Marine and Petroleum Geology 73 (2016) 21e35

Contents lists available at ScienceDirect

Marine and Petroleum Geology

journal homepage: www.elsevier.com/locate/marpetgeo

Research paper

Applications of 3D seismic attribute analysis in hydrocarbon prospect

identification and evaluation: Verification and validation based on
fluvial palaeochannel cross-sectional geometry and sinuosity, Ness
County, Kansas, USA
A.E. Raef*, T.N. Meek, M.W. Totten
Geology Department, Kansas State University, 108 Thompson Hall, Manhattan, KS 66506, USA

a r t i c l e i n f o a b s t r a c t

Article history: Due to their high resolution and established success rates, the analysis of geometrical and stratigraphic
Received 13 August 2015 3D seismic reflection data attributes have become one of the most important tools in many hydrocarbon
Received in revised form exploration and development programs. In areas with high drilling risk as a result of the lack of spatial
19 January 2016
continuity and lithological variation of potential prospects, analysis of relevant seismic attributes is
Accepted 15 February 2016
Available online 18 February 2016
essential to successful placement of wells. This study of the York field, Ness County, Kansas, presents a
corroborative finding, based on the analysis of a set of 3D seismic attributes, revealing a meandering
fluvial channel system incised on the Mississippian stratigraphic unconformity north east of a previously
Keywords:
Palaeoriver
interpreted Weirman field palaeoshoreline and submarine channels. The York field prospect identifica-
Seismic attributes tion and appraisal program targeted the Cherokee sands where reservoir properties are favorable and
Lithofacies where it was thought that thick Cherokee sands correlate with structural Mississippian highs. We pro-
Seismic interpretation duce evidence that Cherokee sand-prone fluvial facies is mappable based on synergistic analysis of
River-channel seismic arrival time, amplitude, spectral decomposition, acoustic impedance, and waveform attributes.
Hydrocarbon Seismic amplitude highs and relative acoustic impedance lows showed moderate conformance with
Sand facies Mississippian-horizon time lows. Controls exerted by channel post-confluence water and sediments
supply provide corroborating hydraulic geometry aspects for the interpreted fluvial system.
Time-structural maps, in addition to time and horizon slices of several 3D seismic attributes including
amplitude, spectral decomposition components, and relative acoustic impedance all seem to indicate
that five previously drilled dry wells within the study area were outside the boundary of a meandering,
Cherokee fluvial system of potential reservoir quality. Additionally, comparisons of the results of this
research to previous studies conducted in the south west of Ness County have provided an opportunity to
support and contribute to, interpreted paleodepositional setting established by Raef et al,, 2015 where a
palaeoshorline was interpreted. The results of this study support a broadly NEeSW trending meandering
channel system with SE flow direction.
© 2016 Elsevier Ltd. All rights reserved.

1. Introduction
1.1. Background relevant studies
Reducing drilling risk when targeting thin, very often below
seismic temporal resolution, Cherokee group basal/fluvial sands
prospects of Kansas is an industry priority amid low-price com-
modity environment and the need to optimize placement of
* Corresponding author.
E-mail address: abraef@ksu.edu (A.E. Raef).
multiwall pads. Since their introduction in the 1970s, seismic at-
tributes have allowed geoscientists to break down and view seismic
traces in terms of unique, quantifiable components such as fre-
quency, phase, and amplitude (Mourning, 2014). Several authors
(e.g. Chopra and Marfurt, 2008; Suarez et al., 2008; and Verma
et al., 2009; Bahorich and Farmer, 1994) have reported on the use
of seismic attributes in the characterization of hydrocarbon reser-
voirs and their utilization in defining channel fill zones. A
comprehensive classification of seismic attributes and the different
modes of extraction and application is covered by Chen and Sydney
(1997). Imaging channels boundaries based on most-negative

