Chapter 6
◆
ABSTRACT
Petrofacies analysis is defined as the characterization and classification
of pore types and fluid saturations as revealed by petrophysical measure-
ments of a reservoir. The word “petrofacies” makes an explicit link
between petroleum engineers’ concerns with pore characteristics as
arbiters of production performance and the facies paradigm of geologists
as a methodology for genetic understanding and prediction. In petrofacies
analysis, the porosity and resistivity axes of the classical Pickett plot are
used to map water saturation, bulk volume water, and estimated perme-
ability, as well as capillary pressure information where it is available.
When data points are connected in order of depth within a reservoir, the
characteristic patterns reflect reservoir rock character and its interplay with
the hydrocarbon column. A third variable can be presented at each point
on the crossplot by assigning a color scale that is based on other well logs,
often gamma ray or photoelectric effect, or other derived variables.
Contrasts between reservoir pore types and fluid saturations are reflected
in changing patterns on the crossplot and can help discriminate and char-
acterize reservoir heterogeneity.
Many hundreds of analyses of well logs facilitated by spreadsheet and
object-oriented programming have provided the means to distinguish pat-
terns typical of certain complex pore types (size and connectedness) for
sandstones and carbonate reservoirs, occurrences of irreducible water satu-
ration, and presence of transition zones. The result has been an improved
means to evaluate potential production, such as bypassed pay behind pipe
and in old exploration wells, or to assess zonation and continuity of the
reservoir.
73
74 Watney et al.
Kansas
14 13
16 15 ARNOLD 13-1
17
14-1 KENDRICK
ARNOLD 1
FRETZ 16-2
SMITH TRUST 22-1 23-1 KENDRICK
23 24
21 22 IRWIN TRUST
20
22-2 SANTA FE
22-1 SANTA FE 22-1 KENDRICK
21-1 SANTA FE
1 mile
Figure 1. (a) Index map of Arroyo field identifying cross section (shown in Figure 2) and with well names and
distribution.
up to 20% and averages 14%. All positions of the sand- of this field on the Kansas Geological Survey’s Digital
stone have been perforated in the field, with some Petroleum Atlas. This atlas is located on the Survey’s
wells reported as only gross intervals. Initial reservoir Internet Home Page (http://crude2.kgs.ukans.edu/
pressure was 1434 psi. DPA/Arroyo/arroyoMain.html).
The upper portion of the sandstone has produced The lenticular lower Morrowan sandstone is com-
only natural gas, and the lowest portion has produced prised of a series of upward-coarsening, marginal
significant amounts of both oil and gas. No water has marine shoreface deposits that are mostly confined to
been produced in any of the wells. Also, no oil-water within an 0.8 km (0.5 mi) wide meandering valley up to
contact has been recognized. The reservoir drive 48 m (157 ft) deep (Figure 1). The sandstone was previ-
appears to be gas expansion. ously correlated and subdivided into five separate
A considerable amount of supporting data on sandstone-dominated genetic units (1, 3, 5, 9, 11) using
Arroyo field, including digital well logs, completion gamma ray, porosity, and resistivity logs and one spec-
reports, and interpretive maps, cross sections, and syn- tral gamma ray log. Each genetic unit is delineated by
thetic seismograms, are included in a digital publication bounding surfaces usually characterized by abrupt
76 Watney et al.
Figure 1. (b) An isopachous map of the lower Morrowan interval (top middle Morrowan limestone to top
Mississippian). Contour interval is 25 ft.
changes in lithofacies. The surfaces either represent fluid producibility may be deduced by the use of pat-
subaerial exposure or flooding surfaces, or both. Each tern recognition informed by basic reservoir engineer-
genetic unit is believed to represent temporally distinct ing principles. A template Pickett plot is shown in
episodic deposition. Only several of the genetic units Figure 3 for the upper Morrowan in the Arroyo field.
