Comparison of Water Saturation Values

Determined from Capillary Pressure

Measurements and Old Logs
R. AGUILERA
Servipetrol Ltd.

Abstract
Calculations of water saturation were carried out for the A
Sand of the Bridger Lake Field using techniques based on: (1)
capillary pressure measurements, (2) old well logs and old con-
ventional interpretation charts, and (3) cross plot and pattern
recognition. It was found that approaches (1) and (3) provided
very close results for average water saturations (31.0 and
30.7%). Calculations using method (2) gave misleading results
(65 and 100%).
It is concluded that techniques based on capillary pressure
measurements and crossplot pattern recognition provide reliable
values of average water saturation. The former has the major
disadvantage of being susceptible to error in the determination
of the volumetric midpoint of the reservoir. The latter has the
advantage of accounting for water saturations in different zones,
which still allow us to determine an average water saturation for
the whole sand.
It is recommended to exercise special care in the use of con-
ventional “cookbook” interpretation charts and computer pro-
grams. The indiscriminated use of these charts and programs can
result in serious error.

Introduction
FIGURE 1: Location map of the Bridger Lake Field.
The Bridger Lake Field is located about 95 miles east of Salt
Lake City (Utah) at the northern foot of the Uinta Mountains on -m
the extreme south flank of the Green River Basin. The legal loca- F = aφ ..................................................................................................(3)
tion according to Utah regulations is Range 14 East, Township 3
North. Figure 1 shows the location of the Field.
The A Sand is located in the lowest part of the Cretaceous
Dakota formation. The Dakota formation lies between the Mowry Water Saturation From Capillary Pressure
formation and the Morrison formation at an average depth of
6,500 ft. below mean sea level. Core analyses of these zones indi- Measurements
cated good oil show with very erratic values of mean permeabili- For the calculation of the average water saturation, capillary
ty. The sand grains are basically light brown, and their size varies pressure measurements were made on 13 samples for different
from fine to medium. water saturations as shown in Figures 2, 3, 4, and 5. Table 1 sum-
The purpose of this study was to compare the values of water marizes the properties of these samples. The logarithm of perme-
saturation determined from capillary pressure measurements and ability for each sample was plotted against water saturation hold-
logs. The former requires knowledge of the geometric mean per- ing capillary pressure constant. This plot yielded approximate lin-
meability and the volumetric mid-point of the reservoir. The latter ear trends represented by straight lines in Figure 6.
is based on the following relationships(1): The next step was to calculate geometric mean permeability.
This was accomplished by means of the equation(3):
−1/ n
Sw = I .............................................................................................(1) n

I = Rt / FRw = Rt / Ro .........................................................................(2)
log kg = ∑ F log(k )
j =1
j a j
......................................................................(4)

34 Journal of Canadian Petroleum Technology

 FIGURE 2: Capillary pressure curves, samples 171 A, 172 A, and FIGURE 3: Capillary pressure curves, samples 164 A, 173 A, and
179 A. 177 A.

where curve represents the average capillary pressure of the reservoir.

kg = geometric mean permeability, millidarcies
Fj = frequency of j interval, fractional
The height above the water-oil contact to the volumetric mid-
(ka)j = arithmetic average permeability of logarithmic class point of the reservoir was estimated to be 80.5 ft. from a structural
interval j map taking into account the highest point in the reservoir and the
n = total number of classified intervals water oil contact at 6,746 ft. below mean sea level.
Table 2 shows the calculation of geometric mean permeability
which was found to be 77 millidarcies. On inclusion of this value
in Figure 6, values of water saturation were found for different This distance was converted to capillary pressure by the
capillary pressures and plotted as shown in Figure 7. The resulting equation:

FIGURE 4: Capillary pressure curves, samples 169 A, 170 A, 176 FIGURE 5: Capillary pressure curves, samples 160 A, 162 A, and
A, and 175 A. 165 A.

September 2000, Volume 39, No. 9 35

 FIGURE 6: Correlation of water saturation with permeability for
various capillary pressures. FIGURE 7: Bridger Lake average capillary pressure.

TABLE 1: Properties of samples in which capillary pressure measurements were made.

Depth, Feet Permeability Porosity

Sample No. From To Millidarcies Per Cent

160A 15,559 15,560 0.45 8.9

162A 15,561 15,562 73.20 13.3
165A 15,564 15,565 267.00 13.5
169A 15,568 15,569 148.00 13.2
170A 15,569 15,570 220.00 13.2
175A 15,574 15,575 0.97 5.7
176A 15,575 15,576 29.70 11.2
164A 15,563 15,564 209.00 12.6
173A 15,572 15,573 182.00 12.1
177A 15,576 15,577 18.70 8.3
171A 15,570 15,571 258.00 12.8
172A 15,571 15,572 223.00 14.6
179A 15,578 15,579 25.50 13.1

TABLE 2: Geometric mean permeability of the Bridger Lake Field.

