Reservoir Volumetric Parameters Defined by Capillary

Pressure Studies

Downloaded from http://onepetro.org/trans/article-pdf/210/01/252/2176489/spe-710-g.pdf by guest on 27 February 2024

S. H. ROCKWOOD
MEMBER AIME
SHELL OIL CO.
G. H. LAIR
MIDLAND, TEX.
B. J. LANGFORD
JUNIOR MEMBERS AIME

T. P.4636

ABSTRACT capillarity in oil production; Jan hydrocarbons and water in the reser-
Law's' and A. C. Bulnes3 statistical voir. This is the reservoir's total
Volumetric reservoir analysis is treatment of core analysis data; porosity. When true total porosity
dependent upon a firm relationship Chalkley's method,,5 for estimating values are not available, correlations
between porosity, connate water, and specific surface areas and porosity; between porosity and capillary pres-
net pay. Capillary pressure data on Walter Rose's· report on porosity, re- sure characteristics lose their signifi-
core samples interrelate these three serves, and recovery; Archie's'" tex- cance since they cannot be related
factors. It is shown that other reser- tural classification of carbonate rocks; through oil-base mud core data to
voir problems may be resolved: (1) Stewart and Spurlock's' large core the porosity and connate water values
whether water production is indi- analysis; Purcell's'o mercury injection found in reservoirs.
genous or extraneous to the oil pro- method in studying capillary pres- Methods used to obtain reliable
ducing interval, (2) an oil-water con- sure phenomena; Stewart, Craig, and porosity data will not be discussed;
tact is being approached, and (3) Morse's" model multiple-phase flow however, care should be exercised
whether the oil-water contact is re- test in investigating the relative per- in selecting samples for measure-
motely situated. meability effect; and Fatt's" illumi- ment. In homogeneously porous for-
nating network model of porous me- mations, small core plugs sampled
INTRODUCTION dia. frequently will furnish adequate in-
Volumetric analysis of carbonate The present study shows that the formation. Where heterogeneity, par-
reservoirs challenges the ingenuity static reservoir parameters of poro- ticularly vugginess, is present, large-
of the reservoir engineer due to an sity, connate water, and net produc- core porosity data may be required;
absence of data on the minimum oil- tive thickness may be interrelated by in such reservoirs a suite of capillary
bearing and oil-producing pores of a utilizing a statistical arrangement of pressure samples may be selected
given reservoir rock. Many studies mercury capillary pressure data in from small plugs.
with this as the ultimate objective addition to routine core analysis.
are available on the physical char- This report summarizes the techni- CAPILLARY PRESSURE
acteristics and fluid-flow behaviors of ques that are followed to obtain
these relations which are applicable Capillary pressure data provide a
the carbonate reservoir. Some of these
to carbonate and sandstone reser- method for relating two carbonate
pUblications are: Hassler, Brunner
voirs that do not contain clays. Capil- reservoir parameters, porosity and
and Deahl's' study of the role of
lary pressure statistical studies may connate water saturation, which in
be extended to include clay-bearing turn permit delineation of productive
Original manuscript received in Society of
Petrolewn Engineers office on June 15, 1956. rocks by utilizing water capillary rock types necessary for selecting the
Revised manuscript received June 10, 1957.
Paper presented at Petroleum Branch Fall pressure data. third parameter, productive thick-
Meeting in Los Angeles, Oct. 14-17, 1956. ness. Specific applications of capillary
lReferences given at end of paper. POROSITY pressure data to petroleum reservoirs
Discussion of this and all following tech-
nical papers is invited. Discussion in writing It is necessary that the laboratory were suggested by Archie' and Pur-
(3 copies) may be sent to the offices of the
Journal of Petroleum Technology. Any dis- determine the porosity value that in- cell.'o Their work demonstrated that
cussion offered after Dec. 31, 1957, should be
in the form of a new paper. cludes the pores that are filled with in preferentially water-wet rock, pore

PETROLEUM TRANSACTIONS, AIME SPE-710-G 252

structure and fluid distribution are the wetting phase. In reservoirs the plugs are selected will determine how
related because of capillarity, and wetting phase is usually connate well a resulting correlation will rep-
that mercury injection curves are water, thus a plot of estimated con- resent a cored interval of interest.
valuable data for defining the rela- nate water saturation vs capillary Sampling of a suite of cores should
tive number of pores, their inter- pressure can be obtained from the be performed so that each suite
connection, size, and size distribu- mercury injection curve. Such a plot of core plugs represents: ( 1) only
tion. is illustrated in Fig. 2. one rock type of the cored interval,
As a result, mercury capillary pres- With homogeneous rock it is rela- (2) the complete porosity range en-
sure curves are used to provide a tively simple to represent reservoir countered in the interval; (3) the
direct estimate of the oil-occupied capillary conditions with a few capil- incidence of permeability existing
pore spaces in a reservoir. Water- lary pressure curves. When the rock within a chosen porosity increment;
occupied pore spaces are determined is heterogeneous, as typified by car- and (4) the porosity region where
by difference of total and oil-occu- bonate reservoirs, the problem re- pore sizes change from a predom-
pied pore spaces. These water values quires special techniques. Neverthe- inance of the larger pores in the
result from the complementary na- less, mercury injection curves are higher porosities to mainly smaller
ture of mercury and water capillary especially useful in evaluating car- pores in the lower porosities, this
pressure data. Consequently, the bonate reservoirs. Although their pore region lying between the asymptotic
water saturations may be related to geometries vary greatly, both vertic- portions of the estimated connate
reservoir capillary conditions through ally and laterally, definite character- water vs porosity curves shown
the more conveniently determined istic trends may be observed for a herein.

