Sei sulla pagina 1di 18

SPWLA TWELFTH ANNUAL LOGGING SYMPOSIUM, MAY 2-5, 19

A COMPARATIVE LOOK AT WATER SATURATION


COMPUTATIONS IN SHALY PAY SANDS

By

Walter H. Fertl and Gregory W. Hammack


Continental Oil Company

ABSTRACT

There are numerous domestic and overseas interpretation techniques


available to the log analysts for determining water saturation in shaly pay
sands. The theory of these different interpretation methods is briefly
outlined and the advantages and limitations are discussed. A comparative
study of these various techniques was made using actual field examples with
the effect of varying amounts of shaliness.

This comparative study illustrates the practical application of


these various approaches to the interpretation of shaly pay sands and pro-
vides the log analyst with more accurate results upon which to make his
recommendat ions .

INTRODUCTION

At present there are several interpretation methods available from


which to compute water saturations in a shaly pay sand. A study was made of
these various methods on a comparative basis using the same input parameters
in each of the equations tested.

This comparative study indicates why the use of one shaly sand
interpretation method over another may or may not comdemn a potential shaly
pay sand. It should also help to avoid mis-interpretati ons of future well
log data by selecting the method most suitable for that reservoir.

In this study we have attempted to show several of the more well


known techniques to compute water saturation in shaly pay sands using reser-
voir parameters representative for a shaly sand province, such as the Gulf
Coast area. These methods have been briefly outlined in theory and their
advantages as well as their limitations are discussed. In addition to the
more well known techniques, we have developed a new method of interpretation
of shaly sands which compares very favorably with several of these other
approaches.

We feel that this comparative study of these various shaly sand


interpretation methods plus the development of this new and simplified approach
should provide the basis for an improved evaluation of most shaly pay sands
that are encountered.

-1-
SPWLA TWELFTH ANNUAL LOGGING SYMPOSIUM, MAY 2-5, 1971

THE LAMINATED CLAY MODEL

One of the first models developed for shaly sand interpretations


has been proposed by Poupon etal. The model consists of two types of media
occurring in alternate layers , such as clean sand and shale laminae. The
approach assumes the conductivities of the shale and sand laminae to be
strictly additive. The model also assumes the same electrical resistivity
for shale laminae and adjacent shale formations.

Conductivity of such a sandwich type shaly sand may be expressed by


the combined effect of the shale and clean sand portion

c = Vsh csh + (1 - Vsh) csd (1)

where: c = measured conductivity of the shaly sand


csh = conductivity of laminated shale component
cstj= conductivity of clean sand
Vsh = proportion of shale lamination of the total volume

Substitution of Equation (1) in the basic Archie-formula yields


the following expression for the water saturation which holds for both
clean and shaly sands

Application of the Humble equation and a saturation exponent of


n = 2.0 allows the computation of water saturation for the laminated clay
model as follows

(3)

As may be seen from Equation (3) reliable data are necessary for
the porosity (~) formation water resistivity (Rw), the amount of shaliness
(Vsh) and the resistivity of an adjacent shale formation. In Figure 1 we
have shown the dependence of the computed water saturation from the forma-
tion resistivity and the amount of shaliness (Vsh = 10, 20, 30%) present in
a pay zone. Other input data, comparable to Gulf Coast conditions, include:
Porosity ($) = 30%, RW =0.045, and Rsh = 1.0

THE CLAY SLURRY MODEL

This clay - sand model was developed by L. DeWitte in 1950. It is


not the same as his more generally known SP model. This clay slurry model
consists of a clean sand pore structure with the clay dispersed within the
pore space. In other words, the clay minerals in the formation are assumed
to exist in a slurry with the formation fluids.