http://dx.doi.org/10.1016/j.marpetgeo.2016.02.023
0264-8172/© 2016 Elsevier Ltd. All rights reserved.
22 A.E. Raef et al. / Marine and Petroleum Geology 73 (2016) 21e35
curvature was shown, by Suarez et al. (2008), to suffer from
acquisition footprints in a study to calibrate the response of various
attributes to a well-understood fluvial system.
The challenges of detecting and undertaking a seismic based
channel systems characterization calls upon adopting corrobo-
rating sources of independent information such as those based on
hydraulic geometry in response post-confluences changes in sedi-
ments and water supply. The York Lease in north east Ness county
(Fig. 1), Kansas, poses a drilling risk that has manifested itself when
five unsuccessful wells were drilled. This drilling risk is to a great
extent related to the targeted basal/fluvial Cherokee sands as very
narrow and thin, as well as highly discontinuous spatially. This
study will utilize several different seismic attributes, including
acoustic impedance, amplitude, spectral decomposition, two way
travel time (TWTT) and amplitude, in order to properly delineate
potential hydrocarbon reservoirs (e.g. meandering channel sands)
in an area of Ness County, Kansas, USA. To this end, we find it
essential to shed some light, in the following three paragraphs, on
previous efforts of Cherokee sand, for which palaeodepositional
setting was described by Walters et al., 1979 prospect identification
and evaluation in Ness County area.
In 2003, Coral Coast Petroleum drilled wildcat well, Keith #1,
targeting a Cherokee sandstone west of northeastern Ness County,
Kansas. This well was part of the Wierman field (see Fig. 1) and
produced 162 barrels before it was plugged as dry and abandoned
(Abbas, 2009). The well was close to a reservoir, but the operator
concluded that not enough sand development was present for
economic reserves, and that the well was on the edge of a channel.
The focal point of Abbas (2009), was essentially answering the
question of “what went wrong” with this particular prospect. In his
thesis, Abbas stated that personal communications with the well
operator revealed that the potential sandstone target was identified
primarily through the occurrence and tracking of a doublet signal
reflection found on a seismic section at the base of the Cherokee
formation, just above the top of the Mississippian formation (Fig. 2).
In order to assess the situation, Abbas used surface 3D seismic
reflection data, well logs, and drilling reports or documents related
to the Weiman field, and then created a workflow incorporating 3D
seismic attributes in an attempt to explain the results of the Keith
#1 well. Ultimately, despite the appearance of this doublet event
within the seismic data, analysis of relative acoustic impedance,
RMS amplitudes, average energy, and amplitude attenuation
Fig. 1. Location of Ness County within Kansas, and the location of the Wierman Field within Section 18 of Township 16 south, Range 22 west (modified after: Abbas, 2009).
indicated that Keith #1 and several surrounding dry holes all fell
within unfavorable areas. Similarly, Philip (2011), attempted to
further develop the use of post-stack 3D seismic attributes in the
Wierman field and apply them as part of an integrated workflow
incorporating well logs, core data, production data, and modern
depositional analogs. The attributes Philip selected for the study
area were: acoustic impedance, amplitude attenuation, RMS
amplitude, and spectral decomposition, as well as curvature and
coherence attributes. Many of these attributes have been similarly
utilized in this particular study, and more detailed explanations of
these attributes and their applications can be found later within the
body of this text.
This leads into some of the most recent work in the area, with
emphasis on Cherokee sands, by Raef et al. (2015); this study pre-
sents a synergistic approach integrating post-stack seismic attri-
butes (TWTT, amplitude, coherency, parallel-bedding indicator, and
curvature) and well-log facies analysis in order to understand the
development and depositional setting of the Cherokee sands of
Wierman Field in central Kansas. The ultimate conclusion of this
study is that the dimensions and spatial relationships of the
interpreted geobodies are in conformity with the modern shoreline
analogs of a barrier beach or strand-plain adjacent to an estuary.
Fig. 3 illustrates interpreted paleovalleys, structural closure, and the
interpreted paleoshoreline.
Raef et al. (2015), suggests that “higher rates of drilling success
in the basal Cherokee sands are attainable by:
a) focusing on the proximity of thicker sands, as evidenced by
amplitude brightening of seismic reflections close to the Cher-
okee basal reflections,
b) [considering] the proximity of identified amplitude anoma-
lies to the interpreted paleoshorline,
c) favoring locations of paleotopographic highs on the Missis-
sippian unconformity, when associated with brightening of
seismic amplitudes in the basal Cherokee reflections and
dimming amplitudes below the Cherokee bottom reflection,
d) [locating] structural closures, [which] in association with the
above factors, are of significant weight in reducing drilling risk.”
Thus, the focus of this new study is three-fold: [1] apply
methodologies guided by three previously mentioned studies to a
seismic attribute analysis being conducted in a nearby area of Ness
 A.E. Raef et al. / Marine and Petroleum Geology 73 (2016) 21e35 23