are developed at any particular location in the field A water line (Ro) expresses the theoretical resistivity-
(Figure 2). porosity coordinates of all zones that are completely
The sandstones were deposited in a meandering val- saturated with water. The water line is established by
ley system during overall rise in sea level. Maps of each the first Archie equation that links the formation fac-
genetic unit record episodes of infilling of this valley, tor, F, to the ratio of the resistivity of the completely
each unit with varying geometries and sand abundance water-saturated rock, Ro, to the resistivity of the for-
and quality. The lowest sandstones are more limited in mation water, Rw, to the porosity of the rock:
distribution, filling only the lowest (deepest) portions of
the valley, while the higher sandstones locally extend
F = Ro Rw = a φ m
beyond the confines of the valley. For these reasons, the
stratigraphic distribution was believed to be a control-
ling factor on flow unit definition. using an Arroyo field formation water resistivity, Rw,
of 0.04 ohm-m and Archie parameter values of a = 1
and m = 1.8, which express pore geometry in the Mor-
METHODOLOGY rowan sandstone. Contours for different values of
water saturation parallel the water line, with spacing
Volumetric properties of pore space and fluid satu-
determined by the saturation exponent, n (generally
ration can be calculated from porosity logs (density, with a value of about 2 in water-wet rocks) in the sec-
neutron, or sonic) and resistivity logs using the stan- ond Archie equation:
dard Archie equations. When plotted on a double-
logarithmic plot of porosity versus resistivity (a
Pickett plot), additional information on pore size and I = Rt Ro = 1 Sw n
Petrofacies Analysis—A Petrophysical Tool for Geologic/Engineering Reservoir Characterization 77
Figure 2. West-to-east stratigraphic well log cross section through Arroyo field containing Lauman 28-1, Santa
Fe 21-1, Santa Fe 22-1, Santa Fe 22-2, and Kendrick 22-1 wells. Datum of section is middle Morrowan lime-
stone. Correlated stratigraphic intervals are correlated through the lower Morrowan sandstone interval. Line
of cross section is shown on Figure 1a.
where I is the resistivity index and R t is the actual least productive at the bottom. Notice that the bottom
resistivity of the rock. Lines can also be drawn on the example was not completed for production, but aban-
plot that are contours of bulk volume water (BVW), doned because a DST (drill-stem test) yielded nothing
where water content is expressed as a proportion of but saltwater. The associated Pickett plot shows a
the total rock, rather than in terms of the pore space as rather ragged scatter of mostly low-porosity zones
is the case with saturation. with high water saturations that probably reflect resid-
The disposition of crossplotted zones with respect ual hydrocarbon saturations. This pattern is common
to the log axes of resistivity-porosity and the com- at the margins of fields, as is this well, and contrasts
puted reference axes of water saturation and bulk vol-
ume water (BVW) gives useful clues on both pore type
and producibility. These properties can be seen when
BVW
relating Pickett plots to production histories (Figure 4) 100
from some example wells in the Arroyo field. Notice 1 10 100
how overall well performance is determined to a large Rw=0.04
extent by higher porosities and lower water satura- a=1
Sw 0.04
tions; however, the location of the data-cloud with =50 0.03
m=1.8 Sw= %
respect to the BVW contours reflects the pore size and 100
%
likely water-cut. Lower values of BVW are matched n=2 10
with coarser pores; higher values of BVW are linked
Porosity%
Well #20686
100K Arroyo field, Stanton
10 rate County, Kansas. Each Pickett
10K plot is identified with a five-
digit well number. The corre-
1K sponding well names are as
Perforated follows: Well #20686: Huber
28 ft
1 10 #10-1 Cockreham, SE NE SE
100 0 Sec. 10-T29S-R41W; Well
MCF
1 10 100 #20649: Huber #26-2 Pro
Porosity%
cumulative
1000K Farms, SW NW NW Sec. 26-
T29S-R41W; Well #20671:
Well #20649
100K Huber #23-2 Kendrick, C NW
10
Sec. 23-T29S-R41W; Well
10K #20692: Petroleum Inc. #1-29
rate Spikes, NENESE Sec. 29-
1K T29S-R41W.
Perforated
33 ft 10
1
100 0
1 10 100 MCF
Porosity%
1000K
cumulative
Well #20671
100K
10
10K
rate
1K
Perforated
1 16 ft 10
100 0 1 2 3 40
1 10 100 year
Porosity%
Well #20692
No production
10 DST: 1196 feet
salt water
1
1 10 100
Resistivity, ohm-m
with an idealized situation far away from fields where plot. This, in turn, helps to correlate the geology to
zones are completely water-saturated and form a fluid-related parameters and to delineate specific
trend on a Pickett plot that conforms closely to the changes in fluids and variations in the pores between
water line (Ro). the different wells.