Considering both the resistivity of the mixture of dispersed clays


and connate water and the contribution by the resistivity of the
dispersed clay material (Rc) one can write

2
SPWLA TWELFTH ANNUAL LOGGING SYMPOSIUM, MAY 2-5, 1

Volume Slurry = Vol Brines + Vol Clay


Rz Rw Rc

or

$ Sw + Vsh =+s~+v~h
R= Rw ~

where:

R= = resistivity of the mixture of dispersed clays and


connate water.

R= = resistivity of the dispersed clays

($ Sw+ Vsh) = (Rz/Rt) 1/2

=
()
1 ($ sw+vsh)2= (($ Sw+vsh) (j&+vsh
Rt Rz RWF (4)

Rewriting the above equation yields an expression of the type:

/qx2+Bx+C=0

or

1 ?:SW2 + $ Sw
~ (k+kp+ (&~. =0

Thus, values for the water saturation S w can be solved as follows:

SFF,[Y+JW)} w
where:

In Figure 2 we have shown the computed values for water saturation


as a function of formation resistivity and shaliness. It is interesting
to note that application of Equation (5) in low resistivity pay sands gives
values of water saturation which are far too low, as a matter of fact Sw
values may even become negative. This condition becomes enhanced with in-
creasing shaliness in a potential shaly pay sand.

-3-
SPWLA TWELFTH ANNUAL LOGGING SYMPOSIUM, MAY 2-5, 1971

APPLICATION OF A PARABOLIC EQUATION

One of the most commonly applied equations for computing water


saturation in shaly pay sands is an expression of the form y = a + CX2.
In terms of a clay-sand model the resistivity of a shaly pay zone is expressed
by two additive electric components, the contribution by the clay present,
and the fluid distribution in the pore space.

Mathematically this model can be expressed such as

c = Vsh csh + @m . swn


a. Rw (6)

From the above equation the water saturation in either a clean or shaly
sand can be computed. Substituting the Humble relation in the equation and
a saturation exponent of n = 2.0 the water saturation can be computed as
fol 1 Ows

Sw= 0.9 1 Vsh Vsh (7)


T iiy~
J )

Again we have shown the change in computed water saturation as a function


of formation resistivity and degree of shaliness (Figure 3). Similar to
DeWitte’s approach our experience indicates that computed values of water
saturation in shaly pay sands using Equation (7) are far too optimistic.
Such Sw values are usually far too low, and with increasing shaliness of the
pay zone Sw values may even become negative.

THE GENERALIZED ARCHIE EOUATION

Another equation to describe a clay-sand reservoir model has been


proposed by Hossin. His model allows to compute the water saturation in
shaly pay sands provided the formation resistivity, porosity and shaliness
of the zone of interest are known. Using the concept of two electrically
parallel conductors, such as shale and sand, which are additives in their
effect one can write

c = Csh + Cs(j (f3)

where: c = measured conductivity


Csh= conductivity by shale component
Csd= conductivity contributed by sand component

In terms of physical parameters Hossin proposes an expression of


the following form:

c = Csh + Csd = vsh2cc + @m . Swn (9)


a. Rw

where: 1 = Rc = the resistivity of the dispersed clay. This value,


~ usually based on area experience, can be approximated

-4-
SPWLA TWELFTH ANNUAL LOGGING SYMPOSIUM, MAY 2-5, 19

by Rc = 0.4 x Rsh (Rsh being the shale resist ivity in an


adjacent shale zone).

Some simple mathematical modifications of Equation (9) including a substi-


tution of the Humble equation and a saturation value of n = 2.0, gives
the following expression for water saturation
Ii \ \
(lo)

In Figure 4we have shown the computed water saturation as a func-


tion of formation resistivity and shaliness.

DOLL METHOD

In an unpublished method by Doll an approach quite similar to the


method proposed by Hossin is suggested. Doll too uses the additive concept
of the clay and clean sand component and expresses the formation resistivity
as follows:

(11)

Some mathematical modifications of this equation and substituting a = 1.0,


a cementation exponent (m) = 2 and a saturation exponent (n) = 2.0 the water
saturation in a pay sand can be computed as follows:

(12)

A brief look at Equation (12) shows that the boundary conditions are satisfied
(Rc+ Rw and Vsh +0).