Fig. 2. Seismic doublet found at the top of the Mississippian formation (from Abbas, 2009).

Fig. 3. Seismic attributes; (a) time structural map of Mississippian-top seismic time horizon with interpreted paleovalley (arrows), (b) Cherokee-top time horizon with structural
closures around the producer well Squier 1e18 and a lack of such closure around Keith 1, interpreted tidal channel branching off towards shoreline (c) isochron (time thickness)
color map of Cherokee group and contours of the time structural map of Cherokee top seismic horizon, and (d) stratal amplitude time slice at 25 ms above the Missippian horizon
(flattened), an interpreted paleoshoreline at the time of Cherokee sand deposition is marked by a dashed line (from Raef et al., 2015).
24 A.E. Raef et al. / Marine and Petroleum Geology 73 (2016) 21e35

Fig. 4. Ness County and location of study area: Approximate study area (red box) relative to the previously mentioned Wierman field (green box). The well, Conrath 1e10, is
highlighted in yellow for reference (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article).

County and [2] validate results of interpreted palaeo-fluvial chan-

nel system based on sinuosity and aspects of post-confluence hy-
draulic geometry signatures, and [3] outline most prospective
area(s).

1.2. Study area

Data for this particular study primarily originates from sections

10 and 11 in Township 16 south, Range 22 west of Ness County,
which is located in the central-western part of the state of Kansas.
The eastern side of the county is situated along the western edge of
the Central Kansas uplift. The first oil well in Ness County was
drilled in 1922, and in 1929, the Aldrich #1 well was drilled on the
Beeler anticline, and produced 100 bpd at a depth of 4422 feet on
top of the Mississippian (Abbas, 2009). The majority of current oil
production in western Kansas originates from Mississippian lime-
stone; however, as was mentioned previously, production from the

Fig. 6. Major tectonic features and location of Ness county: location of study area in
Fig. 5. Stratigraphic column of the Ordovician through Pennsylvanian systems (from the north east of Ness County is in proximity to the drainage basin boundary along the
Ramaker, 2009). early Pennsylvanian crest of Central Kansas Uplift, Modified after Clark (1987).
 A.E. Raef et al. / Marine and Petroleum Geology 73 (2016) 21e35 25