The Pickett (porosity-resistivity) crossplots are the The definition of flow units might be refined to
fundamental components in the petrofacies analysis. include regions of similar or related BVW and pore
The connection of data points by depth and the abil- type using the petrofacies analysis. Often, permeabil-
ity to annotate the data points with a third variable ity data are either lacking or are limited to averaging
help establish relationships between the petrophysi- from core-log porosity and permeability correlations.
cal response and the geology—lithologies, strati- In these cases, assessment of pore type using petrofa-
graphic units, and structure; i.e., the petrofacies. cies analysis may help to provide novel constraints to
Template lines identify minimum BVW and associ- flow units lacking other substantial data. Of course,
ated water saturations and porosities on the Pickett production and transient test data and geochemical
Petrofacies Analysis—A Petrophysical Tool for Geologic/Engineering Reservoir Characterization 79
Figure 5. Pickett plot of lower Morrowan sandstone in the Santa Fe 22-1 annotated with gamma ray. Note
that this Pickett plot has contours of bulk volume water and permeability, the latter estimated from Timur
equation.
tests are necessary to more definitively constrain the and (4) integrate this information to define flow
definition of flow units. units by comparison with stratigraphic zonation,
A west-to-east cross section was chosen to further Pickett plots, color log cross sections, cluster analy-
characterize the sandstone reservoir using the petrofa- sis, and well productivities.
cies analysis approach (Figure 2). The cross section
crosses the valley in two places, separated by an inter-
vening high area.
RESULTS
The questions addressed in this analysis include the The digital data from five well logs comprising a
following: (1) Is additional evidence available to con- west-to-east cross section in Arroyo field were examined
firm or reject the continuity of sandstones across the using petrofacies analysis. The objective was to compare
intervening high region residing between the valleys? well data on the cross section to test for evidence of lat-
(2) What is the evidence of vertical conformance and eral continuity. Correlations shown in Figure 2 suggest
lateral continuity? (3) How do properties of the sand- that most of the units are continuous. Units that are not
stones compare on either side of the valley? (4) Can the shown as continuous may also be connected from out-
definition of flow units be improved? (5) How do the side of the plane of the cross section.
flow units compare with the detailed stratigraphic
subdivisions? Pickett Plots
In addition to the stratigraphic analysis, the pro-
cedure included four operations: (1) construct Pick- LAS (log ASCII standard) digital well log files were
ett plots for each of the wells on the cross section, (2) read into an Excel-Visual Basic® program called PfEF-
perform a cluster analysis of basic petrophysical FER™ to generate the Pickett crossplots. The initial
data to independently define similar reservoir prop- Pickett crossplots provide a visual differentiation of
erties, (3) prepare a series of color cross sections of the variation in the porosity, resistivity, water satura-
selected petrophysical variables with datums on sea tion, and bulk volume water. Permeability lines are
level elevation and a stratigraphic marker (middle annotated on the crossplots, estimated using the
Morrowan limestone located above the sandstone), empirical relationship between water saturation and
80 Watney et al.
Sante Fe 22-1
BVW=0.05
BVW=0.04
BVW=0.03
Sw=20% 1 BVW=0.02
Sw=40%
Sw=60%
Sw=80%
Sw=100%
Stratigraphic
POROSITY, %
0.1 Sequence
11
9
5
3
1
0.01
1 10 100
RESISTIVITY, Ohm-m
Figure 6. Pickett plot of lower Morrowan sandstone in the Santa Fe 22-1 annotated with stratigraphic units
shown in type log in Figure 2.
5380 5380
5380 5380 5380 5380 5380 5380 5380
5400 5400
5400 5400 5400 5400 5400 5400 5400
5420 5420
5420 5420 5420 5420 5420 5420 5420
Figure 7. Display of well log suite from the Santa Fe 22-1 well accompanied by derived information including
water saturation (Sw), bulk volume water (BVW), permeability derived from Timur equation (Timur k), strati-
graphic units, and amalgamation groups from cluster analysis.
porosity developed by Timur. The relationship holds gamma ray values are located on the left side of the
for clean sandstones when water saturations are at crossplot at BVWs in excess of 0.06 (Figure 5). This
irreducible values. We believe that to be the case here. location presumably represents more shaly and finer
The crossplot (Figure 5) with points annotated with pores. The permeability lines are not applicable to
gamma ray values indicates that points with higher these points.