We have shown in Figure 5 the water saturation computed by this


method as a function of formation resistivity and shaliness. It is interesting
to note that this method gives very optimistic results, i.e., low water
saturation values. This is especially true in pay zones exhibiting high per-
centage of shaliness.

PATCHETT METHOD

Many quantitative shaly sand techniques have been suggested or


formulas of the type where

SW2 (13)

where A and B, expressed in terms of physical Parameters, stand for the


conductivity of the shale (CS) and the reciprocal of the-clean-sand forma-
tion factor (F). Furthermore, it is well knwn that the SP-curve is affected

-5-
SPWLA TWELFTH ANNUAL LOGGING SYMPOSIUM, MAY 2-5, 1971

by shaliness. A modified SP-equation has been proposed previously by DeWitte


which takes the clay effect into account such as

PSP = K log CSF + sw/Rw


CSF + Sw/Rmf

Patchett combined these two equations , expressing the water saturation by


the following mathematical relationship

-J
F Rw Rmf
—. .—x -1 (14)
Sw = Rt Rmf - Rw x

where: x loPSP/K
F ~ formation factor from density log using the Humble
equation
Rw= formation water resistivity from reliable source

It appears that this approach gives an improved evaluation of


shaly pay sands, provided the SP-curve can be used in a straightforward
manner. This statement implies that this technique should not be used when
the SP-curve becomes largely affected by invasion or in cases where the
resistivity value of the mud filtrate closely approaches that of the forma-
tion water. Obviously, this method cannot be applied in wells drilled with
oil base muds or in gas filled holes.

In Figure 6 we have shcwn the computed water saturation as a function


of the formation resistivity and shaliness of the reservoir rock. As common-
ly practiced we have expressed shaliness as a function of SP deflection such
as et = PsP/ssP.

THE SIMANDOUX EQUATION

Extensive studies by Simandoux on artificial media composed of


sand and clay have suggested that the conductivity can be expressed by the
relation

Ct = Vsh cc;: + Swn +m (15)


a.Rw

cc;: is the conductivity of the dispersed clay. If one assumes a


saturation exponent of n = 2.0 a parabolic equation, similar to that by
Waxman etal (1968) can be written

y=bx+cx2

Some mathematical modifications and substitution of the Humble equations


allows us to express water saturation such as

(16)

Laboratory experiments at the French Institute of Petroleum give


support to this relationship. We feel that at the present time this approach

-6-
SPWLA TWELFTH ANNUAL LOGGING SYMPOSIUM, MAY 2-5, 1

offers the best way to compute water saturation in a shaly pay sand. This
equation has become an integral part of the computerized Saraband sandstone
analysis used by Schlumberger domestically and is being supplied by Schlum-
berger Overseas at their computation center in Paris.

In Figure 7 we have shown the computed values of water saturation


as a function of formation resistivity and shaliness.

NEW SHALY SAND METHOD

As has been discussed previously, the straightforward application


of the Doll approach yields values for water saturation which may become
very small or even negative. However, we feel that these limitations can
be overcome. Based on Doll’s basic considerations we propose a modification
of his method. The approach consists of using two different saturation
exponents, a value of n = 2.0 in the clean sand term, and a value of nl = 1.0
for the shale component. Therefore, the proposed modification of the Doll
equation can be written as follows:

.
1 1
Sw=l (0.81 Rw/Rt);- Vsh (Rw/(o.4 Rsh))fi (17)
I
T {

After testing several values of n, we have chosen a value of


nl = 1.0. Figure 8 illustrates the change of computed water saturation as
a function of formation resistivity and shaliness of the reservoir rock.
As discussed later on in this report this proposed method appears to give
water saturation values very similar to the Simandoux equation over a wide
range of shaliness, in particular for reservoir characteristics as found
in the Gulf Coast area.