Downloaded from http://onepetro.org/trans/article-pdf/210/01/252/2176489/spe-710-g.pdf by guest on 27 February 2024

laboratory mercury capillary pres- given set of conditions. Consequent- The sampling procedure and tech-
sures. ly, a suite of cores may be sampled niques for deriving the above men-
so that data on small pieces of rock tioned statistical relations can best
MERCURY CAPILLARY PRESSURES
represent the capillary characteristics be illustrated by an example. Twenty-
Fig. 1 illustrated a plot of mer- of the cored interval on a statistical eight capillary pressure curves were
cury volume injected vs injection basis. This does not mean that the run on the carbonate interval 10,323
pressure, or capillary pressure for a capillary pressure curves obtained to 10,360 ft in Shell-University
mercury-air system. Mercury, a non- from a sample are representative of 9-B-1. The capillary pressure data
wetting phase, is injected into an the capillary conditions for the foot are given in Table 1, and the average
evacuated (40 microns or less) core of core from which the sample is
sample and the injected volumes are selected. To accomplish this, it would
recorded at various pressures. Ex- be necessary to analyze large cores
ternal vug volume is filled at zero or to select five or six samples from
capillary pressure, and therefore must each foot. What is important is the
be determined by a separate measure- fact that the capillary pressure curve
ment as discussed SUbsequently. Note does represent the capillary proper-
that at some extremely high capillary ties for rock of a particular porosity
pressure the mercury volume in- and type.
jected would equal the pore volume Since the major objective is to
of the sample. Throughout the pres- determine a statistical relationship ESTIMATED CO~"TE WAlER (~) - - 100
(PIRE VOLUlE UHOCCUPIED BY MEROJRY ")
sure range, the trace of the curve between capillarity and porosity, the
represents the volume of non-wetting FIG. 2-AVERAGE MERCURY CAPILLARY
manner in which capillary pressure PRESSURE CURVES.
fluid that fills the larger pore spaces
at any capillary pressure. Thus it TABLE 1-CAPILLARY PRESSURE DATA, DEVONIAN FORMATION, SHELL·UNIVERSITY 9·B·1, BLOCK 9
represents the effect that the pore FIELD, ANDREWS COUNTY, TEX.
Deplh Per Cent Pore Space Unoccupied by Hg at
geometry has upon hydrocarbon sat- II Lilhology
Porosity
Per Cent
Perm.
Md 5 atm 11 aim 21 aim 62 aim
uration. At any pressure, the por- 10360·61 IiTi:A 2.3 0.04 98 ~ ~ ~
tion of the pore space that is not 10362·63 III LA 2.4 0.21 92 85 78 68
10348·49 III LA 2.8 0.05 98 95 62 44
occupied by mercury is occupied by 10350·51 III MA 2.9 0.04 97 94 82 37
AVERAGES 2.6 0.09 96 93 80 60
PW:E $:ZE DISTRIBUTiON CURV;: 10353·54 IIIMA 3.3 0.08 97 92 87 43
(Mwcwy Injection Data)
10336·37 III MA 3.3 0.14 97 87 67 53
r INFIHITY
10338·39
10355·56
10333·34
III MA
III·I·MB
IIIMA
3.5
3.6
3.7
0.08
0.10
0.35
94
96
81
69
91
43
49
64
32
40
42
27

I t I.,\ I I I I I 10344·45
10337·38
III LA
IIIMA
3.5
3.9
0.15
0.09
90
98
63
62
51
43
45
38

.. 10343·44
AVERAGES
10331·32
III LA

III MB
3.9
3.6
4.6
0.15
0.14
1.07
64
90
43
51
70
24
49
55
17
42
41
15
10356·57 1I1·I·MB 4.0 0.11 98 96 61 35
50 10342·43 III·LA 4.5 0.36 66 48 43 38
10341·42 III MA 4.6 0.31 86 56 53 44
AVERAGES 4.4 0.46 73 56 44 33.
• 10357·58 III·I·MB 5.1 0.56 57 24 17 13
ij 10329·30 III MB 5.5 1.34 75 58 61 35
f-~ 3 10323·24 III MB 5.0 1.33 43 28 23 23
w 10328·29 III MB 5.6 2.31 39 27 24 19

f-~
~
\ 2
AVERAGES 5.3 1.39 54 34
21
28 23
16
~ 10324·25 III MB 6.2 6.09 28 17
g [\. \
10325·26
10327·28
III MB
111MB
6.3
6.5
2.51
1.88
36
49
26
35
22
32
20
28
10322·23 III MB 6.7 3.40 35 25 20 17
t
EFFECT OF
t'--- t-- .~ 10326·27
10323·24
III MB
III MB
6.9
6.9
4.71
8.10
30
23
23
16
20
13
18
11
1 I 6 S , 3 2 0°
t- c. c. MERCURY INJECTED AVERAGES 6.6 4.45 34 24 21 18
o C. Co WEnIHG PHA.SE _ 6.6
III MB 7.7 7.65 24 17 15 12
10330·31
FIG. I-LABORATORY DERIVED MERCURY 10324·25 III MB 7.8 6.76 30 23 19 17
CAPILLARY PRESSURE CURVE. AVERAGES 7.7 7.21 27 20 17 15

253 VOL. 210, 1957

capillary pressure curves for each to generally produce 100 per cent CAPILLARY PRESSURE
porosity group are plotted in Fig. 2. water. ILLUSTRATrONS
Note that only a few pressure points A third general concept is based
VARIATIONS RESULTING FROM
are needed to adequately represent on the relative amounts of large
CHANGES IN ROCK TYPE
each curve. An arbitrary selection capillaries to total capillaries able to
of 5, 11, 21, and 62 atm has been hold hydrocarbons. This is shown by A group of illustrations is avail-
used, where the 5-atm curve repre- the spread between the 62 and 5-atm. able to show that the statistical ap-
sents large pore effects and the curves. The large capillaries repre- proach of applying capillary pres-
62-atm curve the small pore effects. sented by the 5-atm curve will hold sure data yields unique relations be-
Furthermore, the average porosity hydrocarbons in very weak capillary tween porosity and water saturation
for each group is represented by environments, while hydrocarbons for a given rock type. When rock
capillary pressure d-ata typical of that held by finer capillaries are in in- type changes, however, one such re-
group. In addition, rock lithology creasingly stronger capillary environ- lation will not adequately represent
according to Archie's" classification ments as the 62-atm curve pore the data. From the information that
is denoted for each sample. In this sizes are approached. According to has been processed, it appears that
illustration, all samples have the Fig. 3, the 7 per cent porosity inter- in a given reservoir, three or more
same general lithology, but if rock vals contain 83 per cent large pores relations might be needed: one for
type changes occur, separate porosity based on the 62-atm hydrocarbon- Types I and III carbonates, one for
groupings are made for each type occupied pore spaces. Type II carbonate rock, and one for
present. This will be illustrated by A fourth consideration is obtained clastic sediments. Grain size of the
rock matrix of each rock type also

