SCA2004-55 1/6

VALIDATING LABORATORY MEASURED ARCHIE

SATURATION EXPONENTS USING NON-RESISTIVITY
BASED METHODS

D. Boyd and K. Al Nayadi, ZADCO
A. Al Hendi, O. Farisi and M. I. Al-Marzouqi, ADMA-OPCO
O.B. Wilson, ResLab, and G. Sinclair, El ResLab
M. Fleury, IFP

This paper was prepared for presentation at the International Symposium of the
Society of Core Analysts held in Abu Dhabi, UAE, 5-9 October 2004

ABSTRACT
Archies second law has considerable impact on calculated water saturation, especially in
the low to intermediate water saturation range. Despite this fact, Resistivity Index (RI)
measurements for our limestone reservoirs were sparse and performed at non-reservoir
representative conditions. Our lack of knowledge of the proper saturation exponent to
apply forced us to use the industry standard value of two. To rectify the situation,
reservoir condition RI experiments were carried out using different techniques on
separate reservoir units. The first experiment employed the traditional porous plate with
continuous scanning, live crude and hot toluene cleaned core plugs. The second
experiment used fresh state core, dead crude oil, and the Fast Resistivity Index
Measurement (FRIM) technique. The third technique employed as received single
point RI measurements on plugs drilled from an oil base core. All three methods were
performed at reservoir temperature and simulated overburden pressure. To give us
confidence in the correctness of the saturation exponents measured, we compared water
saturation obtained with non-resistivity based methods to water saturation calculated with
our new Archie parameters.

Introduction
Water saturations calculated from open hole resistivity measurements is a primary input
to oil-in-place calculations for Abu Dhabi reservoirs. Additionally the water saturation
profiles in cored wells are compared to Hg capillary pressure derived water saturations to
validate rock typing approaches in our 3D geological models. The assumption is, the
resistivity derived water saturations are correct. Whether this is true or not, partially
depends on the accuracy of the Archie cementation and saturation exponents.
Cementation exponent measurements are inexpensive, take minimal time, and the
laboratory techniques well established. Conversely, saturation exponent measurements
are very expensive, time consuming, and vulnerable to many factors.

Porous plate desaturation is the traditional technique for saturation exponent
measurement. It normally requires several months to complete. To shorten the time
involved, the core analysis industry has developed alternative techniques such as
SCA2004-55 2/6

Continuous Injection (CI)
[1]
, centrifuge
[2]
, Fast Resistivity Index Measurement (FRIM)
[3]
,
steady state injection
[4]
, evaporation
[5]
, and as received (native state) single point
[6]
. In
the technical literature these experiments are conducted under a variety of conditions:
Native or Restored State Core
Air - Brine, Oil Brine
Lab Oil or Reservoir Crude
Dead Crude or Live Crude
Ambient or Elevated Temperature
Ambient or Overburden Condition
Uniform or Non-uniform Saturation Distribution
Reservoir versus Minimal Pore Pressure

Ideally the technique selected is the one with the least cost which delivers the correct
answer. Reservoir condition live-oil displacing brine porous plate is the accepted standard
other techniques are compared to, since it mimics closely the conditions under which the
hydrocarbon was originally emplaced. A major issue is whether to use native state or
restored state plugs. Native state assumes the reservoir wettability of the core can be
retained during coring, at surface through proper preservation techniques, and after soft
cleaning using mild solvents. Restored state attempts to return the sample to a water
wet condition assumed to exist before oil accumulated in the reservoir by using
aggressive solvents. The sample is then returned to the reservoir wetting state by
saturating and ageing with formation brine and then reservoir crude. To compare
techniques and the effect of conditions, ideally tests are run on sister samples or even the
same sample twice. In practice, this evaluation process or pre-study is not normally
performed due to the expense and time involved. Even when full reservoir condition
porous plate tests are carried out, there is a lingering doubt whether a reservoir correct
saturation exponent was obtained.

Recognizing our sparse saturation exponent data sets (4 and 10 values respectively for
two reservoirs) conducted at non-reservoir representative conditions (air-brine @
ambient) were a major uncertainty in our open hole log calculated water saturations, a
program of data gathering was initiated two years ago. RI measurement techniques were:
(1) traditional porous plate with live crude, (2) FRIM using dead crude, and (3) as
received plugs from an oil base core. All measurements took place under simulated
overburden stress and at reservoir temperature. To verify the reservoir correctness of
the answers delivered, we compared open hole log saturations calculated with the new
Archie parameters saturation measurements to saturations obtained independent of
resistivity measurements.