COMPARISON OF THE DIFFERENT METHODS

In Figures 10 and 11 we have attempted to show the relationships


of the previously discussed shaly sand techniques when compared to the
standard Archie - type equation in clean sands. From these graphical pre-
sentations it is obvious that depending on which shaly sand method the log
analyst has chosen, one may recommend or condemn any further development of
a potential pay zone, Also note that such apparent discrepancies become
enhanced with the increasing shaliness of a reservoir rock. For example,
evaluation of a rather shaly reservoir rock (Vsh = 30%) may give extremely
optimistic Sw values when using the DeWitte or Doll approach, whereas Poupon
approach in the range of Sw > 7S% becomes even more pessimistic than a clean
sand Archie approach. Specific inter-relationships of the various shaly
sand methods, as shown in Figures 9 and 10, are self explanatory.

Of even greater interest is a direct comparison of all shaly


sand techniques, discussed in this report, using Simandoux equation as a
reference basis. The graphical presentations of the computed results for
a formation with 10 percent shaliness and rather shaly reservoir rock

-7-
SPWLA TWELFTH ANNUAL LOGGING SYMPOSIUM, MAY 2-5, 1971

(v*h = 30%) are shown in Figure 11.

As expected the apparent differences between all shaly sand techni-


ques become enhanced with the increase of the shale content in a potential
pay sand. Several important conclusions can be derived.

In fairly clean, i.e., not too shaly formations (V~h < 10%), the
methods by DeWitte, Doll, and I%tchett give reasonably close answers. The
same is true for Hossin’s method in the lower range where Sw < 40%. However,
the best agreement is obtained between the Simandoux method and our newly
developed shaly sand method (Figure llf).

in the shalier reservoir rock (Vsh = 30%), the above observations


become enhanced. The Doll and DeWitte approaches are too optimistic by 10
to 25 saturation percent, whereas Poupon and Hossin for Sw > 26 to 28 percent,
and Patchett over the total saturation range give a too pessimistic evaluation
of the water saturation present. However, again excellent agreement is ob-
served between our newly developed method and the Simandoux approach (Figure
llf).

Finally, we have checked the range of agreement between the Simandoux


and our new method under reservoir conditions different from the ones en-
countered in the Gulf Coast area. Conditions studied include low to high
porosity rocks, fresh and saline waters, several values for shale resistivity,
and shaliness in the reservoir rock ranging from 10 to 40 percent.

n general, the difference in computed water saturations by our


new method and the Simandoux equation can be expressed mathematically such as

A SW.

or

ASW = f] + fz

Looking at the terms fl and f2 independently, the following conclusions can


be drawn:

(1) fl is independent of the formation resistivity

(2) f2 will become smaller with increasing Rc

Obviously, the agreement between the Simandoux and our new equation
would not be expected to hold true under all reservoir conditions, since the
difference in computed water saturations (ASW) is a combined function of
porosity, shaliness, resistivity of formation water, shaly slurry and formation.
A case where the disagreement is large is illustrated in Figure 12 where
Rw = 1.0 and Rc = 6.o. Similar conditions, for example may be encountered in
the Rocky Mountain area.

-8-
SPWLA TWELFTH ANNUAL LOGGING SYMPOSIUM, MAY 2-5, 1

In general, however, this new method can be expected to yield less


than five saturation percent difference from the Simandoux equation where
Rw is less than 0.065 and the 0 is greater than 25 percent. Furthermore,
Rt in the pay zone should be higher than 0.8, and the shaliness in the reser-
voir rock should not exceed 30 percent. Figure 13 represents conditions
one would expect to encounter in the Gulf Coast area, such as Rw = 0.04
and Rc = 0.4. Note that under these conditions for all Gulf Coast sands
of commercial interest very good agreement is observed between the Simandoux
and the new method equations. Also note that in sands having about 25 percent
porosity or higher and the amount of shaliness ranging from 10 to 30 percent
an excellent agreement between both equations is obtained.