Downloaded from http://onepetro.org/trans/article-pdf/210/01/252/2176489/spe-710-g.pdf by guest on 27 February 2024

further examples. from the data composing each point.
A given rock type will have over-all influences these relationships. For
STATISTICAL PRESENTATIONS
trends in basic petrophysical data the characteristics of each carbonate
Fig 2 shows that capillary pres- when there is a smooth-line relation- type, refer to Table 2 which presents
sure, connate water, and porosity are ship of the 62-atm porosity-small the carbonate classification proposed
interrelated. To provide a basis for pore values, and the individual values by Archie:
interpreting reservoir capillary pres- composing each point show only An example of the difference be-
sure effects, a statistical relationship minor variations from the average. tween capillary pressure characteris-
is obtained between porosity and The associated 5-atm data may show tics of Type I-III rock and Type II
connate water at constant capillary a trend in average values, but the rock is shown in Fig. 5. Note that
pressure for the Devonian Type III individual data composing each point the earthy Type II material has much
reservoir in the vicinity of Shell- may show either a good agreement higher water saturation and finer
University 9-B-1. This statistical pre- or a wide-spread variation. A good pores than the Type I-III material.
sentation of the trend of pore size agreement shows good large-pore in- These finer pores are a result of the
variation with porosity changes terconnection while the latter shows very fine crystals inherent to Type
should be observed for five general poor large-pore interconnection. II rock. The heterogeneity of the
concepts. The first concept is evi- A fifth consideration is the rapid Monahans Clear Fork reservoir is of
denced by the 62-atm curve. Since increase in connate water at constant an extent that requires zoning in or-
it shows the very-fine-pore changes, capillary pressure as porosity de- der to apply these statistical rela-
it is related to average connate water creases. This characteristic of these tions to core and logging data from
saturation for each porosity of the statistical correlations is used to de- other wells in the same field.
reservoir rock that is situated a great termine minimum hydrocarbon-bear- An explanation of a single sta-
distance above the free water table. ing porosities. As shown in Fig. 3, tistical relation for both Type I and
This curve also shows the irreducible this porosity value is about 1.6 per Type III rock is illustrated by Table
minimum connate water values at cent for the 62-atm curve, and is 3 (data shown graphically in Fig.
lower positions, and may be com- increasingly higher as the free water 5). Note that the Type I classifica-
pared to actual connate water values level is approached. This porosity tion, associated with low porosity,
to determine if water production is value is of sufficient importance to changes to Type I-III for interme-
indigenous to the hydrocarbon pro- be checked by use of residual core diate porosity and Type III for high
ducing interval. Applications of this saturation data. For example, Fig. porosity. If the capillary pressure
statistical method do not always re- 4 shows core water and oil satura- data for a reservoir containing both
tions plotted vs porosity for Shell- Type I and Type III rock yield simi-
quire conversion of height to labora-
Desert Creek 26-36. Four plots are lar smooth-line relationships between
tory capillary pressure values since shown to represent all carbonate porosity and connate water satura-
it will be shown by subsequent illus- types present in the reservoir. Each tion, then it can be assumed that
trations that log calculated or oil- plot has the characteristic sharp porosity change is accompanied by a
base mud porosity and connate water increase in water saturation at change in rock type. In other words,
values may be used to reproduce the lower porosities. Of particular signifi- the change in lithology is one of the
62-atm capillary pressure statistical cance, however, is the fact that both direct physical factors that cause
relationships of producing sections Laterolog calculated water saturation changes in pore volume and pore
in many thick carbonate reservoirs. and core oil saturation indicate size distribution.
A second general concept is based equivalent minimum oil-bearing poro- Other cases have been noted where
on the statistical effects of the large sities. In the event core saturations Type I rock grades into Type III
pores. The 5-atm curve has been are not available, fluorescence and rock, or into Type II rock, in a man-
selected herein to relate connate cut data may be used in conjunction ner that permits a unique correlation
water values and porosities close to with the estimated connate water vs between porosity and water satura-
the water table. Water ilaturations porosity curves to determine mini- tion. However, correlations that en-
higher than this curve are indicated mum oil-bearing porosity. compass a gradation from Type I to

PETROLEUM TRANSACTIONS, AI ME 254

 • AT.5 ATMS Hg (ABS) -e CORE OIL SATURATION
oAT 11 ATMS Hg (ABS) X CORE WA TER SATURA TION
D AT 21 ATMS Hg (ABS) - - -lATEROLOG CONNATE WATER
6 AT 62 ATMS Hg (ABS) -4---- OIL SA TURATION("l -OIL SATURATIO~ C!il
1 100 80 iO 40 20 o 100 80 iO AO 20 o
2& 2&
10 \ OOLOMITE I VF ,: OOLOMITE mVf/ I
~ZO ~.
, .j ~2O 't'- __ •
lCK:TYpk "t-. __ .'-- ':~Olt.~
, ..,1
9 R ;::1& ., , EI&
I~:~ kltl' ;;;
,
ilO :ii!IO
8 ,
~ 5 &
g7 ~ mMEt
0 ,
Ii!

00
. '1
~\ ~
0 20 40 80 100 20 40 iO 80 100
mMB
~6
WATER SATURATION ( % ) - WATER SATURATION (")~

~~ ~ r...
~5 ~~ ~ ~o
mMB
f"'-.. r-..,mL&M A&S
~OIL SATURATION (%) ~Oll SATURATION (%)
04 ........,
3
mL&M A
~~
.......
t-- r--.....
~
b..
100
2&
80 iO

LIMESTONE mVF
40 20
t---
100
25
80 iO

LIMESTONE I YF
40 20 0

2
mLA r-...... r.:::::: ~ ~~5
0 ~20
;::15
r-- t-- ;;; \
i,o \ t,- ,
100" WATER_=~ ~I 0
1 --: -... '""'.' ••
....V 1-:.,-
~ ~ 5 ~
OfCUrIE~ P~OSlp
o0
&
.. .. .
::;-. .L j9·OIL .1--.- 1---- ~OIL-

Downloaded from http://onepetro.org/trans/article-pdf/210/01/252/2176489/spe-710-g.pdf by guest on 27 February 2024

'
1o 20 3D 040 50 60 70 80 90 10o 0
o 20 40 80 100
0
0 20 40 iO 80 100
ESTIM4TED COIfIIATE WATER (%) WATER SA TURATION {~----+- WATER SATURATION (")~

FIG. 3-TYPICAL CARBONATE RESERVOIR CAPILLARY FIG. 4-MINIMUM OIL·BEARING POROSITY.