Reservoir Number One
This reservoir is a clean limestone (packstone/grainstone) with high porosity and
permeability. Pore systems are intergranular between the grains, with small vugs, and a
significant microporosity component within the grains. Plugs were drilled from a core
SCA2004-55 3/6

obtained with a bland water base mud and following exposure to the atmosphere for six
months. Hot toluene and methanol Soxhlet cleaning were expected to return the rock to a
water wet condition. After cleaning, the wettability of the three samples was not checked.

The full reservoir condition (live crude), porous plate, restored state, oil displacing brine
followed by brine displacing oil experiment (3 samples concurrently) took fifteen months
to complete. Gamma scanning delivered the saturation at regular intervals and allowed us
to check for uniform saturation. The electrode configuration is displayed in Figure 1.
Multiple methods of saturation were used for comparison to the resistivity calculated one:
(1) Coring with a deuterium traced mud system followed by Dean Stark analysis, (2)
Thermal decay log interpretation with rock, oil and formation water capture cross section
determined through elemental analysis, (3) Oil-brine capillary pressure, and (4) Hg-Air
capillary pressure converted to an oil-brine system.

Results
Our three plugs delivered linear oil-brine resistivity index saturation exponents from 2.39
to 2.42. The brine-oil displacement saturation exponents are nearly identical. The
composite (8 plugs) cementation exponent measured at overburden and temperature was
2.28. Archie water saturations calculated using the new m and n parameters were
significantly higher than those delivered by the other methods (Figure 2a). Back
calculating the saturation exponent using the other saturation measurements as the target,
delivered a value of 1.65 using our average m of 2.1 obtained in previous studies
(Figure 2b), and 1.45 with our new m of 2.28. Alerting the laboratory to the water
saturation discrepancies prompted a review of the electrical measurements. Quality
checks by the lab uncovered calibration and calculation errors resulting in elevated
cementation and saturation exponents. Revised cementation and saturation exponents are
2.05-2.13 and 1.71-1.83 respectively.

Reservoir Number Two
This reservoir is a clean limestone with high porosity and low to moderate permeability.
Pore systems are residual intergranular with small vugs and a significant microporosity
component within the grains. Plugs were drilled from a core obtained with a synthetic oil
mud designed to contain no water and advertised by the manufacturer to not alter
wettability. Core plugs drilled with mineral oil at the rig were preserved with wax. Upon
reception at the lab two days later, thirty-six plugs resistivity were measured (two
electrode configuration) at reservoir temperature and overburden pressure prior to Dean
Stark (DS) for as received RI. Whole core preserved in wax were sent to a second, out-
of-country laboratory for FRIM analysis. After CT scanning the whole core, four plugs
were drilled and cleaned with mild solvents (ethyl alcohol followed by isopropyl
alcohol). Assumption is the oil base mud and soft cleaning did not alter wettability,
though we have no proof of this. After saturating with simulated formation brine, each
sample was desaturated with dead crude oil at reservoir temperature and overburden
pressure using the FRIM technique (Figure 3).

SCA2004-55 4/6

Results
FIRM analysis took 10 to 20 days per sample to complete. With the exception of one
sample, the oil-brine RI versus SW measurements are linear on log-log scales. The four
saturation exponents delivered were 1.9, 2.25-3.0, 1.9, and 1.75 for different reservoir
layers. Single point saturation exponents ranged from 1.58 to 2.25. For the 22 to 50%
saturation range, Archie calculated SW (one foot resolution induction for RT) compared
well to Dean Stark SW when as received saturation exponents are used. Below 22%
SW, saturations are overestimated. FIRM saturation exponents match well the oil-brine
Pc SWi for two depth levels and poorly for the remaining two. The picture changes when
inversion processed RT is used to correct for the 48 degree inclination (bedding is flat)
and bed thicknesses of 10 to 35 feet. FRIM delivers a better match with SWi than the
single point RI method. Figure 4 compares water saturations obtained with Archie, DS
and Pc.

Conclusions
1. Saturation benchmarks using non-resistivity based methods are an important
quality check of RI and FRF measurements used for Archie calculated SW.
2. Single point, oil base core RI measurements permit the gathering of a large
number of saturation exponents quickly at Swi.
3. FRIM produces the full RI curve at a fraction of the time required for the
traditional porous plate.

