Improving the Reliability of Synthetic S-Wave Extraction Using Biot-Gassman Fluid Substitution

Sudarmaji.et al. Int. Journal of Engineering Research and Applications www.ijera.com
ISSN: 2248-9622, Vol. 6, Issue 4, (Part - 4) April 2016, pp.52-56
www.ijera.com 52|P a g e
Improving the Reliability of Synthetic S-Wave Extraction Using
Biot-Gassman Fluid Substitution
Sudarmaji1
, Sismanto2
, Waluyo3
And Bambang Soedijono4
1
phd Student of Department Of Physics, Gadjah Mada University, Indonesia
2,3
Department of Physics, Gadjah Mada University, Indonesia
4
Department of Mathematics, Gadjah Mada University, Indonesia
ABSTRACT
Usage of Castagna’s relation by mean the P-wave log directly in estimation of the S-wave log gives a large error
to the original S-wave log so that the result is not reliable for further analysis. This paper offers new method for
S-wave log estimation based Castagna’s relation which is combined with fluid replacement modeling method
based on Biot-Gassman substitution and the use of petro physical data as input. The S-wave log result of the
offered estimation method has a small error to the original S-wave log so that more reliable and accurate for
further analysis. The offered method could be used for S-wave log estimation in various litology such as sand,
limestone, dolomite and shale. The S-wave log result of the offered estimation method has successfully used for
cross plot analysis of Vp/Vs ratio as function of acoustic impedance and Gamma Ray for delineation of
sandstone bearing hydrocarbon from two different field.
Keywords - Biot-Gassman substitution, S-wave log, P-wave log, Vp/Vs ratio
I. INTRODUCTION
Shear wave (S-wave) log data is always
used in the determination of the elastic parameters of
reservoir, analysis of rock physics, inversion and
lambda-mu-rho [1]. However, most of the well data
is not completed by the shear wave log data. Various
of S-wave log estimation has been proposed [1, 2].
The famous one is Castagna’s relation [3]. The use
of Castagna’s relation by mean the P-wave log
directly in estimation of the S-wave log gives a large
error if compared to the original S-wave log because
castagna’s relation is a good empirical tool for the
wet sands and shale only, so the result is not reliable
for further analysis. This paper offers new method
for S-wave log estimation based on Castagna’s
relation which is combined with Fluid replacement
modeling method based on Biot-Gassman
substitution and the use of petro physical data as
input. The weakness of direct S-wave estimation by
Castagna’s relation is compensated by applying
Castagna’s relation in the wet condition of reservoir
only. The S-wave log resulted from the offered
estimation method has a small error to the original
S-wave log so that more reliable for further analysis.
II. TEORY
Castagna’s relationship between P-wave and S-
wave velocity
Simple empirical relationship between P-
wave and S-wave velocity for unconsolidated and
partially consolidated sand has been derived by
Castagno [3] in the form of second order
polynomial fit as follows
𝑉𝑠 = 𝑎𝑉𝑝
2
+ 𝑏𝑉𝑝 + 𝑐 (1)
where 𝑎, 𝑏 and 𝑐 are polynomial coefficient. The
relationship is lithology dependent and different
coefficient are proposed to increase the accuracy in
km/s as [1].
Sand:
𝑉𝑠 = +0.804𝑉𝑝 + 0.856 (2)
Limestone:
𝑉𝑠 = −0.055𝑉𝑝
2
+ 1.017𝑉𝑝 − 1.030 (3)
Dolomite:
𝑉𝑠 = +0.583𝑉𝑝 − 0.078 (4)
Shale:
𝑉𝑠 = +0.770𝑉𝑝 − 0.867 (5)
Unfortunately these equations are a good empirical
tool for wet sand and shale only.
The Biot-Gassmann equation
In porous and saturated rocks, basic
equations for P and S-wave velocity of saturated
rocks are.
𝑉𝑝_𝑠𝑎𝑡 =
𝐾 𝑠𝑎𝑡 +
4
3
𝜇 𝑠𝑎𝑡
𝜌 𝑠𝑎𝑡
(6)
𝑉𝑠_𝑠𝑎𝑡 =
𝜇 𝑠𝑎𝑡
𝜌 𝑠𝑎𝑡
(7)
where 𝜌𝑠𝑎𝑡 is saturated density, 𝜇 𝑠𝑎𝑡 is saturated
shear modulus and 𝐾𝑠𝑎𝑡 is the bulk modulus of
saturated rocks. The problem was first addressed by
Biot and then Gassmann using apparently different
RESEARCH ARTICLE OPEN ACCESS