PRESSURE CORRELATION. Paradox formation, Shell-Desert Creek 26-36, Desert Creek area, Utah.
Devonian formation, Shell-University 9-8-1, Block 9 field,
Andrews County, Tex.

both Types II and III rock exist lations are generally not possible. An lary pressure data are illustrated by
only if very fine crystals are present example is a case where Type I or Figs. 3 and 6. Both cored intervals
in all types. With all data processed Type III rock has medium-fine crys- show Type III rock, with pore sizes
to date, the smoothest correlations tals, whereas any Type II present ranging from Type A matrix pores,
occur when a uniform crystal size is would have very fine crystals. Such not visible at 10 power magnifica-
present throughout the entire porosity a reservoir would require zoning for tion, through the various secondary-
range, as noted by the data of Table an effective reservoir analysis. type pores B, C, and D, as explained
3 and most of the figures presented by Table 2. The significant difference
PORE SIZE EFFECTS
in this paper. When uniform crys- between the two wells is indicated by
tal size does not exist, unique corre- The effects of pore size on capil- the relative shapes and separation of
TABLE 2-CLASSIFICATION OF CARBONATE ROCKS
Approximate Porosity (Per Cent)
Visible Porosity
(Per Cent of Surface) Total Porosity
Class Appearance Appearance or Luster Crystal or Size of Pore-mm (Sum of Matrix and
Type A Matrix
(Texture of of Hand Under Microscope (About lOX) Grain Size ·0.1 mm 0.1·2.0 mm 2.0 mm _ _ _V.:.:i"si:::b",le,,)-,---=-_
(Not Visible
Matrix) Sample General Range (mm) lOX) Type B Type C Type 0 A·B A·C
I Crystalline, hard, Matrix made up of crystals Resinous L* 0.4 2 e.g. 10 e.g. 15 12 17
Compact dense, sharp edges tightly interlocking, allowing
and smooth faces no visible pore space be- M 0.2
on breaking. tween the crystals, often to
Resinous producing "feather-edge" 0.1
appearance on breaking due Vitreous"''''
to fracturing of clusters of VF 0.05 5 e.g. 10 e.g. 15 15 20
crystals in thin flakes, prob-
ably along cleavage planes
IJ Dull, earthy or Crystals, less effectively in-
Chalky chalky appearing, terlocking than above, join-
hard to soft, crystal-ing at different angles. Ex-
line appearance ab-tremely fine texture may still
sent because the appear "chalky" under this
small crystals are power, but others may start
less tightly inter- appearing crystalline. Grain Chalky** VF 0.05 15 e.g. 10 e.g. 15 25 30
locked, thus re- size for this type is less
flecting light in than about 0.05 mm. Coarser
different directions, textures classed as Type III
or made up of ex-
tremely fine granules
or sea organisms.
May be siliceous or
argillaceous
III Sandy or sugary Crystals less effectively in- Finely 0.1 10 e.g. 10 e.g. 15 20 25
Sucrosic appearing terlocked, fracture generally Sucrose
(Sucrose) along individual crystal faces M 0.2
giving a rough or sandy ap-
pearance. Generally more
space between crystals. Ool- 0.4 e.g. 10 e.g. 15 15 20
ite, pisolite and other granu-
lar textures also fall in this Coarsely
class Sucrose

*L-Large (coarse); M-Medium; F-Fine; VF-Very Fine

In cases where cuttings are between vitreous and chalky in appearance, designate as I-II or II-I. Samples are considered in the Vf group when the grain
or crystal size is difficult to distinguish (12X-18X). Put in f group when grains are easily distinguished
When pores are greater than about 2.0 mm and therefore the pores occur at the eoge of the cuttings (for example, sub·cavernous pores). Amount of such
porosity is indicated by the per cent of cuttings in an interval showing eyidem;1I' of farge ~ores
SYMBOLS,
I I I F-B10: finely sucrose (therefore, matrix porosity about 10 per cent), visible porosity about 10 per cent, total porosity about 20 per cent
(II-I) VF-A: Chalky to vitreous, very fine texture (therefore, matrix porosity about 8 per cent), no visible porosity, total porosity about 8 per cent

255 VOL. 210, ]957

 IHE RAIN E0 CR'fS TALLIN E & SUCROSIC TABLE 3-ROCK TYPE AS RELATED TO POROSITY, CLEAR FORK FORMATION, MONAHANS FIELD, WARD
AND WINKLER COUNTIES, TEX.
!IY C
ESTIMAT ED CONNATE WATER Fine Grained Very Fine Grained
VS. Crystalline and Sucrosic Rock Earthy Rock
;lel POROSITY
, CLEAR FORK FORMATION Porosity Per Cent Rock Type Porosity Per Cent Rock Type

5 •
1-'
W~RO
"0 NAHANS FIELD
tNf
WINKLER COUNTIES, TEXAS 1.5
2.4
IFA
I FA
4
J
L, VERY FIN E
3.4
4,4
1 F AS
(1,1·111) F AS 4.6 I Y FA
2 1\' GRAINED fA. RTHY 5.4 (1·111) F AB 5.5 1·11 Y FA
I 0 6.6 1·111 f BC 6.4 (I-II, I) YF AB
0 7.6 1·111 F BC 7.4 (I-II, I) YF BC
9 • 8,4 (1-111,111.1) F BC 8.3 (1-11,1·111) YF BC
8 •
~
p\
'0
9.3
10.3
(I-III, III-I) F BC
1·111 FM BC
9.5
10.3
II-III YF AC
(I-II, II) YF BC
7 11.5 (1·111, 111·1) F BC 11.3 I-II, YF BC
6 12.4 (III-I, I-III) F BC 12.4 (I-II, 11·1, II-III) YF BC
0 13.5 (111-1,1.111) F BC
5 13.5 II-III YF BC
4 62 AIMS. 14.3 (111·1 F, I-III r, III M) BC 14.4 (1-111,11·111) YF B
J
15.3 (1·111, III-I, III) F BC 15.3 (I-III, 11·111) Y FC
2 • 61 AIMS.

