Spectral BVI coefficients.

NMR-derived capillary pressure curve is treated with a model

equation to extrapolate the curve to very large capillary
pressure levels to attain the theoretical Swir. The same model
also provides the threshold pressure as Sw approaches 100%
by the parameter estimation technique provided that the
hydrocarbon correction of the T2 distribution is performed.
The form of the model equation is as follows:
A seven saturation units uncertainty (or error) is imposed on
irreducible water saturation by a 50% error in the scaling Pd P P
S w = S wir + (1 − Swir ) + (1 − S wir ) d Ln ( c )
coefficient C. This may not be a drastic penalty in Pc Pc Pd
consideration with the better (lesser error) selection of C if Swir
is the only desired parameter. The capillary pressure curve where:
shape is of extreme importance when OWC level(s), relative Sw = wetting phase saturation
permeability, fractional flow and pore-size distribution are to Swir = Irreducible water saturation
be computed. However, the transformation of wireline NMR Pd = Threshold (entry) pressure
data into synthetic capillary pressure curves for each level of Pc = Capillary pressure
the log data provides a wealth of reservoir and production
engineering applications provided that the selection of the The above equation21 satisfies the boundary conditions of the
scaling coefficient and the treatment for hydrocarbon shift of capillary pressure phenomenon: Sw approaches Swir when Pc
the T2-spectra are performed correctly. approaches infinity and Pc approaches a finite number (Pd)
when Sw approaches 100 %. Theoretical irreducible at infinite
SPE 71703 CAPILLARY PRESSURE DATA FROM NMR LOGS AND ITS IMPLICATIONS ON FIELD ECONOMICS 7

capillary pressure helps to eliminate the ambiguity in capillary pressure in atmosphere. A permeability range
permeability models that are based on BVI by equating the varying from 1 to 1450 mD has a lithology factor average of
computed permeability to absolute permeability. 0.216 with minimal variability.

2. Capillary pressure profile 4. Relative permeability by Corey-Burdine12 equations

The capillary pressure profile for the entire coverage of the
NMR log is created. Theoretical (absolute) Swir is computed by
using the preceding model equation. Pore throat size
histograms for any given interval can also be computed
provided the shape factor that links the pore body and pore
throat size is not grossly in error and the hydrocarbon shift of
the T2-spectra is corrected.
Relative permeability to wetting and non-wetting phases is
computed by interrelating the synthetic capillary pressure with
the Corey-Burdine equations. These equations have recently
been re-worked by Huang et al13. for the drainage and
imbibition cycles, and are presented below:
Wetting phase relative permeability:

S w*
1
∫ P 2 dS w
*

( )
k rw = S w*
2
⋅ 0
1
c

1
∫ P 2 dS w
*

0 c

For the non-wetting phase, the equation is as follows:

S w*
1
∫ P 2 dS w
*

(
k rnw = 1 − S w* ) 2
⋅ 0
1
c

1
∫ P 2 dS w
*

0 c

where:
3. Purcell’s permeability model7 Sw* is normalized wetting phase saturation:
Purcell relates the permeability of a unit system of cylindrical S w − S iw
capillaries with various radii to the porosity and the capillary S w* =
pressures and volumes of its components by: 1 − S iw − S or

(σ ⋅ cos θ )2 φ N
S
k =
2 ⋅10 4 ∑ Pw2
i =1 c

The above equation is for a theoretical porous media and does

not consider the interconnectivity of the pore networks; hence
Purcell introduced a lithology factor to account for the
differences between theoretical and hypothetical porous
media. The equation is then reduced to:

Sw =100 dP
k =0.66 ⋅ F φ ∫
Sw= 0
Pc2

where k is the permeability in mD, φ is the porosity in percent,

Sw is the wetting phase saturation in percent and Pc is the
8 M. ALTUNBAY, R. MARTAIN, M. ROBINSON
5. Relative permeability by regression models
Empirical relative permeability models compiled and
derived9,10,11 from the historical SPE database classified by
system, lithology, and wettability are used in conjunction with
the resistivity-based or DTW-based Sw analysis to derive
relative permeability characteristics of the porous media.
6. Fractional flow of water
Fractional flow of water16 can be computed for each level by
using:
1 more responsive and more accurate BVI values then the
fw =
k rnw µ w
1+ ⋅ carbonates.
k rw µ nw
and plotted (fw versus Sw) for any given interval or for the
entire profile.
7. Effective permeability and productivity analysis
The permeability we obtain by using the theoretical Swir (as Pc
goes to infinity Sw approaches to Swir) eliminates the
ambiguity of the permeability type we derive from the Bound
Water Model. The permeability we obtain by using synthetic
BVI (i.e. Swir*φe) with the Bound Water Model is the closest
approximation of the absolute permeability. In addition,
Purcell’s permeability also approaches to the absolute perm.
Having both the relative and absolute permeability enable us
to compute the effective permeability to the wetting and non-
wetting phase. The ability to derive effective permeability to
hydrocarbons helps to compute the correct mD-ft for any
interval that is subject of interest in addition to the first
approximations of the productivity to both phases provided
that the pressure and fluid properties data also available (see
figure on page 9).
SPE 71703
Conclusions
1. Reliable and repeatable capillary pressure information can
be obtained from NMR logs for neutral to water wet
formations provided that the correction for “hydrocarbon
shifting of the T2-spectra” is applied.
2. Derivation of Pc from NMR logs for oil-wet formations
and formations with heavy crude and tar require special
treatment of the T2-spectra.
3. Scaling coefficient must be computed for each time
segment (bin) of the data either using an empirical model
or core calibration when available.
4. Synthetic capillary pressure-based BVI estimations are
T2cutoff or SBVI-based BVI values for both clastics and
5. BVI estimations for carbonate formations are more
accurate with Synthetic Pc-based BVI than other
traditional techniques.
6. Purcell’s permeability model provides an alternative to
the Bound Water and Kozeny models and eliminates the
needs for the calibration of coefficient and exponents of
Bound Water and Kozeny models and the T2cutoff time.
7. Using theoretical BVI (Swir*φe as Pc approaches to
infinity) with the Bound Water Model for permeability
provides a closer approximation of the absolute
permeability then the T2cutoff-based BVI.
8. Relative permeability derived from synthetic capillary
pressure tends to be reasonably accurate for practical
engineering calculations of productivity.
9. The mD-ft computations for hydrocarbons and water
provide a solid framework for completion / hydraulic
fracturing decisions and, consequently, increasing the
utility of NMR logs in engineering applications
10. Reasonably accurate predictions of unmeasured rock
properties at wireline log scale enhance the reservoir
performance estimations via reservoir and frac simulation
at a fraction of the cost for unnecessary measurements
11. “Reservoir Description and Characterization” studies
improve and gain accuracy with the generation of
Advance Rock Properties information from NMR log
data.
Nomenclature
BVI : bulk volume irreducible
D : diffusion constant
DTW : dual wait time
F : lithology factor
φe : effective porosity
φi : partial porosity
Fs : pore shape factor
fw : fractional flow of water
γ : gyromagnetic ratio
G : gradient field
k : permeability
SPE 71703 CAPILLARY PRESSURE DATA FROM NMR LOGS AND ITS IMPLICATIONS ON FIELD ECONOMICS 9

krnw : relative permeability of the non-wetting nw (subscript) : non-wetting phase