Fig. 7. Surface P-wave seismic description: (a) seismic amplitude time section and Cherokee (blue) and Mississippian (green) horizons, and synthetic seismic overlay at Wyman
1e11, (b) amplitude spectrum, and (c) amplitude tuning effects and dominant seismic wavelet (For interpretation of the references to color in this figure legend, the reader is
referred to the web version of this article).
26 A.E. Raef et al. / Marine and Petroleum Geology 73 (2016) 21e35
Fig. 8. Time structural map of the Mississippian unconformity and location of all
drilled dry-hole wells (drilled on structural highs) and position of seismic section of
Fig. 6a above.
relatively younger Cherokee group sandstones is of particular
importance for this study. Fig. 4 illustrates the location of this study,
relative to the nearby Wierman field, which is the focus of Philip
(2011) and Raef et al. (2015).
1.3. Sequence stratigraphy and depositional history
The target lithologies for this attribute analysis are from within
the previously mentioned Cherokee shale, a group of Pennsyl-
vanian-age shales and limestones, with thin layers of sands
(Merriam, 1963). These sediments range in thickness from 5 to 200
feet and were deposited in a continental environment transitioning
into a marginally marine environment as the sea transgressed over
the Mississippian unconformity, out of the Hugoton Embayment of
the Anadarko basin, and onto the Central Kansas uplift (Cuzella,
1991).
Fig. 5 outlines the general stratigraphic succession of the Cher-
okee shale. Ness County itself is located on the northeastern shelf of
the Hugoton Embayment, where sediments of the Desmoinesian
Stage rest unconformably on Mississippian carbonates. As previ-
ously mentioned, these sediments consist primarily of interbedded
carbonate beds and thin shales. Closer to the uplift, fine-grained
clastics give way to coarser clastics (sandstones and conglomer-
ates) that parallel the ancient shorelines. By the end of Missourian
time, the Central Kansas uplift was inundated and covered with
marine sediments.
Shoestring sand lenses (known to operators in Kansas and
Oklahoma as the Bartlesville sand) occur in the lower part of the
Cherokee shale (Bass, 1936; Doveton, 2002). These slender Cher-
okee sands are associated with fluvially dominated delta systems
where the sediment entered the seaway (Brenner, 1989). It is worth
noting that many of the reservoir sandstones found within the
Cherokee group are only 10e15 ft (4.6 m) thick; highly variable in
size, shape, orientation, and composition; and are difficult to pre-
dict (Cuzella, 1991).
According to Clark (1987), fluvial drainage systems began to
develop in the Early Pennsylvanian around the flanks of the
Hugoton embayment and on the northern shelf of the Anadarko
basin (Fig. 6); sediment was eroded from the Central Kansas uplift
and transported into the area. Marine transgression into the area
during Morrowan time resulted in the deposition of coarse-grained
basal sandstones and marine shales. The focus of this study is the
basal fluvial to transitional marine facies overlying the Mis-
sissippianePennsylvanian unconformity. This facies in the subsur-
face of Ness county ranges from transitional shallow marine in the
south west (Raef et al., 2015) to fluvial facies to north east of the
county; the fluvial facies is north east of a palaeoshoreline that was
interpreted by Raef et al. (2015); the fluvial setting is in agreement
with both the incised valleys findings of Archer et al. (1995) and the
palaeo incised valley imaging of this study.
The Mississippi unconformity is a karstic surface of Mississip-
pian limestones, from which dolines, as deep as 164 ft (50 m) and
measuring 0.99mi (1.6 km) across, have been described in the study
area, as well as an extensive network of palaeovalleys; most of
these valleys display a northesouth trend. In the easternmost part
of Ness County NEeSW oriented valleys join into larger north-
esouth oriented valleys towards the west, and whereas the
palaeovalleys in easternmost Ness County are relatively narrow,
ranging from 0.5 mi (0.8 km) to 1.5 mi (2.4 km) with a depth
ranging from 9.8 ft (3 m) to 39.0 ft (12 m), the approximately
northesouth oriented valleys in the west are 0.5mi (0.8 km) to
2.5mi (4 km) wide and up to 92 ft (28 m) deep (Ramaker, 2009).
The Mississippian surface is overlain by cherty conglomeratic
breccia, glauconitic sandstone, and variegated silty mudstone,
filling the palaeovalleys (Ramaker, 2009). This deposition interval
did not result in a very thick clastic sequence, as indicated by the
depths of the palaeovalleys. Thus, the thickest cherty conglomeratic
breccia measures 23 ft (7 m), and the thickest and most persistent
sandstone accumulation 57.0 ft (17.5 m) (Ramaker, 2009).
The cherts are reworked residual karst deposits (Nodine-Zeller,
1981). The sand may have been shed from the Paleozoic Reagan
Sandstone of the Central Kansas uplift, and the glauconite sand-
stone probably represents marine deposition in an upper estuarine
environment (Ramaker, 2009).
This clastic sequence is capped by a paleosoil, representing a
sequence boundary, and followed by a thin 15.7e19.6in (40e50 cm)
laterally continuous, wackestone-packstone horizon reflecting an
open, shallow-marine shelf environment within the photic zone as
well as marine flooding (transgressive sequence, Ramaker, 2009).
Gray shale of 1.96 ft (0.6 m) to 3.9 ft (1.2) m thickness, deposited on
top of the limestone, represents an offshore environment
(Ramaker, 2009). Updip to the northeastern part of Ness County, the
sequence boundary is overlain by coarse-grained and cross-
laminated sandstones, which were deposited when accommoda-
tion space was created by flooding (Ramaker, 2009).
2. Data and methods
In this section we present the 3D surface seismic reflection and
geophysical well-log datasets utilized in this study. We also
describe techniques and methods of analysis and interpretation
with emphasis on seismic attributes for channels/palaeovalleys
facies mapping. Borehole drilling results in this study provide
ground truthing facies control on seismic facies interpretation.
Furthermore, hydraulic geometry aspects of modern river conflu-
ence analog was utilized in corroborating, a seismic-based, inter-
preted palaeo-river channel.
2.1. 3D surface seismic reflection data and wireline well logs
Data utilized in this study includes post-stack P-wave 3D surface
seismic reflection data and wireline well logs in addition to drilling
and completion reports of five wells. We have also utilized, for
 A.E. Raef et al. / Marine and Petroleum Geology 73 (2016) 21e35 27