~ 100 % WATER
16.7
17.4
(111·1 F, I-III F, III hi) BC
(III M, 111·1 F) BC
16.7
17.0
II-III YF BCD
II YF A
1 OCCUPIED POROSITY 20.3 III F BC 18.5 11·111 YF B
00 10 20 30 40 SO 60 70 80 90100 19.5 11·111 YF BC
ESTIMATED CONNATE WATER (%)

MERCURY CAPILLARY PRESSURE DATA AT 6? AIMS. lABS.)·

• FINE GRAINED CRYSTALLINE & SUCRDSIC SAMPLES
lower capillary pressures, it should should have no effect on the capil-
o VERY FINE GRAINED EARTHY SAMPLES be possible to calculate the oil in lary pressure correlations unless
FIG. 5-EFFECT OF CABONATE ROCK TYPE place for this type reservoir that thickness is related to rock type
ON CAPILLARY PRESSURE CORRELATIONS. would be affected by such a water change_ Such a relation has not been

Downloaded from http://onepetro.org/trans/article-pdf/210/01/252/2176489/spe-710-g.pdf by guest on 27 February 2024

drive from porosity data and the 5- observed. As an example, the capil-
the 5- and 62-atm curves. The data atm curve which is used to estimate lary pressure data of a 391-ft core
for Shell-University 9-B-l (Fig. 3) connate water values, The ultimate from Shell-State GA-2 yielded a
show an increasingly larger spread recovery from a given rise in the smooth relation between estimated
between the 5- and 62-atm curves water table then could be calculated connate water and porosity (Fig. 8).
as porosity decreases. This is to be to determine the amount of by-pass-
expected since the intermediate size CLOSURE OF MERCURY AROUND
ing that has occurred, and the de-
pores (type B) that are present in THE TEST SAMPLE
velopment pattern and/or producing
the higher porosities are absent in method could be evaluated, When obtaining capillary pressure
the lower porosity range. Further, data, one phase of the laboratory
the wide separation indicates that a DOLOMITIZATION EFFECTS procedure is quite critical. Particu-
relatively large transition zone should Experience to date indicates that lar care must be exercised to deter-
exist between the producing water dolomitization of a limestone does mine the point of complete closure
level and the free water level, de- not have any unique relation to capil- of mercury around the sample_ If
pending on transition zone porosities. lary pressure correlations. An ex- complete closure is not obtained, the
This is verified somewhat by residual ample of the presence of both lime- effective bulk volume of the sample
oil in cores to depths 280 ft below stone and dolomite in a cored inter- in the cell will be greater than the
the producing water level of this res- val is shown in Fig. 7 for a 68-ft true bulk volume. As a result, the
ervoir. section in Shell-Hilburn No.3. Note measured volumes of injected mer-
In contrast to this, Shell-Ratliff that the capillary pressure correla- cury will be falsely high. The effects
Bedford No. 15 (Fig. 6), an El- tions are represented by a smooth of this will be particularly noticeable
lenburger producer, shows two sig- curve. in low porosity rock as evidenced
nificant petrophysical characteris- in Fig. 9. Note that the correlation
tics: (1) a small separation of the GROSS THICKNESS points for Shell-Foster No_ 5 form
5- and 62-atm curves, and (2) a Theoretically, gross thickness a smooth curve, whereas for the
close similarity in shapes of these
curves. The latter reveals that there "f--Hlr-+--t- ESTIMATED CO'flATE WATER VS. POROSITY
ESTIMA. TED CONNATE WATER VS. POROSITY DOLOMITE \ LIMESTONE
is a uniform pore size distribution \tt;;:ru!v";'. ELLENBURGER FORMA TlON WOLf CAMP fORMATION
SHELL.RATLlFF·BEOFORD NC. 15 SHELL· HILBURN NO.3
throughout most of the porosity BEDFORD FIELD
ANOREWS COJNTY, TEXAS TOWNSEND FIELD
range. Also the small spread shows LE A CDUNTY, NEW MEX ICO
15 f--H-II-++...., 26~r-r-.-.-~-r-r-r-.-'
that low capillary forces are in- ~OCK TYPE
lCl--+-+-+-+-:.:r"T-'-F-t--+---t
volved in the pores that are holding 13 ~22'~+-+-+-~~~~~-+---t
most of the oil in place since pro- 11 '--t-+t--r-
>-
~~~-+-f-f~~~~t-+-i
duction data on wells discussed sub- ll-H-II-+ ~ 18:f+,ol--+-+--+
E
sequently will show that water will ~ 10 ~ 16f+-HHH=t-t-t-t-t-t
be produced whenever the connate ~ ,
water value of a given zone falls be-
tween the 21- and 5-atm correlation
lines. These two petrophysical facts
show that there should be a very
sharp "oil-water" contact with a very
thin transition zone. Of greater im-
portance, the low capillary forces of
the hydrocarbon-bearing pores should
AT 62 ATIIS Hi (ABS)
permit a water sweep for a known +COI'IIATE WATER CALCULATED
FROIiLATEROLOG
water drive to be nearly 100 per
FIG. 6--EFFECT OF CARBONATE RESERVOIR FIG. 7-DoLOMITIZATION HAS NO EFFECT
cent efficient. As this sweep is mainly PORE SIZE ON CAPILLARY PRESSURE ON CAPILLARY PRESSURE CORRELATIONS.
through pores with only 5 atm and CORRELATIONS.

PETROLEUM TRANSACTIONS, AIME 256

other wells, the lower porosity points TABLE 4-WATER CAPILLARY RISE ABOVE A FREE WATER TABLE
Calculated
depart from a smooth curve. This Height
Above Mercury Lithology
Theoretical
is explained by the fact that the sam- Free Capillary Pore Factor
Perm. Water Pressure Radius ( Length Core) 2 Porosity Resistivity
ples from Foster No. 5 were core (md) (Ft) (atm) (em) length Pores (Per Cent) Factor
plugs such that mercury closure was ----0:1 150 --10- 7 x 10- 5 0.030
--6-
~
l. 100 7 10" 0.056 9 200
easily obtained, and the samples for 10. 50 3 2 X 10- 4 0.1l9 13 65
the other wells were small, reverse- 100. 20 I 6 X 10- 4
10- 3
0.147 17 40
1000. 10 0.6 0.198 22 23
circulation diamond cores such that Water-Oil density difference is 0.23; interfacial tension times cosine 8 for Mercury-Air 340, and for
closure was more difficult to obtain. Oil·Water - 35
This error would be amplified
presented in Table 4 to illustrate re- 1 md for the rock matrix permeabil-
throughout the entire porosity range