Average
Permeability Permeability Number Cumulative
Range of Range of Frequency Frequency Fjx
(Millidarcies) (ka)j Samples Fraction Fj log (ka)j

1.3 – 2.5 1.9 0 0.0000 0.0000 0.0000

2.6 – 5.0 3.8 4 0.0952 0.0952 0.0552
5.1 – 10.0 7.5 3 0.0714 0.1666 0.0625
10.1 – 20.0 15.0 2 0.0476 0.2142 0.0560
20.1 – 40.0 30.0 1 0.0238 0.2380 0.0352
40.1 – 80.0 60.0 8 0.1906 0.4286 0.3390
80.1 – 160.0 120.0 6 0.1429 0.5778 0.2971
160.1 – 320.0 240.0 15 0.3571 0.9349 0.8500
320.1 – 640.0 480.0 3 0.0714 1.0063 0.1914

42 1.8864

Geometric mean permeability = antilog 1.8864 = 77 millidarcies

36 Journal of Canadian Petroleum Technology

TABLE 3: Basic data from logs and core analyses.

Depth ∆t Rt(6FF40) Porosity** ∆t – ∆tm

Zone Feet µsec/ft. ohm-m Per Cent µsec/ft.

1 15555 – 58 65 28 8.37 11
2 15558 – 60 68 30 11.20 14
3 15560 – 64 74 22 14.50 20
4 15564 – 66 67 60 13.50 13
5 15566 – 74 68 55 14.20 14
6 15574 – 78 66 22 11.80 12
7 15578 – 82 72 20 15.50* 18
8 15582 – 85 67 40 11.10* 13
9 15585 – 87 64 60 8.70* 10
10 15587 – 89 80 5 22.50* 26

* Estimated from Figure 10

** Arithmetic average porosity for each zone (A value of porosity was available every foot)

TABLE 4: Initial values of water saturation. h(ρ w − ρ o )

Pc =
144 ...................................................................................(5)
Sw
Zone I Per Cent Values of water and oil densities were determined to be 62.85
1 1.00 100.0
and 52.73 lb./cu ft. respectively by standard procedures. On inclu-
sion of the above values in Equation (5), the capillary pressure
2 30/15 = 2.00 70.0 was calculated to be 5.65 psi. Entering this value in Figure 7, the
3 22/8 = 2.75 59.0 average water saturation was determined to be 31.0%.
4 55/18 = 3.05 55.0
5 31/15 = 2.01 70.0
6 1.00 100.0 Water Saturation From Cross Plot and
7 20/10 = 2 70.0
Pattern Recognition Technique
8 40/19 = 2.10 70.0
9 60/31 = 1.93 70.0 The determination of water saturation from logs using this
10 1.00 100.0
approach was carried out as explained by Pickett(4), with the use
of the following equation(2):

FIGURE 8: SP-Induction log. FIGURE 9: Gamma ray, caliper, and sonic logs.

September 2000, Volume 39, No. 9 37

TABLE 5: Laboratory waterflood study.

Fork A Well No. 2

Oil Saturation
Oil Saturation Oil Saturation* at 30:1
Before Flood After Flood Water-Oil Ratio
Sample No. Per Cent Per Cent Per Cent Sand

61A 90.2 45.3 47.0 A

63A 85.2 62.7 65.0 A
64A 86.0 48.2 52.0 A
67A 86.5 63.7 66.0 A
Average 87.0 55.0 57.5

* Water-Oil Ratio in excess of 100:1

Fork A Well No. 3

Oil Saturation
Oil Saturation Oil Saturation* at 30:1
Before Flood After Flood Water-Oil Ratio
Sample No. Per Cent Per Cent Per Cent Sand