approaches but, as shown by Krief [4,5], these two
approaches lead to the same results.
In equations (6) and (7), the saturated
density can either be measured in-situ or computed
from the equation h
𝜌𝑠𝑎𝑡 = 𝜌 𝑚 1 − 𝜙 + 𝜌 𝑤 𝑆 𝑤 𝜙 + 𝜌h𝑐 1 − 𝑆 𝑤 𝜙 (8)
where 𝜌 with the subscripts 𝑚, 𝑤, and h𝑐 indicate
matrix, water, and hydrocarbon, 𝑆 is the fraction of
saturation of each fluid component and 𝜙 is porosity
[6].
Once we have the measured or computed
saturated density value, the value of the shear
modulus in equations (6) and (7) can be computed
from the S-wave velocity using equation (8) to give:
𝜇 𝑠𝑎𝑡 = 𝜌𝑠𝑎𝑡 𝑉𝑠_𝑠𝑎𝑡
2
(9)
The value of the saturated bulk modulus
can then be found from equation (3), giving:
𝐾𝑠𝑎𝑡 = 𝜌𝑠𝑎𝑡 𝑉𝑝 𝑠𝑎𝑡
2
−
4
3
𝜇 𝑠𝑎𝑡 (10)
In order to perform fluid replacement
modeling using Biot-Gassmann equation, it is
assumed that the shear modulus is independent of
fluid content (but not porosity) or
𝜇 𝑠𝑎𝑡 = 𝜇 𝑑𝑟𝑦 (11)
where 𝜇 𝑑𝑟𝑦 is the shear modulus of the dry rock,
which is the rock for which the pore fluids have been
fully evacuated.
The Biot-Gassmann equation can be written:
𝐾𝑠𝑎𝑡 = 𝐾𝑑𝑟𝑦 +
1−
𝐾 𝑑𝑟𝑦
𝐾 𝑚
2
𝜙
𝐾 𝑓𝑙
+
1− 𝜙
𝐾 𝑚
+
𝐾 𝑑𝑟𝑦
𝐾 𝑚
2
= 𝐾𝑑𝑟𝑦 +
𝛽2
𝜙
𝐾 𝑓𝑙
+
𝛽− 𝜙
𝐾 𝑚
and (12)
𝛽 = 1 −
𝐾 𝑑𝑟𝑦
𝐾 𝑚
(13)
where 𝐾 𝑚 is bulk modulus of matrik, 𝐾𝑓𝑙 is bulk
modulus of fluid, 𝐾𝑑𝑟𝑦 is bulk modulus of dry, 𝐾𝑠𝑎𝑡
is modulus bulk of saturated rock, 𝜙 is porosity and
𝛽 is called the Biot Coefficient, the ratio of the
volume change in the fluid to the volume change in
the formation, when hydraulic pressure is constant.
If 𝛽 = 0 then 𝐾𝑑𝑟𝑦 = 𝐾 𝑚 and 𝐾𝑑𝑟𝑦 = 𝐾𝑠𝑎𝑡 or purely
elastic condition. Mavko rearranged the Biot-
Gassmann equation to the form [7]:
𝐾 𝑠𝑎𝑡
𝐾 𝑚 − 𝐾 𝑠𝑎𝑡
=
𝐾 𝑑𝑟𝑦
𝐾 𝑚 − 𝐾 𝑑𝑟𝑦
+
𝐾 𝑓𝑙
𝜙(𝐾 𝑚 − 𝐾 𝑓𝑙 )
(14)
The bulk modulus of fluid 𝐾𝑓𝑙 is normally calculated
using the Reuss average given by
1
𝐾 𝑓𝑙
=
𝑆 𝑤
𝐾 𝑤
+
𝑆ℎ 𝑐
𝐾ℎ 𝑐
(15)
The Reuss average is used for uniform
distribution of the fluids, for patchy saturation, the
linear Voigt average could be used.
There are many alternate forms of the Biot-
Gassmann equations besides equation (14), but this
form to be the easiest to work with when performing
fluid replacement modeling.
It assumed that the saturated, matrix
(mineral) and fluid bulk modulus are known, as well
as the porosity, from laboratory or petro physics
analysis then the only unknown in equation (14) is
the dry rock bulk modulus. This can then be
computed by re-arranging equation (14) to give
𝐾𝑑𝑟𝑦 = 𝐾 𝑚
𝑥
1+𝑥
(16)
Once 𝐾𝑑𝑟𝑦 is able to be computed, then the
fluid value can be changed by using new saturations
in equation (15) and then re-compute 𝐾𝑠𝑎𝑡 and thus
𝜌𝑠𝑎𝑡 and 𝑉𝑝_𝑠𝑎𝑡 . For the case of a fluid change only,
the shear modulus and dry rock bulk modulus will
not change. However, if the porosity in equation (14)
changes, both the dry rock and shear modulus will
change. To change these parameters, there is no
clear consensus as to which method to use [6]. .
The P wave modulus
It is needed the elastic parameter of P-wave
modulus for fluid replacement modeling. The P-
wave modulus is defined as
𝑀 = 𝑉𝑝_𝑠𝑎𝑡
2
𝜌𝑠𝑎𝑡 (17)
where 𝑀 is P-wave modulus, 𝑉𝑝_𝑠𝑎𝑡 is P-wave
velocity of saturated, and 𝜌𝑠𝑎𝑡 is saturated density.
If modulus bulk of matrix and modulus shear of
matrix has been known from laboratory or petro
physics analysis then P-wave matrix modulus
defined as:
𝑀 𝑚 = 𝐾 𝑚 +
4
3
𝜇 𝑚 (18)
where 𝐾 𝑚 is bulk modulus of matrix, and 𝜇 𝑚 is
shear modulus of matrix.
III. METHOD
The workflow of S-wave velocity
estimation using fluid replacement modeling could
be seen in Figure 1. The input for this process is
density log, P-wave velocity log, porosity log, water
saturation log and bulk modulus of matrix, density
of matrix, bulk modulus and density of fluid. The
porosity log, water saturation log and bulk modulus
of matrix, density of matrix, bulk modulus and
density of fluid could resulted from petro physics
analysis. The S-wave estimation process is done
after the rocks are really fully saturated by water (S-
wave in wet condition) so the weakness of
Castagna’s relation can be compensated. Then the
estimated S-wave is calculated from S-wave in wet
condition. By using this S-wave, modulus bulk dry
of the rock could be calculated and could be used to
calculate the true modulus bulk of saturated
reservoir rock. Finally the true saturated density, P-
wave and S-wave could be calculated as well.