lationship of matrix permeability and ity may hold 100 per cent water for
in data from drill cuttings.
height above a water table for typical 50 and 100 ft respectively above a
Other than the case just cited, carbonate reservoir properties, inter- free water level compared to only a
mercury closure effects may appear facial tension variations being omit- few feet for vuggy-type pores. At
in a suite of vuggy samples. The ex- ted 13 • There is a given set of condi- these heights above the free water
ternal vug volume shown in Fig. 1 tions for each rock sample that in- level of an oil-bearing vuggy reser-
is determined by comparing the bulk fluences the height to which water voir, vugs that are connected through
volume measurement of a sample will rise above the free water level, such a matrix might produce water-
with empty vugs to the bulk volume and more than one rock type may free oil initially, but a subsequent
of the sample with wax-filled vugs. exist in a given reservoir. From these pressure draw-down would cause

Downloaded from http://onepetro.org/trans/article-pdf/210/01/252/2176489/spe-710-g.pdf by guest on 27 February 2024

The capillary pressure curve is ob- facts stems the carhonate reservoir water to be produced from the rock
tained by injecting mercury into the problem. matrix in amounts that would depend
rock matrix of the dewaxed plug Reference to Table 4 indicates mainly upon the degree of vug in-
and adding the vug and injected mer- that the 62-atm estimated connate terconnection, and only slightly on
cury volumes of each pressure step. water-porosity curve is representative relative permeability relationships.
of capillary conditions at great
CONNATE WATER CAPILLARY PRESSURE COMPARISONS
heights above the free water level.
Field data also present some evi- Conversion to feet above the free
From the foregoing considerations, water level of a given reservoir is
dence that this curve may be valid
which are based on actual reservoir usually impossible. Consequently this
in depletion reservoirs where inter-
sample data, it has been shown that statistical approach offers a conven-
connection is poor between the res-
mercury capillary pressure correla- ient correlation technique of reser-
ervoir and an aquifer. However, with
tions may be used to represent the voir and laboratory data. Several ex-
other depletion reservoirs and all
trend of variations of connate water amples are available to provide a
water-drive reservoirs, when the pro-
saturation with porosity. One method comparison of estimated water satu-
ductive members approach the vi-
of application to a specific reservoir ration from capillary pressure curves
cinity of the free water level, con-
requires that mercury capillary pres- with water saturations indicated by
nate water saturations increase be-
sure be related to heights above a oil-base mud core data or log calcu-
cause of capillarity. Thus, an esti-
free water level. This is estimated mated connate water-porosity curve lations.
by converting a mercury-rock capil- representing a lower capillary pres- Fig. 3 depicts a clean oil-produc-
lary system to an oil-water-rock sure must be used. tive section located a considerable
capillary system, and by equating Again referring to Table 4, it distance above the water level. Note
capillary forces and gravity forces in shows that pores having 10 and that good comparison is obtained be-
the latter system. An example is tween water saturation at 62-atm
NO.SAMPlE$ ROCK TY-;PE_-::-:,=-::-:-:':7.'1 capillary pressure and water satura-
IG mMe AT62ATIII5.~asl tion calculated from the Laterolog.
mlLco IDOL~ AT Q ATIIS. ... (A8SI 15 mM8 eSHfLL-wrdlTON
• NOS.9~1I
An equally good comparison is
10 1 1 OSllll·FOmIt shown in Fig. 7 for similar condi-
I I tions.
" ID ~CO DOLO. Another example of good agree-
• I I ment between capillary pressure data
and oil-base mud core saturations is
l - f-
1 1
IDLCII
"lID ~C
.0.
r"
shown in Fig. 10. The departure of
this good agreement at lower porosi-
• " II LC DOLO.
ties may have been caused by evapor-
6
ID .. & L co DOLO. ative losses before the core was
• m IIICO DOLO. canned. These losses become more
) " noticeable for lower pore volumes.
0 l m L:D DOLO.
1 IIICO DOLO.0
J Note that the in situ water satura-
o 0
ID . 1 _ DOLO.
• "I >C~ 1"~"'tm'II"r"~
tions for both the sucrosic and earthy
't-i- '1' r
6}.!. &II·IL&IIIC&O DOLO.
0 & LIIL
rocks are represented by the 62-atm
• ..... I .& .... &0 DOL0. correlation curves. This indicates
I&ID·I"'''&C DOL~~] 'j or
that the productive zones are far re-
.. . ..
1& l.m,&",u DOLO.
l

10 10 II
I
..
1 I 'A DOLO. A moved from the free water level, or
'" '" '0
that they are a depletion-type reser-
FIG. 8-GROSS THICKNESS HAS NO EFFECT
voir. Since Sealy Smith No. 61 pro-
ON CAPILLARY PRESSURE CORRELATIONS. FIG. 9-MERCURY CLOSURE EFFECT ON duced 11 per cent water on final
Dolomite separated by limestone, Devonian for· CAPILLARY PRESSURE CORRELATIONS. completion, the 12 per cent per-
motion, Shell·State GA·2, Caudill field, tea Clearfork formation, North Russell field, Gaines
County, N. M. County, Tex. meability-feet of Type II earthy rock

257 VOL. 210, 1957

 "FINE: ",RAINED CRYSTALLINE: & SUCRO$IC
20
0

:1 ,
MERCURY CAPILLARY PRESS URE CORRELA TION LINE

0
ESTIMATED CONNA TE WATER VS. POROSITY
CLEAR F~ K FORMA TION
7r-~~
"'c !;t N=..!
19 oc
w~
:g- IOl~ o t J o ~J t--
i3/'" + WATER SATURATION _ t--
-
MONAH ANS FIELD
6r-i· 18
7 WARD'AND WINKL ER COUNTIES, TEXAS
... ! :fl~
I II--I 17 oct-
6 15 r-f~ 0..", ......
\ ~~ -
5
x VERY FINE GR AINED ENHHY \C.,\ \ 16 -t=- on
4 I ~~RCURY CAPI LLARY PRESSURE I~
It x \
J
ORRELATION UNE

3 \\ \ 15
,
1
\+
1 \ ~ ~ "- 14
0 \ \ '\
9 f'-c -I1 i\. ~ C\ 13
\
t:
8
7 • ~1 0 r\. N -12
,!!
\ \ \
~I I
;;;
\

-
6
5 ~9 r-INITIAL WATER CUT
o \
t-
enlO
+ \
4
x
61 ATMS.
A.8 C·, r- ~ \ o \ .\
x t-f"62 ATMS, ~9
~
J
1
x 100% WATER