Petrofacies Analysis—A Petrophysical Tool for Geologic/Engineering Reservoir Characterization 81
Sante Fe 22-1
BVW=0.05
BVW=0.04
BVW=0.03
Sw=20% 1 BVW=0.02
Sw=40%
Sw=60%
Sw=80%
Sw=100%
POROSITY, %
0.1
Cluster Analysis
5
4
3
2
1
0.01
1 10 100
RESISTIVITY, Ohm-m
Figure 8. Pickett plot of lower Morrowan sandstone in the Santa Fe 22-1 annotated with amalgamation groups
from cluster analysis.
The reservoir has no water leg, and no water has the similarity among petrophysical data. The
been produced. The points are annotated by strati- method provides a consistent automated treatment
graphic interval and form rather tight clusters or of the data to aid in comparing considerable amounts
bands for each stratigraphic unit (Figure 6). The clus- of data among the zones and wells. Ward’s Method
tering of points in distinct stratigraphic intervals at was selected as the clustering technique. The method
lower water saturations suggests that these zones are consists of a series of clustering steps that begins
near their respective minimum BVWs and closely with “t” clusters, each containing one object. The
correspond to particular stratigraphic zones. Also, clustering ends with one cluster containing all
the bands parallel water saturation lines. This pat- objects. At each step, a merger of two clusters is
tern is ascribed to changing minimum BVW and made that results in the smallest increase in the vari-
pore size within a zone, which has implications to ance (Romesburg, 1984).
fluid flow. If any portions of the reservoir were in a The petrophysical variables included in the cluster
gas or oil/water transition zone, the bands of points analysis are gamma ray, deep induction resistivity, Pe
may have more likely paralleled porosity lines, if the (photoelectric index), S w (water saturation), BVW
pore type were not changing; however, this is not (bulk volume water), and apparent permeability using
seen and no wells have experienced any water-cut the Timur equation
oil or gas production. The variations suggest possi-
ble changes in pore type and evidence for reservoir k a = 1 × 10 4 ∗ φ 4.5 Sw 2
continuity or lack thereof.
defined in Figure 2.
-1930 11
-1930.5 11
-1931 11
-1931.5 11
-1932 11
-1932.5 11
-1933 11
-1933.5 11
11
-1936 11
-1936.5 11
-1937 11
-1937.5 11
-1938 11
9
-1962.5 9 11 11 11
-1963 9 11 11 11
-1963.5 9 11 11 11
-1964 9 11 11 11
-1964.5 9 11 11 11
-1965 9 11 11 11
-1965.5 9 11 11 11
-1966 9 11 11 11
-1966.5 9 11 11 11 11
-1967 9 11 11 11 11
-1967.5 9 11 11 11 11
-1968 9 11 11 11 11
-1968.5 9 11 11 11 11
-1969 9 11 11 11 11
-1969.5 9 11 11 11 11
