Sie sind auf Seite 1von 6

SCA2006-56 1/6

EVALUATION OF WATER SATURATION FROM


LABORATORY TO LOGS
M. S. Efnik1, M. Dernaika2 and M.Z. Kalam1
Abu Dhabi Company for Onshore Oil Operations1 and ResLab L.L.C., Abu Dhabi2
This paper was prepared for presentation at the International Symposium of the
Society of Core Analysts held in Trondheim, Norway 12-16 September, 2006

ABSTRACT
An extensive laboratory study to determine initial water saturation as well as remaining
oil in water flooded regions has been initiated using porous plate measurements at
reservoir temperature and reservoir overburden pressure conditions. We present the
results of the primary drainage experiments which capture the capillary pressure (Pc) and
the electrical resistivity (Sw-RI) changes all the way down to Swi. The impact of the
drainage laboratory results for the well and for the reservoir is also considered.

The initial water saturation was estimated using the RI-Sw curves in drainage. Very low
saturation was reached covering the range of saturation of interest in the field. The
saturation exponents ‘n’ appear to be around 2 for the reservoir rock types (RRT)
containing most of the STOOIP while the tighter RRT’s show lower values. Capillary
pressure data when grouped according to lithofacies as J-function plots confirm the
distinct RRT characteristics.

The value of the well defined laboratory data and its integration to logs is presented.
Despite intermediate to oil wet conditions, ‘n’ values are lower than the default value of 2
and much lower than that expected for oil wet conditions. The laboratory measurements
reduced the uncertainties in the oil in place estimations and allowed a realistic evaluation
of the water flooding performance.

INTRODUCTION
Capillary pressure curves (Pc) and electrical resistivity measurements (Sw-RI) are two
most important factors in formation evaluation. Representative capillary pressure curves
can give reliable information for the evaluation of the reservoir fluid distribution. The
importance of the capillary pressure parameter comes from the fact that connate water in
the reservoir occurs through the mechanism of immiscible fluid displacement which
involves capillary action. The electrical resistivity of a fluid-saturated rock is its ability to
impede the flow of electric current through that rock. Sw-RI data allow laboratory
determination of the saturation exponent ‘n’ which in turn is used to quantify fluid
saturations from resistivity logs across the zones of interest in a hydrocarbon bearing
formation.
SCA2006-56 2/6

EXPERIMENTAL PROCEDURES & DISCUSSIONS OF RESULTS


Sample Selection And Preparation
Forty four reservoir core plug samples were screened to obtain representative rock types
present in a carbonate reservoir zone from a producing field in the Middle East. The core
plugs were 1.5” in diameter and the sampling strategy for SCAL was based on X-ray CT
scanning, Thin Section description, mercury injection measurements and poroperm
analysis. The reservoir core samples examined comprised 8 RRT’s with porosity ranges
from 10-33% and permeability ranges from 0.01-6000 mD.

Figure 1 presents the porosity permeability trends for the selected SCAL plugs (20
plugs). This shows the distinct RRT classification in the groupings of samples selected.
NMR porosity had also been acquired on the samples and showed good agreement with
the Helium porosity results (figure 2). The selected core samples underwent two-stage
flush cleaning cycles with various solvent combinations in order to render the samples
more water wet. The cleaning procedure was an optimised technique that were used in
previous studies on similar rocks and verified through restored wettability experiments.

Formation Resistivity Factor


Figures 3 and 4 show the evaluated cementation exponents ‘m’ at ambient and reservoir
conditions, respectively. These confirm ‘m’ to increase slightly from 1.95 to 1.98 upon
applying a reservoir temperature of 121ºC, at the same reservoir overburden pressure.
Figure 5 shows a decrease in FF with increasing permeability below approximate value
of 10 mD. This electric behavior is controlled by changes in pore geometry which can
also be linked to changes in permeability. However, the fairly unchanging FF with
permeability for the rocks higher than 10mD should be due to similarities in the pore
geometry for those rocks despite the large changes in permeability.

Capillary Pressure And Resistivity Index


Combined capillary pressure curves and resistivity index plots were generated by the
ceramic porous plate method [Longeron and Wilson et al, refs. 1 &, 2]. All measurements
were performed under net overburden pressure and reservoir temperature using stock
tank oil as the displacing fluid. The core plugs were de-saturated by increasing capillary
pressure in steps and recording resistivity and expelled water on a daily basis. Figure 6
shows the impact of porosity on the evaluated Swi at the end of primary drainage cycle.