66A 87.7 50.2 56.0 A

76A 86.7 49.7 52.0 A
79A 85.0 50.7 53.0 A
Average 86.7 50.2 52.7

* Water-Oil Ratio in excess of 100:1

log Rt = − m log( ∆t − ∆t m ) + m log B + log Rw + log I ....................(6) were in the order of 55%. This value was corroborated in a publi-
cation which showed a typical relative permeability curve for the
Bridger Lake Field (Figure 12)(5). The above information clearly
Equation (6) tells us that a plot of log Rt vs. log(∆t - ∆tm) indicates that a high amount of residual oil was lost in bringing
should result in a straight line with a slope of -m. From the above the cores from the reservoir to the laboratory.
equation we can see that the resistivity index I can be calculated As a consequence the water-bearing line was shifted (solid
from such a plot even if B, Rw, or m are unknown. line) so as to make the dashed line represent a hydrocarbon satura-
To apply the aforementioned method, the A Sand was divided tion of 55%. Using this new water-bearing line, values of Sw (23
into ten zones, as indicated in Figures 8 and 9. Readings of depth, to 42%) were determined for each zone as indicated in Table 6.
∆t, and Rt were taken for each zone. These values are presented in The average water saturation for the whole A Sand was deter-
Table 3. Values of porosity were determined from core analyses. mined to be 30.7%, by weighting the water saturations for each
A plot of ∆t versus φ was drawn with the above data (Figure zone with thickness (Table 6). The agreement of this value with
10). The response of the sonic log at the matrix (∆tm) was deter- the Sw calculated from capillary pressure measurements (31.0%)
mined to be 54 ohm-m by extrapolating the straight line to zero is excellent. Consequently it is concluded that both methods pro-
porosity. Values of ∆t - ∆tm are indicated in Table 3. The next step vide a valid way to calculate average values of water saturation.
was to draw a log-log plot of Rt vs. ∆t - ∆tm (Figure 11). The An important corollary from this conclusion is that the volumetric
dashed line was taken initially as the water bearing trend. On this mid-point of the reservoir truly represents the average water satu-
basis values of I (1.9 to 3.05) and Sw (55.0 to 70.0%) were deter- ration of the whole sand for this data set.
mined for each zone as indicated in Table 4. Other properties derived from the application of this technique
Residual oil saturations determined from a conventional fresh were:
core analysis were in the order of 15%. However, waterflood Water resistivity, Rw = 0.062 ohm-m
experiments(2) (Table 5) indicated that residual oil saturations Cementation factor, m = 1.88

TABLE 6: Final values of water saturation.

Sw Thickness, h
Zone I Per Cent Feet Sw x h

1 27/5.4 = 5.00 42.0 3 126.0

2 30/3.4 = 8.82 31.0 2 62.0
3 22/1.8 = 12.2 26.0 4 104.0
4 60/4.0 = 5.00 24.0 2 48.0
5 55/3.4 = 16.2 23.0 8 184.0
6 22/4.5 = 4.90 42.0 4 168.0
7 20/2.1 = 9.55 30.0 4 120.0
8 40/4.0 = 10.0 29.0 3 87.0
9 60/6.4 = 9.40 30.0 2 60.0
10 5/1.0 = 5.00 42.0 2 84.0

34 1043.0
Average Water Saturation = 1043.0/34 = 30.7 per cent

38 Journal of Canadian Petroleum Technology

 saturation determined from capillary pressure measurements and
old logs. The Bridger Lake Field was taken for practical applica-
∆t vs. Porosity tion. The following conclusions were derived from the preceding
discussion:
1. Techniques based on capillary pressure measurements and
crossplot pattern recognition provide reliable values of aver-
age water saturation.
2. The volumetric mid-point of the reservoir truly represents
the average water saturation of the whole sand for this data
set.
3. Special care should be exercised when using conventional
“cookbook” interpretation charts and computer programs.
The undiscriminated use of these charts and programs can
result in serious error.
4. Improved results can be obtained by properly combining
information obtained from logs and core analyses.
5. The method based on capillary pressure measurements has
the major disadvantage of being susceptible to error in the
determination of the volumetric mid-point of the reservoir.

NOMENCLATURE
∆t = response of the sonic log, µsec./ft.
∆tm = value of ∆t at zero porosity (matrix ∆t), ∆sec/ft.
FIGURE 10: Core porosity vs. sonic transit time (µsec/ft.).
φ = porosity, fractional
ρw = water density, lb./cu ft.
Slope of the linear relation between ∆t and porosity, B = 118 µ- ρo = oil density, lb./cu ft.
sec/ft. porosity (from Figure 10). a = constant in formation factor equation
B = slope of the linear relation between ∆t and porosity
F = formation resistivity factor
Water Saturation From Conventional h = distance from water-oil contact to volumetric mid-point,
feet
Charts I = resistivity index
Old Schlumberger charts(6) were used to calculate water satura- m = porosity (or cementation) exponent
tion for each zone. In an effort to improve results the whole sand n = saturation exponent
was taken as a unit, and the maximum SP (-52 mv) was used to Pc = capillary pressure, psi
determine Rw (0.6 ohm-m). Resistivities were read directly from Ro = resistivity of a non-shaly sample 100% saturated with
the old 6FF40 log, and porosities were determined from core water of resistivity Rw, ohm-m
analyses. With the use of Equations (1), (2), and (3) or the con- Rt = true formation resistivity
ventional Nomograph, values of Sw were determined to range Rt = true formation resistivity, ohm-m
between 65 and 100%. These water saturations were obviously Rw = water resistivity, ohm-m
wrong. Consequently it was concluded that a straight forward Sw = water saturation, fractional
“cookbook” technique was not adequate for this case.
Calculations using this technique are indicated in Table 7. REFERENCES
1. ARCHIE, G.E., The Electrical Resistivity Logs as an Aid in
Determining Some Reservoir Characteristics; Tran. AIME, 146, pp.
Conclusions 54-62, 1942.
2. AGUILERA, R., Naturally Fractured Reservoirs; Second Edition,
The purpose of this study was to compare the values of water PennWell Books, Tulsa, OK, p. 521, 1995.