-1970 9 11 11 11 11
11
-1970.5 9 11 11 11 11
-1971 9 11 11 11 11
-1971.5 9 11 11 11 11
-1972 9 11 11 11 11
-1972.5 9 11 11 11 11
-1973 9 11 11 11 11
-1973.5 9 11 11 11 11
-1974 9 11 11 11 11
-1974.5 9 11 11 11 11
-1975 9 11 11 11 11 11
-1975.5 9 9 11 11 11
-1976 9 9 9 11 11 11
-1976.5 5 9 11 11 11
-1977 5 9 9 11 11 11
-1977.5 5 9 11 11 11
-1978 5 5 9 11 11 11 11
-1978.5 5 9 11 11 9
-1979 5 9 11 11 9
-1979.5 5 9 11 11 9
-1980 5 9 11 11 11 9 9
-1980.5 5 9 9 11 9
-1981 5 9 9 11 9
-1981.5 5 9 9 11 9
-1982 5 9 9 9 11 9
-1982.5 5 9 9 11 9
-1983 5 9 9 11 9
-1983.5 5 9 9 11 9
-1984 5 9 9 11 11 9
-1984.5 5 9 9 9 9
-1985 5 9 9 9 9
-1985.5 5 9 9 9 9
-1986 5 9 9 9 9 9
-1986.5 5 9 9 9 9
-1987 5 9 9 9 9
-1987.5 5 9 9 9 9
-1988 5 9 9 9 9
5
-1988.5 5 9 9 9 9
-1989 5 9 9 9 9
-1989.5 5 9 9 9 9
-1990 5 9 9 9 9
-1990.5 5 9 9 9 9
-1991 5 9 9 9 9
-1991.5 5 9 9 9 9
-1992 5 9 9 9 9 9
-1992.5 5 5 9 9 9
-1993 5 5 9 9 9
9
-1993.5 5 5 9 9 9
-1994 5 5 5 9 9 9
-1994.5 5 5 9 9 9
-1995 5 5 9 9 9
-1995.5 5 5 9 9 9
-1996 5 5 9 9 9
-1996.5 5 5 9 9 9
-1997 5 5 9 9 9
-1997.5 5 5 9 9 9
-1998 5 5 9 9 9
-1998.5 5 5 9 9 9
-1999 5 5 9 9 9
-1999.5 5 5 9 9 9
-2000 5 5 9 9 9
-2000.5 5 5 9 9 9
-2001 5 5 9 9 9
-2001.5 5 5 9 9 9
-2002 5 5 9 9 9
-2002.5 5 5 9 9 9
-2003 5 5 5 9 9 9
-2003.5 3 5 9 9 9
-2004 3 5 9 9 9 9
-2004.5 3 5 9 5 9
-2005 3 3 5 9 5 9
-2005.5 3 5 9 5 9
-2006 3 5 9 5 5 9
-2006.5 3 5 9 5 9
-2007 3 5 9 5 9
-2007.5 3 5 9 5 9
-2008 3 5 9 5 9
-2008.5 3 5 9 5 9
-2009 3 5 9 5 9
-2009.5 3 5 9 5 9
-2010 3 5 9 5 9 9
-2010.5 3 5 9 5 5
-2011 3 5 9 5 5
-2011.5 3 5 9 5 5
-2012 3 5 9 9 5 5 5
-2012.5 3 5 5 5 5
-2013 3 5 5 5 5
-2013.5 3 5 5 5 5
-2014 3 5 5 5 5 5
-2014.5 3 5 5 5 5
-2015 3 5 5 5 5
-2015.5 3 5 5 5 5
-2016 3 5 5 5 5
3
-2016.5 3 5 5 5 5
-2017 3 5 5 5 5
-2017.5 3 5 5 5 5
-2018 3 5 5 5 5
-2018.5 3 5 5 5 5
-2019 3 5 5 5 5
-2019.5 3 5 5 5 5
5
-2020 3 5 5 5 5
-2020.5 3 5 5 5 5
-2021 3 5 5 5 5
-2021.5 3 5 5 5 5
-2022 3 5 5 5 5
-2022.5 3 5 5 5 5
-2023 3 5 5 5 5
-2023.5 3 5 5 5 5
-2024 3 5 5 5 5 5
-2024.5 3 5 5 3 5
-2025 3 5 5 3 5
-2025.5 3 5 5 3 5
-2026 3 5 5 5 3 3 5
-2026.5 3 3 5 3 5
-2027 3 3 5 3 5
-2027.5 3 3 5 3 5
-2028 3 3 3 5 3 5
-2028.5 3 3 5 3 5
-2029 3 3 5 3 5
-2029.5 3 3 5 3 5
-2030 3 3 5 3 5
-2030.5 3 3 5 3 5
-2031 3 3 5 3 5
-2031.5 3 3 5 3 5
-2032 3 3 5 3 5
-2032.5 3 3 5 3 5
-2033 3 3 5 3 5
-2033.5 3 3 5 3 5
-2034 3 3 5 3 5
-2034.5 3 3 5 3 5
-2035 3 3 5 3 5
-2035.5 3 3 5 3 5
-2036 3 3 5 3 5
-2036.5 3 3 5 3 5
-2037 3 3 5 3 5
-2037.5 3 3 5 3 5
-2038 3 3 5 3 5
-2038.5 3 3 5 3 5
-2039 3 3 5 3 5
-2039.5 3 3 5 3 5
-2040 3 3 5 3 5 5
-2040.5 3 3 5 3 3
-2041 3 3 5 3 3
-2041.5 3 3 5 3 3