t--.OCCUPIED
C·6
r-C.ll
0..
8 \ [,+, \
1
0
o 10 20 30 40 50 60 70 !'IO .0100,
POROSITY

6
35" i
7
\ .\ \ \
ESTIMATED CONNATE WATER (%)
SHELL· SEAL Y SMITH /10. 61: 5
\ 6 \ +
"\..
OIL BASE MUD CORE WATER: ESTIMATED CONNATE WATER '4. POROSITY + \
X FINE GRAINEI) CRYSTALLINE 8. SUCROSIC 'iAMPLE5 CALCAREOUS DES MOINES SAND 5
OVERY FINE GRAINED EARTHY SAMPLES
.INDUCTION LOG CONNATE wATEq
SHELL. BURNETT NO. C-6
4 r-... + ~
BURNETT RANCH DES MOINES SAND FIELD
ESTIMATED CONNATE WAT~ .......
FIG. 10--GOOD AGREEMENT BETWEEN OIL· KINGCDUNTY. TEXAS 3
s::

Downloaded from http://onepetro.org/trans/article-pdf/210/01/252/2176489/spe-710-g.pdf by guest on 27 February 2024

VS. POROSITY
BASE MUD CORE SATURATIONS AND MER· + WATER SA lURA TlON CALCULA TED
CURY CAPILLARY PRESSURE CORRELATIONS. 1 FROM ELECTRIC LOGS ON - t-- 2 SAN ANDRES FORMATION
~L- UNIVERSITY NO.30-A-a
r-
BURNETT NOS. C.2. C-6. C.1I
o I DUNE FIELD r-
that had greater than 50 per cent I02ll:.l«J:lU6Qmlll Q II CRANE COUNTY TEXAS
ESTIMA TED CONNA TE W" TEi'! ClIl
connate water and less than 0.1 md 00 !O 20 30 40 50 60 70 80 90 100
FIG. ll-INITIAL WATER CUTS RELATED ESTIMATED CCNIIATE WATER("Iol
is presumed to be the main cause of TO CAPILLARY PRESSURE CORRELATIONS. FIG. 12-CORE AND CAPILLARY PRESSURE
this water production. In many cases DATA FOR A WELL PRODUCING INITIALLY
this type of rock inherently has an ative permeability to water. Since 46 PER CENT WATER.
amount of water saturation that is Curve A frequently represents the
accompanied by finite values of rel- capillary relationships in carbonate has been established that the per-
ative permeability to water. reservoirs, connate water saturations meability measured in the laboratory
Examples also are available of res- higher than the irreducible values is a function of the technique ap-
ervoirs that produce both oil and frequently are the cause of "unex- plied and the sample orientation. An-
water from intervals relatively close plained" water production. Although other drawback is the lack of any
to the water level. These examples the cited example represents extreme criteria for determining permeability
are illustrated by Figs. 11 and 12. opposites in pore size distribution, values applicable to changing capil-
For each figure, note that the in situ it serves to illustrate the difficulty lary conditions, especially as related
connate water saturations, deter- of predicting whether a reservoir to the height above the free water
mined either from oil drilled core will produce water unless in situ table.
or log data, are represented by capil- water saturation can be compared From the many studies that have
lary pressures somewhat lower than with capillary pressure data. been made on carbonate reservoirs, it
62 atm. This fact alone does not in- appears that the most accurate and
dicate that water should be pro- PRODUCTIVE THICKNESS consistent method for selecting pro-
duced; however, it is a clue that ductive intervals is based upon con-
water saturations are higher than ir- Before volumetric reservoir anal- nate water saturation-porosity rela-
reducible values. Whether or not ysis is possible, the productive thick- tions. This method first involves de-
water will be produced depends upon ness, or net pay, of the gross inter- termination of that relation from oil-
the relative permeability of the rock val under consideration must be de- base mud core data, capillary pres-
to water at the water saturation in termined from laboratory or logging sure data, log calculations, or com-
question. Any increase in relative data. This item may be apparent in binations of these. Next, a porosity
permeability to water caused by an sand-shale reservoirs; however, lime- profile for the interval in question
increase in water saturation is a func- stone reservoirs embrace a wide range is obtained by employing core data
tion of the pore size distribution and of rock porosities, most of which or neutron logs. Connate water satu-
pore network of the rock. may have some hydrocarbon satura- rations for each foot also may be de-
As an example, consider two sam- tion. rived from core data, or for ap-
ples with capillary pressure charac- In many instances, permeability is propriate intervals of the porosity
teristics represented by Curves A the criterion used to delineate pro- profile of logging data.
and B in Fig. 13. A small increase ductive intervals. Usually the selected The collection of data outlined
in water saturation above the irre- intervals have air permeabilities above will enable selection of a cut-
ducible value for Curve A will be greater than 0.1 md. Other limiting off porosity. This cut-off porosity
associated with a large decrease in values are used, however, depending is selected in the region of the cor-
the forces holding water in place. As upon reservoir conditions. This pre- relation curve that is associated with
a result, the relative permeability sents one of the disadvantages in a drastic change in pore size distribu-
to water may materially increase. using permeability, since no good tion. As a result, there is a negligi-
Conversely, with the sample repre- method is available to predict cut- ble volume of hydrocarbons present
sented by Curve B, an equal increase off permeabilities for specific reser- in porosities lower than the cut-off
in water saturation would cause only voir geometries. Furthermore, per- value due to high connate water
a slightly lower capillary pressure, meability is a parameter that cannot saturation, low pore volume, and
resulting in a small increase in rel- be measured in a well bore, and it lower recovery efficiency for the

PETROLEUM TRANSACTIONS, AIME 258

 rosity. Basing a net pay estimate on 14
WATER - OIL CONTACT AT 5321'
the latter reservoir property has many n
distinct advantages over methods lOr-!il1.R..~ ",'..r PRESSURE (ATMS.)