CONCLUSIONS

Several shaly sand techniques are currently being applied by the


industry. The differences in the computed water saturations increase with
the increase of shaliness in the reservoir rock. Techniques by DeWitte
and Doll give too optimistic appraisals and in very shaly pay zones become
completely unrealistic. In shaly pay sands located in transition zones, i.e.,
exhibiting high water saturation, the methods by Hossin and Poupon become far
too pessimistic and their application may condemn pay zones of commercial
significance.

Based upon this study we recommend application of the Simandoux


equation or our newly developed shaly sand technique, in particular if reser-
voir characteristics are similar to the ones experienced in the Gulf Coast
area.

-9-
SPWLA TWELFTH ANNUAL LOGGING SYMPOSIUM, MAY 2-5, 1971

BIBLIOGRAPHY

DeWitte, L., “Relation Between Resistivities and Fluid Contents of Porous


Rocks”, Oil and Gas Journal, pp 120-132, August 24, 1950.

Doll, H.G., method unpublished.

Hossin, A., “Calcul des saturation eneau par la methode cu ciment argileux
(formule d’ Archie general isee). Bull. A.F.T.P.

Patchett, J, G., “An Approach to Determining Water Saturation in Shaly


Sands”. Journal Petroleum Technology, October, 1967.

Poupon, A., et al., “A Contribution to Electric Log Interpretation in Shaly


Sands.” Journal Petroleum Technology_, August, 1954, pp 29-34.

Simandoux, P., “Mesures Dielectriques en Milieu Poreux, Application a Mesure


des Saturations en Eau, Etude du Comportment des Massifs Argileux”,
Revue de l’institut Francais du Petrole, Supplementary Issue, 1963.

Waxman, M. H. and Smits, L. J. M., “Electrical Conductivities in Oil-Bearing


Shalv Sands”, Society of Petroleum Engineers Journal, June, 1968.

NOMENCLATURE

SW water saturation, percent

4 effective porosity

F Formation resistivity factor

m cementation factor

n saturation exponent

Rw formation water resistiviry

Rt formation resistivity

c measured conductivity of shaly sand

csd conductivity of clean sand component

csh 1 conductivity (res stivity) of sha e component


~

RZ resistivity of mixture of dispersed clay and connate water.

Rc resistivity of dispersed clay, approximated by Rc = 0.4 Rsh

v~h percent shale of total volume; laminated, dispersed and


structures or any combination.

1o
1 I I I I 20
I I I I I I I I I 1 I
I I
I LAMINATED CLAY-SAND MODEL I CLAY SLURRY MODEL
18 — (L. DEW ITTE)
I (POUPON et 01) I
I
SHALINESS (%)
I
I
I
16
I
I
I
I CLEAN SAND
I

I
I 14 — f’
CLEAN SAND

(
I

\
12 .

\
\
SCALINESS \ \
1 (0 —
\
\
!0 \
\
‘h \ \,

\
\ \
6 -\

Q20

K
\

2 —

I 1 I I I I
10 20 30 40 50 60
I
70
I I I I I I I 1
0 10 20 xl 60 30 60 70
WATER SATURATION. SW wATER SATURATION. SW

FIGURE i FIGURE 2.
SPWLA TWELFTH ANNUAL LOGGING SYMPOSIUM, MAY 2-5, 1971
. J
L
o I
a
$
o
a
a
a
a

._ —- ——
I I I I I 1 I I I
m In * . 0 m m * m
wv 8U ‘A LIA11SIS3U NOIIVWUOd
12
SPWLA TWELFTH ANNUAL LOGGING SYMPOSIUM, MAY 2-5,
I I I I I I I I 1
I
II
m
-13-
SPWLA TWELFTH ANNUAL LOGGING SYMPOSIUM, MAY 2-5, 1971
m
Ill
K
3
L
0
-.
II
: 0
, .
8
m’
0
.
0
0
,
x
z
0
z
a
x
In
SPWLA TWELFTH ANNUAL LOGGING SYMPOSIUM, MAY 2-5, 197
0
lN33M3d ‘(31 H3MV) ‘S
*
o
II
,%
lN3YS3d ‘(31 H3MV) ‘S
15
v~” = 10