Figure 1. Workflow of S-wave velocity estimation
using fluid replacement modeling
IV. EXPERIMENT AND DISCUSSION
S-wave log estimation method is tested to
estimate S-wave log of sands bearing gas from
#PMK-2 well, Mawar field, Kutai Basin, Indonesia
which has original S-wave log. The gas sand
reservoir as objective of tested zone is in the depth
626 ft to 700 ft. The estimated S-wave log is
compared to the original S-wave log and the
estimated S-wave log from directly Castagna’s
relation. The estimated S-wave log using fluid
replacement modeling is more accurate than the
estimated S-wave log from direct Castagna’s relation
(Figure.2). The fourth log in the Figure 2 are the
overlay of original S-wave log (red color), estimated
S-wave log using direct Castagna’s relation
(magenta color) and estimated S-wave log using
fluid replacement modeling (blue color).
The estimated S-wave log using fluid
replacement modeling more coincides with the
original -wave log than the estimated S-wave log
using direct Castagna’s relation. Therefore, the
estimated S-wave log using fluid replacement
modeling more accurate and reliable for further
analysis than the estimated S-wave log using direct
Castagna’s relation. The benefits of the offered
estimated S-wave is that S-wave estimation process
done after the rocks is really fully saturated by water
(brine), so the weakness of Castagna’s relation can
be compensated.
Cross plot of Vp/Vs ratio versus acoustic
Impedance as function Gamma Ray of #PMK-2 well
using the S-wave velocity which is estimated using
fluid replacement modeling method could be seen in
Figure 3. Two characteristic sand reservoirs which
are bearing gas could be clustered in the cross plot as
well. The cluster of two gas sand reservoirs is
separated from the cluster of dry sand, shale and
coal. This cross plot is the basis for simultaneous
inversion and Lambda-mu-rho analysis. The
conventional acoustic impedance inversion will fail
if applied to this sand reservoir because the acoustic
impedance value of the gas sand is superimposed
with dry sand and shale.
Figure 2. The comparison of original S-wave
velocity, S-wave velocity estimated using
fluid replacement modeling method and
direct Castagna’s relation from #PMK-2
well, Mawar field, Kutai Basin,
Indonesia.
Figure 3. Cross plot of Vp/Vs ratio versus acoustic
Impedance as function Gamma Ray using
the estimated S-wave velocity from from
#PMK-2 well, Mawar field, Kutai Basin,
Indonesia.
Wireline Log DataPetrophysic Data
Create Initial Vs
Using Castagna’s equation
Determine target sand
boundary
Calculate ρ for 100%
Brine Saturation
Calculate Vp_sat Modulus
Calculate Matrix Modulus of Vp_sat
Adjust Vp_sat Modulus
to 100% brine saturation
Calculate Vp_wet
Calculate Vs_wet from Vp_wet
Calculate Vs from Vs_wet
Calculate K_sat and Mu_sat
Calculate K_dry
Calculate K_sat for new fluid
Get density ,Vp, Vs for new fluid
Correct Vs Value inside Target
Reservoir1
Reservoir2
Sand
Shale
Coal
Reservoir1 :
Vp/Vs: 1.35 – 1.6
Ip : 21500 – 23750((ft/s)*(g/cc))
Reservoir2:
Vp/Vs: 1.4 – 1.55
Ip : 24000 – 27000((ft/s)*(g/cc))
P-ImpedancevsVpVs-waveCrossplot