Figures 7 and 8 represent typical time steps applied during the equilibrium measurements
for the drainage cycle (brine de-saturation), and confirm the long times required for good
Pc and Sw-RI data. A full drainage cycle can typically take up to 8 months for many of
these carbonate samples. Figures 9 and 10 typify the RI-Sw and Pc data, respectively
from such measurements. In figure 9 it is seen that both transient and equilibrium data
follow the same trend. Figure 10 also shows a comparison with the water-oil Pc derived
from mercury intrusion on the trim of the same sample at no confining stress; the
differences are evident from the errors in assumed conversions and the physics of the
SCA2006-56 3/6

wetting processes. Such plots, however, enhance confidence in the measured water-oil Pc
data from the porous plate.

Figure 11 show the J-function plot of the analyzed samples. It is evident that the
lithofacies are impacting the results with RRT 1 and RRT 2 containing grain stones, vugs
and rudists conforming to one set while RRT 3 – RRT7 containing mostly pack stones
and relatively more homogenous conforming to another set. Similarly, when the RI-Sw
plots are examined (as in figure 12), ‘n’ is shown to vary from 1.99 to 1.61. The range in
the data is reminiscent of previously examined data from another reservoir zone of the
same field [Fleury et al, ref 3] although the non-linear behavior observed earlier in some
of the samples were totally absent in the current study. Further, the data set is consistent
and repeatable in showing ‘n’ approaching 2 in most of the prolific rock types (RRT 1 –
RRT 5) while the less permeable RRT (6-8) samples show ‘n’ approximately around 1.6.
The significance of this analysis is demonstrated by examining logs in the same well, and
in other wells.

Principal Component Analysis (PCA)


A Principle Component Analysis has been performed on resistivity data acquired from
four different fields (A, B, C and D) in the Middle East by three different techniques:
Porous Plate, FRIM and CI. The data in this paper comes from field D and the work has
been presented elsewhere [Kalam et al, ref 4]. The PCA study confirms the statistical
acceptability of the measured samples, despite a wide variation in RRT and poroperms.

Log Calibrations
The new SCAL results are used to calibrate the log response in two wells, one with oil
base mud to capture the Swi as accurately as possible (figure 13) and another one from
the study well where the SCAL plugs are taken (figure 14). The new results are compared
with the old water saturation calculation using previous data where saturation and
cementation exponents were derived mathematically using linear regression analysis. The
old parameters (m=2.15 and n=1.78) are compared with the new parameters using two
ranges: an average range where m=1.95 and n=1.99 representing the two dominant rock
types (RRT4 &RRT5) containing most of STOOIP and another where m=1.95 and
n=1.60 for the poor quality rock types. New SCAL data matches closely to that derived
from oil base mud (filled dots), and hence more dependable.

CONCLUSIONS AND RECOMMENDATIONS


Porous plate at reservoir temperature and reservoir overburden pressure is indeed a
dependable tool for water-oil Pc and RI-Sw measurements during primary drainage of
carbonate reservoir cores. Interesting observations of RRT dependency on RI-Sw
behaviour is confirmed for first time, and shows that ’n’ can be as low as 1.60 in tight
carbonate samples. The impact on the calibrated logs is significant as a lower ‘n’ value
means a higher amount of remaining oil and STOOIP. Non-linear RI-Sw behaviour is not
observed despite presence of local heterogeneity such as vugs and rudists. NMR derived
porosity is an excellent tool to capture representative carbonate porosities, and needs to
SCA2006-56 4/6

be considered as an essential SCAL measurement tool where core matrix preservation is


required prior to various displacement tests.
ACKNOWLEDGEMENTS
The authors acknowledge ADNOC and ADCO Management for permission to share the
contents of this on-going study.
10000.0 0.40
P orosit y
Ambient 0.35
1000.0 Equal line

NMR porosity (fraction)


Gas permeability (mD)

P oroperm
0.30 Linear (P orosit y)
100.0
0.25
10.0 0.20
0.15
1.0
0.10 y = 0.9587x + 0.0104
0.1 0.05 R2 = 0.9951