phase Sw : wetting phase saturation
krw : relative permeability of the wetting phase Sw* : normalized wetting phase saturation
Pc : capillary pressure Siw : connate water saturation
Pd : displacement (threshold) pressure Sor : residual oil saturation
θ : contact angle σ : interfacial tension
ρ : a measure of surface's ability to cause S/N : signal to noise ratio
protons to relax Sw : wetting phase saturation
rb : pore-body radius Swir : irreducible saturation of the wetting phase
rpt : pore-throat radius T2 : rate of decay of transverse magnetization
S : surface area T2bulk : T2 of bulk fluid
C : Scaling coefficient T2cutoff : a value of T2 empirically related to
R : multiplier capillary irreducible phase
w (subscript) : wetting phase TE : inter-echo time
V : volume
10 M. ALTUNBAY, R. MARTAIN, M. ROBINSON
Acknowledgements
The authors thank Baker Atlas management for permission to
publish this work.
We also would like to thank Dr. Daniel Georgi for his
comments and contributions, Dr. Gigi Zhang for her assistance
and Ms Karen Bush for her editorial comments.
References
1. Altunbay, M., Georgi, D., Zhang, G.Q., ”Pseudo Capillary
Pressure from NMR Data”, 12th International Petroleum
Congress of Turkey, October 12-14, 1998
2. Volokitin, Y., Looyestijn, W.J., Slijkerman, W.F.J.,Hofman,
J.P., “A Practical Approach to Obtain 1st Drainage Capillary
Pressure Curves From NMR Core and Log Data”, SCA-9924
3. Zhang, G.Q., Huang, C.C., “Newpc”, A Matlab® program for
calculating C values, 1998
4. Chen, S., Arro, R., Minetto, C., Georgi, D., Liu, C., 1998,
“Methods for Computing Swi and BVI from NMR logs’,
Transactions of 39th SPWLA Annual Symposium, Keystone,
CO, May 26-29, paper HH
5. Marschall, D, Gardner, J.S., Mardon, D., Coates, G.R., “Method
for correlating NMR Relaxometry and Mercury Injection Data”,
Proceedings of the 1995 International Symposium of Society of
Core Analyst, paper 9511
6. Lowden, B.D., Porter, M.J., Powrie, L.S., 1998,”T2 Relaxation
time versus Mercury Injection Capillary Pressure: Implications
for NMR Logging and Reservoir Characterization”, SPE 50607
7. Purcell, W.R., “Capillary Pressures-Their Measurement Using
Mercury and The Calculation of Permeability Therefrom”,
”Trans. AIME (1949),186, pp.38-48
8. Personal Communications, Dr. Songhua Chen. Baker Atlas
9. Mohamad Ibrahim, M.N., Koederitz, L.F., “Two-Phase Relative
Permeability Prediction Using a Linear Regression Model”, SPE
65631, 2000 SPE Eastern Regional Meeting, Morgantown West
Virginia, 17-19 October 2000
10. Mohamad Ibrahim, M.N, “ Two-Phase Relative Permeability
Predictions Using A Linear Regression Model”, Dissertation,
UMR Library, Thesis T7668
11. Honarpour, M. Koederitz, L.F., Harvey, A.H, “Empirical
Equations for Estimating Two-phase Relative Permeability in
Consolidated Rock”, SPE 9966, December 1982
SPE 71703
12. Burdine, N.T., “Relative Permeability Calculations from Pore
Size Distribution Data”, Petroleum Transactions, AIME, Vol
198, 1953, pp 71-78
13. Huang,D.D., Honarpour M.M., Al-Hussainy, R., “An Improved
Model for Relative Permeability and Capillary Pressure
Incorporating Wettability”, SCA-9718, 1997 SCA International
Symposium, Calgary, Canada, September 7-10, 1997
14. Rajan, R.R., Delaney, P., “Reconciliation of Log and Capillary
Pressure Based Water Saturation in the Fateh Thamama
Reservoir, Dubai”, 32nd Annual Logging Symposium, June 16-
19, 1991
15. Huang, D.D., Honarpour, M.M., Al-Hussainy, R., “An
Improved Model for Relative Permeability and Capillary
Pressure Incorporating Wettability”, SCA 9718
16. Amyx, J.W., Bass, D.M., Whiting, R.L., “Petroleum Reservoir
Engineering”, McGraw-Hill Book Company, 1960, page 133-
139
17. Shaffer, L.J., Bernardo, M., “Obtaining Capillary Pressure Data
From NMR T2 Distributions”, Extended Abstract, 1996
SPWLA NMR Workshop, Section 12
18. Kenyon, W.E., Howard, J.J., Sezginer, A.,Straley, C., Matteson,
A., Horkowitz, K., and Elrich, R., 1989, Pore-size distribution
and NMR in microporous cherty sandstones, SPWLA
Thirteenth Annual Logging Symposium, June 11-14, pp. I-
24LL. 11
19. Huang, C.C., “Estimation of Rock Properties by NMR
Relaxation Methods”, MS Thesis, Rice University, Houston, Tx,
Aug 1997.
20. Norgard, J.V., Olsen D, Springer, N, Reffstrup J., “Capillary
Pressure Curves for Low Permeability Chalk Obtained by NMR
Imaging of Core Saturation Profiles”, SPE 30605, SPE Annual
Technical Conference, Dallas, Tx, USA, 22-25 October 1995,
pp:807-816
21. Personal communications, Dr. Jude Amaefule, VRMT
International Inc., Houston, Tx
22. Hassler,G.L., and Brunner, E., “Measurement of Capillary
Pressure in Small Core Samples”,AIME tech. Pub., page 1917
(1945)
23. Wells, J.D., Amaefule, J.O., “Capillary Pressure and
Permeability Relationships in Tight Gas Sands”, SPE/DOE
13879, SPE/DOE 1985 Low Permeability Gas Reservoirs,
Denver Colorado,, May 19-22, 1985