-2042 3 3 5 3 3 3
-2042.5 3 3 5 3 3
-2043 3 3 5 3 3
-2043.5 3 3 5 3 3
-2044 3 3 3 5 3 3
-2044.5 3 1 5 3 3
-2045 3 1 5 3 3
-2045.5 3 1 5 3 3
-2046 3 3 1 1 5 3 3
-2046.5 1 1 5 3 3
-2047 1 1 5 3 3
-2047.5 1 1 5 3 3
-2048 1 1 1 5 3 3
-2048.5 1 1 5 3 3
-2049 1 1 5 3 3
-2049.5 1 1 5 3 3
-2050 1 1 5 3 3
-2050.5 1 1 0 5 3 3
-2051 1 0 5 3 3
-2051.5 1 0 5 3 3
3
-2052 1 0 5 5 3 3
-2052.5 1 0 0 3 3 3
-2053 1 0 3 3 3
-2053.5 1 0 3 3 3
-2054 1 0 3 3 3 3
-2054.5 1 0 3 3 3
-2055 1 0 3 3 3
-2055.5 1 0 3 3 3
-2056 1 0 3 3 3
-2056.5 1 0 3 3 3
-2057 1 0 3 3 3
-2057.5 1 0 3 3 3
-2058 1 0 3 3 3
-2058.5 1 0 3 3 3
-2059 1 0 3 3 3
-2059.5 1 0 3 3 3
1
-2060 1 0 3 3 3
-2060.5 1 0 3 3 3
-2061 1 0 3 3 3
-2061.5 1 0 3 3 3
-2062 1 0 3 3 3
-2062.5 1 3 3 3
-2063 1 3 3 3
-2063.5 1 3 3 3
-2064 1 3 3 3
-2064.5 1 3 3 3
-2065 1 3 3 3
-2065.5 1 3 3 3
-2066 1 3 3 3
-2066.5 1 3 3 3
-2067 1 3 3 3
-2067.5 1 3 3 3
-2068 1 3 3 3
-2068.5 1 3 3 3
-2069 1 3 3 3
-2069.5 1 3 3 3
-2070 1 3 3 3
-2070.5 1 3 3 3
-2071 1 3 3 3
-2071.5 1 3 3 3
-2072 1 3 3 3 3
-2072.5 1 3 0 3
-2073 1 3 0 3
-2073.5 1 3 0 3
-2074 1 1 3 0 0 3
-2074.5 3 0 3
-2075 3 0 3
-2075.5 3 0 3
-2076 3 0 3
-2076.5 3 0 3
-2077 3 0 3
-2077.5 3 0 3
-2078 3 0 3
-2078.5 3 0 3
-2079 3 0 3
-2079.5 3 0 3
-2080 3 3 3
-2080.5 3 1
-2081 3 1
-2081.5 3 1
-2082 3 1
-2082.5 3 1
-2083 3 3 1
-2083.5 1 1
-2084 1 1
-2084.5 1 1
-2085 1 1 1
-2085.5 1 1
-2086 1 1
-2086.5 1 1
-2087 1 1
-2087.5 1 1
-2088 1 1
-2088.5 1 1
-2089 1 1
-2089.5 1 1
-2090 1 1
-2090.5 1 1
-2091 1 1
-2091.5 1 1
-2092 1 1 1
-2092.5 1
-2093 1
-2093.5 1
1
-2094 1
-2094.5 1
-2095 1
-2095.5 1
-2096 1
-2096.5 1
-2097 1
-2097.5 1
-2098 1
-2098.5 1
-2099 1
-2099.5 1
-2100 1
-2100.5 1
-2101 1
-2101.5 1
-2102 1
-2102.5 1
-2103 1
-2103.5 1
-2104 1
-2104.5 1
-2105 1
-2105.5 1
-2106 1
-2106.5 1
-2107 1
-2107.5 1
-2108 1
-2108.5 1
-2109 1
-2109.5 1
-2110 1
-2110.5 1
-2111 1
-2111.5 1
-2112 1
-2112.5 1
-2113 1
-2113.5 1
-2114 1
-2114.5 1
-2115 1
-2115.5 1
-2116 1
-2116.5 1
-2117 1
-2117.5 1
-2118 1
-2118.5 1
-2119 1
-2119.5 1
-2120 1
-2120.5 1
-2121 1
-2121.5 1
-2122 1
-2122.5 1
-2123 1
-2123.5 1
-2124 1
-2124.5 1
-2125 1
-2125.5 1
-2126 1
-2126.5 1
-2127 1
-2127.5 1
-2128 1
-2128.5 1
-2129 1
-2129.5 1
-2130 1
-2130.5 1
-2131 1
-2131.5 1
-2132 1
-2132.5 1
-2133 1
-2133.5 1
-2134 1 1
-2134.5 0
Six separate groups of points were selected from separation of groups at this level. Third, the number
the cluster analysis of each well. Several criteria is comparable to the stratigraphic divisions and
were used to determine this number. First, the num- might show useful groupings and comparison.