References
1. Zeelenberg, H. and Schipper, B., Developments in I-SW Measurements, paper
SCA1991-14, Intl. Symp. of the SCA (EUROCAS), London, U.K., May 20-22, 1991.
2. Durand, C. and Lenormand, R., Resistivity Measurements While Centrifuging,
paper SCA9715, Intl. Symp. of the SCA, Calgary, Canada, Sept. 7-10, 1997.
3. Fleury M., FRIM: A Fast Resistivity Index Measurement Method, paper SCA1998-
29, Intl. Symp. of the SCA, La Hague, The Netherlands, Sept. 14-16, 1998.
4. Durand, C., Improvement of Fluid Distribution by Combined Use of X-Ray CT and
Continuous Local Resistivity Measurement, paper SCA2003-29, Intl. Symp. of the
SCA, Pau, France, Sept. 21-24, 2003.
5. Springer, N. and Korsbech, U., Resistivity Index Measurement Without the Porous
Plate: A Desaturation Techniques Based on Evaporation Produces Uniform Water
Saturation Profiles and More Reliable Results for Tight North Sea Chalk, paper
SCA2003-38, Intl. Symp. of the SCA, Pau, France, Sept. 21-24, 2003.
6. McCoy, D. and Grieves, W., Use of Resistivity Logs to Calculate Water Saturation
at Prudhoe Bay, paper SPE 28578, SPE Res. Eng., Vol. 12, No. 1, Feb. 1997.






SCA2004-55 5/6












Figure 1. Saturation exponent apparatus used for reservoir number one.













Figure 2a. The new m & n parameters produced a poor match with other
measurements of water saturation alerting us to a potential problem. The D
2
O traced
plugs measuring more than 30% SW (invasion corrected) are suspect due to heavy
invasion by mud filtrate.













Figure 2b. Using our normal cementation exponent of 2.1 and back calculating n to
match Archie SW to the other saturation measurements, we arrived at a value of 1.65.
Annular
Electrodes
Brine out
Oil in
V
I
R=V/I
Viton Sleeve
Bypassing electrodes
Conductive filter and
Capillary contact powder
2
nd
axial electrode
1
st
axial electrode
Isolated from rest
of end piece
Brine used for hydrostatic
in contact with end piece
Archie Parameters
A=1, M=2.1, N=1.65
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
MD (ft)
S
W

(
f
r
a
c
t
i
o
n
)
Si gma SW
Sponge Core (1-SOR)
OH SW
Tracer Pl ug SW Inv. Corr.
Li ve Oi l - Bri ne Pc SW
Hg Pc (O-B) SW
Archie Parameters
A=1, M=2.28, N=2.41
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
MD (ft)
S
W

(
f
r
a
c
t
i
o
n
)
Si gma SW
Sponge Core (1-SOR)
OH SW
Tracer Pl ug SW Inv. Corr.
Li ve Oi l - Bri ne Pc SW
Hg Pc (O-B) SW
SCA2004-55 6/6


















Figure 3. FRIM apparatus used for reservoir number two.
















Figure 4. Reservoir No. 2: Comparison of Archie calculated SW using new m and n
parameters and oil base core Dean Stark SW (circles) and oil-brine Pc SWi (squares). At
high SW, core SW is significantly less than Archie SW. A possible explanation is mobile
formation water displaced by the synthetic oil coring fluid. The choice of logging tool RT
has an impact on the saturation comparison.

Note: AIT & Merlin are marks of Schlumberger
Merlin (inversion) RT SW vs Oil Base Core SW
0.00
0.20
0.40
0.60
0.80
1.00
0.00 0.20 0.40 0.60 0.80 1.00
Archie SW
C
o
r
e

S
W

(
P
c

&

D
e
a
n

S
t
a
r
k
)
as received 'n'
FRIM ' n'
AIT 1 FT. RT SW vs Oil Base Core SW
0.00
0.20
0.40
0.60
0.80
1.00
0.00 0.20 0.40 0.60 0.80 1.00
Archie SW
C
o
r
e

S
W

(
P
c

&

D
e
a
n

S
t
a
r
k
)
as receive 'n'
FRIM ' n'
Side View of
Radial Electrodes
Oil in
V
I
R = V/I
Brine out
Water Wet
Ceramic
Viton
Sleeve
Radial
Electrodes
Sample Sample