The S-wave log estimation method using
fluid replacement modeling is applied for S-wave
estimation of oil saturated sandstone in Middle
Missisauga formation from #L-30 well, Penobscot
field, Nova Scotia Basin, Canada. Figure 4. shows
Log data of #L-30 well from Middle Missisauga
formation [8]. The blocky interval is the oil saturated
sandstone with high acoustic impedance. Cross plot
of Vp/Vs ratio versus acoustic impedance as
function Gamma Ray of #L-30 well using the S-
wave velocity estimated using fluid replacement
modeling method is displayed in figure 5. The
cluster of high acoustic impedance sandstone which
saturated with oil can be seen clearly. The low
value of Vp/Vs ratio and high value of acoustic
impedance from sandstone bearing oil could be
clustered which are separated from wet sand and
shale as well, so it is potential to delineated using
simultaneous inversion or lambda-mu-rho analysis.
Gambar 4. Log data of #L-30 well containing oil
saturated sandstone from Middle
Missisauga formation
Figure 5. Cross plot of Vp/Vs ratio versus acoustic
Impedance as function Gamma Ray using
the estimated S-wave velocity from of #L-
30, Penobscot field, Nova Scotia Basin,
Canada.
V. CONCLUSION
We have proposed new method for
estimation of S-wave log based Castagna’s relation
which is combined with Fluid replacement modeling
method based on Biot-Gassman substitution and the
use of petro physical data as the input. The S-wave
log result of the offered estimation method has a
small error to the original S-wave log so that it’s
more accurate and reliable for further analysis. The
offered method could be used for S-wave log
estimation in various litology such as sand,
limestone, dolomite and shale. The S-wave log result
of the offered estimation method has successfully
used for cross plot analysis of Vp/Vs ratio as
function of acoustic impedance and Gamma Ray for
delineation of sandstone bearing hydrocarbon from
two different field.
ACKNOWLEDGEMENTS
The author thanks to Geophysics
Laboratory of Gadjah Mada University, Indonesia
for sponsorship and supporting facilitates. The
author also thanks to Nova Scotia Department of
Energy, Canada for its public set data.
.
REFERENCES
[1]. Veeken, C.H.P., Seismic Stratighraphy,
Basin analysis and Reservoir
Characterization, Amsterdam: Elsevier,
2007
[2]. Mavko, G., Chan, C. and Mukerji, T. 1995.
Fluid substitution: estimating changes in
Vp without knowing Vs, Geophysics 60,
1750–1755
[3]. Castagna, J.P., Batzle, M.L. and Eastwood,
R.L., 1985, Relationship between
Sand SandsaturatedoilShale
P-IMPEDANCE vsVp/Vs (GR)

compressional-wave and shear-wave
velocities in clastic silicate rocks,
Geophysics, 50, 571-581
[4]. Biot, M.A., 1941, General theory of three
dimentional consolidation, Journal of
Applied Physics 12, 155-164.
[5]. Krief, M., J. Garat, J. Stellingwerff and J.
Ventre, 1990, A petrophysical
interpretation using velocities of P and
Swave (full waveform sonic), The Log
Analyst, Vol. 31, p. 355–369.
[6]. Russell, B., and Smith, T., 2007, The
relationship between dry rock bulk modulus
and porosity–An empirical study,
CREWES Research Report, Vol. 19
[7]. Mavko, G., Mukerji, T., and Dvorkin, J.,
The Rock Physics Handbook: Tools for
Seismic Analysis of Porous Media,
Cambridge University Press, 2004.
[8]. Clack, W.J.F and Crane, J.D.T, 1992,
Penobscot Prospect: Geological evaluation
and oil reserve estimates, Report for Nova
Scotia Resources (Ventures) Ltd.
.

Improving the Reliability of Synthetic S-Wave Extraction Using Biot-Gassman Fluid Substitution

More Related Content

What's hot (20)

Similar to Improving the Reliability of Synthetic S-Wave Extraction Using Biot-Gassman Fluid Substitution (20)

Recently uploaded (20)

Improving the Reliability of Synthetic S-Wave Extraction Using Biot-Gassman Fluid Substitution