based only on permeability. The 18 1 + LATEROLOG DATA
-[
porosity method is more consistent
t A in that it is based upon definite re- IrIf Me 1\ 529}Jl}
cf lations for each reservoir, and these u \~',
relations usually may be established. 11 \ 1\ I';""5D! \ 5320. -c- 5327
Conversely, the effects of reservoir ~.l. MC \5:1)3 .", .
geometry, capillarity, and rock type 10
-II MB \~" i"'-.
on the limiting permeability are diffi-
cult to predict. Further, porosity can
8
~~ ~ '"
---
o
s.,,-
FIG. 13-THEORETICAL CAPILLARY
be measured in the wellbore by pres-
ent methods of porosity logging, par- 4
T'-.. . . . . . . . . . . ,. ". . . ~-
PRESSURE CURVES. ticularly by the neutron method, 2 1 MA ==
while rock type may be obtained o ROCK TYPE
10 20 l:) 40 50 60 70 80 90 100
porosIties involved. As an example, from cuttings data. As a result, cor-
refer to any of the estimated con- ESTIMATED CONNATE WATER (%)
ing operations are necessary only on
nate water-porosity correlations in- FIG. 14-EFFECT OF RESERVOIR CAPILLARY
scattered key wells. CONDITIONS ON CUT-OFF POROSITY.
cluded in this report. As mentioned Des Moines limestone, Shell-Harkey No. A-l,

Downloaded from http://onepetro.org/trans/article-pdf/210/01/252/2176489/spe-710-g.pdf by guest on 27 February 2024

previously, water saturations increase CONCLUSION
Toenail Strawn field, Schleicher County, Tex.
sharply as the lower porosities de- REFERENCES
crease. The porosity at which each It may be concluded from the
curve reflects a sharp increase in presented data that oil in place may 1. Hassler, G. L., Brunner, E., and
Deahl, T. J.: "The Role of Capillar-
water saturation is taken as the cut- be delineated through a method ity in Oil Production," Trans. AIME,
off porosity for its associated capil- which is based on capillary pressure (1944), 155, 153.
lary pressure. The curve from which as the rock property that correlates 2. Law, J.: "A Statistical Approach to
it is chosen depends upon the capil- porosity to pore size distribution. the Interstitial Heterogeneity of Sand
Reservoirs," Trans. AIME, (1944),
lary conditions in the reservoir. The This is a practical method since the 155, 202.
value chosen then may be combined oil in place value of a reservoir may 3. Bulnes, A. c.: "An Application of
with either the core or log porosity be estimated from scattered, cored Statistical Method~ to Core Analysis
profile in order to delineate net pro- wells which are used to establish Data of Dolomite Reservoirs," Trans.
AfME, (1946), 165, 223.
ductive thickness by selecting only these relationships. Capillary effects 4. Chalkley, H. W., Cornfield, J., and
those porosities that are larger than then are combined with cuttings data Park, H.: "A Method for Estimating
the cut-off value. for rock type, and neutron log data Volume-Surface Ratio," Science,
for porosity in situ. Furthermere, (1949), no, 295.
Core porosities lower than this cut- 5. Chin, W. S., and Rose, W.: "Exami-
off value may be assigned fracture capillary pressure data may be used nation of the Components of Lime-
or vuggy void space by comparing to estimate the connate water of stone Porosity by the Chalkley Meth-
large-core porosity data with non- many carbonate reservoirs based Oil od," Bull. AAPG, (1951), 35, 615.
core, rock type, logging, and oil- 6. Rose, W.: "Porosity of Carbonate
vuggy small-core porosities for each Rock and Its Bearing on Reserves
foot involved. Although this void water contact data. Also, capillary and Recovery," Proc. Second Oil
space should not contain connate pressure data may be used to deter- Rec. Con£., TPRC, (1951) 51.
water if a few feet above the free mine if water production is indige- 7. Archie, G. E.: "Introduction to
Petrophysics of Reservoir Rocks,"
water level, its volume usually will nous or extraneous to the producing Bull. AAPG, (May, 1950), 34, 5.
be negligible compared to the oil held interval, or if the oil-water contact 8. Archie, G. E.: "Classification of Car-
by the porosities greater than the is being approached. bonate Reservoir Rocks and Petro-
cut-off value. physical Considerations," Bull.
AAPG, (Feb., 1952), 36, 2.
An example of varying cut-off ACKNOWLEDGMENTS 9. Stewart, C. R., and Spurlock, J. W.:
porosity, depending upon capillarity, "How to Analyze Large Core
is shown in Fig. 14. Where its value The authors wish to thank Shell Samples," Oil & Gas Jour., (Sept.
IS, 1952), 89.
would be about 5 to 6 per cent at a Oil Co. for permission to publish 10. Purcell, W. R.: "Capillary Pressures
large distance above the water level this paper. They are particularly in- -Their Measurement Using Mercury
(62-atm curve), it would increase debted to members of the Midland and the Calculation of Permeability
Therefrom," Trans. AIMF:, (1949),
to 10 to 12 per cent near the water Area Production Laboratory for the 186, 39.
level (5-atm curve). excellence of their work that is the 11. Stewart, C. R., Craig, F. F., Jr., and
In summary, the ideal method for foundation for deriving this method Morse, R. A.: "Determination of
Limestone Performance Characteris-
selecting net pay is to base it on the of reservoir analysis in carbonate tics by Model Flow Test," Trans.
actual forces holding connate water rocks. The suggestions and com- AIME, (1953), 198, 93.
in place; in lieu, capillary pressure ments of C. L. Blackburn, G. E. 12. Fatt, I.: "The Network Model of
Porous Media," Trans. AIME, (1956)
data by mercury injection are recom- Archie, C. G. Cooper, and E. D. 207, 144.
mended for determining this parame- Manly of Shell Oil Co. and others 13. Whiting, R. L., and Guerrero, E. T.:
"Capillary Properties of Some Lime-
ter. These capillary data are shown were invaluable for the successful stone Cores," Oil & Gas Jour., (July
to be related to rock type and po- completion of this paper. 26, 1951), 272. ***

259 VOL. 210, 1951