ber is not large enough to produce too many groups, The assigned groupings derived from cluster
which could complicate reservoir modeling. Second, analysis were first compared by depth with the
the cluster dendrogram for each well showed good petrophysical data and stratigraphic zonation. The
Petrofacies Analysis—A Petrophysical Tool for Geologic/Engineering Reservoir Characterization 83
boundaries between the stratigraphic intervals and units, e.g., groups 4 and 5 in stratigraphic sequences
the cluster assigned groupings generally coincide. 5 and 9. It would be anticipated that similar sandstone
The clustering identified a moderate amount of properties would transcend sandstone intervals.
smaller scale heterogeneity within each stratigraphic However, the general finding is that each stratigraphic
interval (Figure 7). This internal variation includes interval is dominated by only one or two assigned
having the same cluster group in different stratigraphic cluster groupings.
84 Watney et al.
Figure 10. East-to-west color-image cross section depicting gamma ray variation across Arroyo field. The
section includes same wells as in Figure 2. Wells are annotated with perforations as bar along right side of
each well. Section also shows correlations of stratigraphic sequences. Cross section is part of an Excel
spreadsheet and is at the resolution of the digitized data (0.5 ft in this example). Vertical scale bar shown on
this and ensuing structural sections. No horizontal scale (wells are equally spaced). (a) [above] Structural
version of cross section with a sea level datum.
Petrofacies Analysis—A Petrophysical Tool for Geologic/Engineering Reservoir Characterization 85
The posting of assigned cluster groups as an conformance. The question remains as to the extent
attribute on the Pickett plot further indicates a close of lateral continuity. Flow units are not fieldwide in
correspondence between stratigraphic units and extent, but are anticipated to be correlatable to some
assigned cluster grouping (as in Figure 8). The clus- degree. This continuity is ultimately established
ter analysis can be adapted in the spreadsheet envi- using petrophysical data, fluid recovery, pressure,
ronment to help facilitate consistent, rapid and fluid chemistry. The suites of petrophysical vari-
assignment of cluster groups and further aid in flow ables including BVW can be used to evaluate lateral
unit assessment. continuity. Continuous trends or constancy of prop-
erties of the sandstone and correlations with struc-
tural elevation suggest possible fluid continuity in
Color Cross Sections
the reservoir.
Computer-assisted generation of color cross sec-
In general, flow units are assigned to zones in the tions based on original digital well log sampling of
reservoir with similar permeability and porosity, 0.16 m (0.5 ft) provides the means to observe and
and that also exhibit lateral continuity. Flow units evaluate detailed subtle changes in reservoir charac-
are inferred to control fluid flow, and confirmation ter and substantially assists in assessing continuity
was sought to substantiate these units. Petrophysi- and assigning flow units. The cross sections are gen-
cal variation within individual well profiles has erated with an elevation (subsea) or stratigraphic
been described up to this point, focused on vertical datum.
86 Watney et al.
Cross sections for key petrophysical parameters induction resistivity (Figures 9–11). Each petrophysical
were generated, including gamma ray, permeability variable is presented as a structural and stratigraphic
calculated with the Timur equation (apparent perme- cross section placed side by side, the latter with a datum
ability filtered on gamma ray and neutron-density shale at the top of the middle Morrowan limestone. Strati-
indicators), Pe (photoelectric effect) from the lithoden- graphic units are identified and correlated. Perforated
sity log, porosity, water saturation, BVW, and deep intervals are shown alongside each well profile.