0.0 0.00
0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40

Helium porosit y (fract ion) Helium porosit y (fract ion)

Figure1. Poroperm Distribution And RRT Figure 2. Porosity - NMR Vs Helium


1000 1000
Formation Factor at OB Stress and

m = 1.98
Formation Factor at OB Stress and

m = 1.95
a=1 a=1
100 100
Ambient T

Ambient T

10 10

1 1
0.01 0.10 1.00 0.01 0.10 1.00
Porosity (fraction) Porosity (fraction)

Figure 3. FF Vs Porosity At OB, Amb T Figure 4. FF Vs Porosity at OB, RT


1000 0.20
Form ation Factor at OB Stress and

Swi (fraction), (Pc=101 psi)

0.15
100
Am bient T

0.10

10
0.05
y = -0.502x + 0.205
R2 = 0.660
1
0.00
0.0 0.1 1.0 10.0 100.0 1000.0 10000.
0.00 0.10 0.20 0.30 0.40
Gas permeability at amb (md) 0 Porosity at OB (md)

Figure 5. FF Vs Permeability At Ambient T Figure 6. Swi Vs Porosity


SCA2006-56 5/6

1 .0 0 1000
P c=3 .6 2 5 p si P c=3 .6 2 5 p si
0 .8 0 P c=1 0 .1 5 0 p si P c=1 0 .1 5 0 p si
Water saturation, Sw

Resistivity Index, RI
P c=2 5 .0 1 3 p si P c=2 5 .0 1 3 p si
100
0 .6 0 P c=5 8 .0 0 0 p si P c=5 8 .0 0 0 p si
P c=1 0 1 .5 0 p si P c=1 0 1 .5 0 p si
0 .4 0
10

0 .2 0

0 .0 0 1
0 25 50 75 100 125 150 175 200 225 250 0 25 50 75 100 125 150 175 200 225 250
T im e, day s T im e, day s

Figure 7. Sw Vs Time-plug B Figure 8. RI Vs Time-plug B

1000 120
T ran sien t dat a M ercu ry cu rv e

Capillary pressure, Pc (psi)


E quilibrium dat a Oil-Brin e Pc
Resistivity index, RI

100 80

10 40
Hg curve was obtained
with no confining stress
0
1
0.00 0.20 0.40 0.60 0.80 1.00
0 .0 1 0 .1 0 1 .0 0 W at er sat urat io n , Sw (frac.)
W at er sat urat io n , Sw (frac.)

Figure 9. Transient RI - plug B Figure 10. Capillary Pressure Curve-Plug B

20.00 1000
-1.38
RRT 1and 2 Regression RRT 1-5
18.00 y = 0.78x
2 Regression RRT 6-7
16.00 R = 0.66 RRT 3-7
14.00 The curve is shifted to the
Resistivity index, RI

100
12.00 right for RRT 1-2 due to
J(Sw)

10.00 differences in pore geometry.


8.00
6.00 -1.23 10
y = 0.16x
4.00 2
R = 0.73
2.00
0.00
1
0.00 0.20 0.40 0.60 0.80 1.00
0.01 0.10 1.00
Water saturation, Sw (frac.) Wat er sat urat ion, Sw (frac.)

Figure 11. Pc J-function plot Figure 12. RI versus Sw for all samples
SCA2006-56 6/6

Figure 13. Comparison of Calculated Figure 14. Study Well


Sw from Resistivity Log Using Measured
Resistivity Data With Sw from Cores Taken
With Oil Base Mud

REFERENCES
D.G. Longeron, M.J. Arquad, and L. Bouvier, “Resistivity Index And Capillary Pressure
Measurements Under Reservoir Conditions”. SPE 19589, San Antonio 1989.
O.B. Wilson, B.G. Tjetland and A. Skauge, ”Porous Plate Influence On Effective
Drainage Rates In Capillary Pressure Experiments” SCA 2001, Edinburgh.
M Fleury, M. S. Efnik and M .Z. Kalam, “Evaluation Of Water saturation From
Resistivity In A Carbonate Field”, SCA 2004 – 22, Abu Dhabi.
M Z Kalam et al, ”Importance Of Porous Plate Measurements On Carbonates At Pseudo
Reservoir Conditions”, (To Be Presented) At SCA 2006 – A 38, Trondheim.