/ 1
20
I I 1 J,
40 00 ao BO 20 40 60 w
SW ( DE WITTE ), PERCENT Sw ( POUPON ) , PERCENT sw (HOss IN) , PERCENT
FIG. I la FIG. Ilb FIG. I IC

1 I I I

2
/
0
/
,

I I I I
/,, ,
20 40 60 ao Zo 40 60 m
SW(OOLL) , PERCENT 5W (pATcHETT ) , PERCENT Sw ( NEW METHOO), PERCENT
FIG, Ild FIG. I 10 FIG. Ilf

COMPARISON OF SEVERAL SHALY SAND METHODS WITH THE SIMANDOUX APPROACH

FIGURE I I
SPWLA TWELFTH ANNUAL LOGGING SYMPOSIUM, MAY 2-5, 1971
I I 1 I r
I I I
-m m
. .
-- 11II
. . -s
II II was >
.
=:
‘(ly. /“
Mm
———————— . .
11 11
.
ru - ua:
. .
II 11
=: ——
d~.” ‘\ ——
— — .
—— ——
mm
. .
II II m.
. .
.
II II
61:
. >
~7s2
1 I I I 1 I 1 1 I I 1 I
w N o m w * N o
W! ‘A11A11SIS3N NO11VWM03
o
I I I I
I I 1 I
I
I I o
N
-1.-.
. .
11 II
z
m: o
‘m 0
-.
. .
II u
z
m; _/ o
R ——— n-l
.— —— ——
7
mm-l
L
. .
11 II
.
m:
- —. .—
r.-, .
11

. .

——
II

m-
a:

e4-
.
11 11
x

.
——— ~
//
.c
mg i

_—-’”
——— — —— ro-
,,yl, ,,, ,,-
. .
1( 11
.
m:
I i I I
* N o In * IN o
‘w.! ‘All AllS;S311 NO11VIW04
17
i SPWLA TWELFTH ANNUAL LOGGING SYMPOSIUM, MAY 2-5, 1971

ABOUT THE AUTHORS

GREGORY W. HAMMACK is a native of Fort Worth and graduated from Texas


Christian University with a 6A degree in Physics. He joined Core
Laboratories where he worked in California, Texas, Oklahoma and Louis-
iana areas. He was later associated with The Elgen Corporation and
held several positions including Assistant Director of Field interpre-
tation. Prior to joining Continental Oil Company, he was Staff Engineer
with the Lane Wells Company. He is presently a Senior Formation Evalua-
tion Engineer with Continental Oil Company’s Production Engineering
Services group in Houston, Texas

WALTER H. FERTL was born and educated in Austria, and graduated with a
Dipl. Ing. degree in petroleum engineering from the Montanistische
Hochschule (Mining University) in Leoben, Austria, in 1963. His summer
training periods were spent with several oil and service companies in
West Germany, Austria, and Middle East. During 1963 - 1965, he asso-
ciated with the Austrian Petroleum Company as assistant to the field
manager in well completion and workovers. He then attended The University
of Texas, department of petroleum engineering, receiving his M.S. degree
in 1966 and his PhD in 1968. During his graduate studies at the Univer-
sity of Texas he was instructor in Well Log Analysis Institutes of this
university and the University of Zulia, Maracaibo, Venezuela. In 1968
he joined the Formation Evaluation group at the Production Research
Division of Continental Oil Company. He is a member of the SPE of Al ME,
the SPWLA, the Canadian Well Logging Society and the Austrian Mining
Society. He has also authored or co-authored several technical papers
in both Engl ish and German.

-18-

Potrebbero piacerti anche