Petrofacies Analysis—A Petrophysical Tool for Geologic/Engineering Reservoir Characterization 87
The five wells in the cross section are perforated appear to delineate most of the petrophysical variation
in two distinct intervals, a lower interval restricted except for several possible re-correlations. These re-
to stratigraphic unit #1 in the Lauman 28-1 and correlations are based on further analysis.
Kendrick 22-1. Also, the lowest part of unit #3 in the Porosity varies from 15 to 20% in the Lauman well to
Santa Fe 2-21 is suggested to be part of the lower 0 to 8% in the Kendrick (Figure 11). Apparent perme-
interval and may possibly be recorrelated with unit ability calculated from the Timur equation and filtered
#1 (Figure 9). on gamma ray and neutron shale indicators shows con-
The lower interval produces significant amounts of siderable changes on the cross section (Figure 12). The
oil and gas. The upper perforated interval includes permeability and porosity are both higher on the west.
stratigraphic units #3, #5, #9, and #11. The upper inter- Permeability ranges between 10 and 100 millidarcies
val produces natural gas and minor amounts of oil (md) in the Lauman 28-1 on the west side to between
from the Santa Fe 2-11 and Santa Fe 2-22. This differ- 0.1 to 1 md in the Kendrick 22-1 on the east side.
ence suggests that the reservoirs are separate. The High permeability and porosity correlate well
lower and upper sandstones are isolated by a promi- with a trend of increased natural gas production and
nent shaly interval, according to the gamma ray and decreased oil production in the Lauman 28-1, with
photoelectric logs. 2.5 Gcf of gas and 75,000 bbl of oil. In comparison, the
On closer inspection of the gamma ray cross section, Kendrick 22-1 well on the east and lowest side of the
the stratigraphic units can be distinguished with the cross section recovered less gas, 1 Gcf, but more oil,
help of the correlation lines; however, there is consider- over 120,000 bbl. Permeability varies considerably in
able variation in the internal properties of the strati- thin streaks near the base of the Kendrick 22-1 well.
graphic units (Figure 10). This variation persists in the Unit #1 is separated from the overlying sandstones
other parameters. The stratigraphic units generally by a thicker shaly interval.
88 Watney et al.
Unit #3 is also thick in the paleovalleys on either side but the mixed production is similar to the recoveries
of a central high. Santa Fe 2-21 is perforated in the basal noted in unit #1, suggesting that they are a common
part of a thin sandstone that is in close proximity to the reservoir in the deeper portions of the paleovalleys.
lower interval, unit #1. This zone in Santa Fe 2-21 has Production from perforations in the upper interval
produced 450 Mcf of gas and nearly 50,000 bbl of oil. in Santa Fe 21-1 and Santa Fe 22-2 is notably different.
Production values are less than Lauman and Kendrick, Santa Fe 21-1 has realized 570 Mcf of gas from the
Petrofacies Analysis—A Petrophysical Tool for Geologic/Engineering Reservoir Characterization 89
each stratigraphic unit. Cluster analysis provides a Kansas Technology Enterprise Corporation and the
consistent means to further delineate reservoir proper- industrial consortium, “Development and Demon-
ties. The boundaries of the clustered groups are com- stration of An Enhanced, Integrated Spreadsheet-
monly those of the stratigraphic units. The clustered based Well Log Analysis Software,” Subcontract No.
groups provide further subdivisions of the reservoir G4S60821 with BDM-Oklahoma and industry consor-
rock that could be used to classify finer scale flow units. tium. PfEFFER application and testing has been con-
Color cross sections further substantiate the use of the ducted with support from “Shaben Field—Class II
stratigraphic divisions as basic templates for distin- Field Demonstration Project,” Contract No. DE-FC22-
guishing flow units. The color cross sections are repre- 94PC14987, and “Digital Petroleum Atlas,” Contract
sentative of the original digitized well log data and No. DE-FG22-95BC14817, both supported by the
provide the means to precisely subdivide the strati- Department of Energy.
graphic units. Petrofacies analysis should prove useful
for evaluating improved petroleum recovery options. REFERENCES CITED