(∆t - ∆tm) vs. Rt

FIGURE 12: kro, krw relative permeability curves (from

FIGURE 11: Pickett plot for “A” sand. Reference 5).

September 2000, Volume 39, No. 9 39

TABLE 7: Calculation of water saturation using “cookbook” technique.

Rt (6FF40) Porosity Sw
Zone ohm-m Per Cent F = φ-m I = Rt/FRw Per Cent

1 28 8.37 105 28/(105 × .60 < 1.0 100.0

2 20 11.20 61 20/(61 × .6) < 1.0 100.0
3 22 14.50 37 22/(37 × .6) = 1.0 100.0
4 60 13.50 44 60/(44 × .6) = 2.3 65.0
5 55 14.20 40 55/(40 × .6) = 2.3 65.0
6 22 11.8 56 22/(56 × .6) < 1.0 100.0
7 20 15.5* 35 20/(35 × .6) = 1.0 100.0
8 40 11.1* 62 40/(62 × .6) = 1.1 95.0
9 60 8.7* 97 60/(97 × .6) = 1.02 98.0
10 5 27.5* 11.2 5/(11.2 × .6) < 1.0 100.0

SP = -52 mv
(Rmf)e/(Rw)e = 3.9 (From Chart SP-1, Reference 6)
(Rw)e = 0.334
Rw = 0.6 (From Chart SP-2, Reference 6)
Rmf = 2.77 @ 77˚ F
Rmf = 1.30 @ 181˚ F (BHT)(From Chart A-6, Reference 6)

*Estimated from Figure 10

3. AMYX, J.W., BASS, D.M., and WHITING, R.L., Petroleum

Reservoir Engineering; New York, McGraw Hill, 1960. Author’s Biography
4. PICKETT, G.R., A Review of Current Techniques for Determination
of Water Saturation from Logs; Journal of Petroleum Technology, Roberto Aguilera is the president of
November 1966.
Servipetrol Ltd. in Calgary, Alberta. He is a
5. BURT, R.A., High Pressure Miscible Gas Displacement Project
Bridger Lake, Unit, Summit County; paper SPE 3487, presented at petroleum engineering graduate from the
the 46th Annual Meeting of the Society of Petroleum Engineers of Universidad de America at Bogota,
AIME, New Orleans, LA, October 3 – 6, 1971. Colombia, and holds masters and Ph.D.
6. Schlumberger—Log Interpretation charts; 1968. degrees in petroleum engineering from the
Colorado School of Mines. He has ren-
Provenance—Original Petroleum Society manuscript, dered consulting and training services to oil
Comparison of Capillary Pressure Measurements and Loss, and gas companies worldwide. He is the
(98-82), first presented at the 49th Annual Technical Meeting, author of the book Naturally Fractured
June 8-10, 1998, in Calgary, Alberta. Abstract submitted for Reservoirs, PennWell, Tulsa, OK (1980,
review December 10, 1997; editorial comments sent to the authors 1995), and co-author of the books Technology of Artificial Lift
November 5, 1998; revised manuscript received March 22, 1999; Methods, Vol. 4, PennWell (1984) with Kermit Brown et al.,
paper approved for pre-press April 6, 1999; final approval August Horizontal Wells, Gulf Publishing Co., Houston, Texas (1991),
30, 2000. and Determination of Oil and Gas Reserves, Petroleum Society
Monograph No. 1 (1994). Dr. Aguilera received the Outstanding
Service Award from the Petroleum Society in 1994. He is a
Distinguished Author of the Journal of Canadian Petroleum
Technology (1993 and 1999) and an SPE Distinguished Lecturer
on the subject of Naturally Fractured Reservoirs for the 2000 –
2001 season.

40 Journal of Canadian Petroleum Technology