Fig. 9. Wyman 1e11 wireline logs: Gamma ray (GR), Sonic transient time (DT), and Resistivity Latero-logs deep (RLD) and intermediate (RLM); zones (dotted boxes) of higher
porosity and low permeability esimilar response of deep induction resistivity (RILD) and intermediate induction resistivity (RILM) logse are not associated with low GR eshale and
shaley sand facies.

independent interpretation-validation criteria, a modern analog of www.kgs.ku.edu/Magellan/Qualified/index.html: date accessed

channel morphology and hydraulic geometry signatures of post-
confluences based on a laboratory and field study by Best (1987).
The seismic data has subsurface coverage of about 12.5 km2 an area
with natural bin-size of 44 m, usable frequency bandwidth of
20e110 Hz, and amplitude-tuning effects starting from time
thickness below 20 ms (Fig. 7 aec) at target zone level (0.9 s). The
processing of the data was carried out through a surface consistent
processing sequence with amplitude preservation and pre-stack
time migration applications. The limit of temporal seismic resolu-
tion of this post-stack dataset according to criteria based on average
dominant-wavelet breadth (e.g., Widess, 1973; Kallweit and Wood,
1982) is 7 ms at best as evidenced by the tuning analysis shown in
Fig. 7c ethe resolution limit is elaborated on further in subsection.
A total of five dry holes with targets in the Cherokee formation
are located within the bounds of this study area, according to re-
cords available on the Kansas Geological Survey database, (http://
20/07/2015). Those wells, listed from north to south are: Dinkel-
Victor 1, Dinkel 1, Conrath 1e10, Wymann 1e11, and Young 1
(Fig. 8). All five wells were drilled on structural highs and were
evaluated as dry holes.
Conrath 1e10 was completed on September 10, 2005, having
been drilled to a total depth of 4530 feet. Drill stem tests were
carried out for two separate intervals identified as Cherokee sand.
DST results were not encouraging, with no blows, fluid pressures
that died within minutes, and no oil or gas shows within the 5e10
feet of mud recovered. The Wyman 1e11 was completed on March
19th, 2006, having been drilled to a total depth of 4500 feet. There
were also no oil or gas shows, and no drill stem tests were run. Full
well completion reports are available for both wells on the Kansas
Geological Survey website. Figs. 9 and 10 below provide wireline
logs (gamma ray (GR), Caliper (DCAL), Sonic transient time (DT),
Density porosity (DPOR) and deep (RILD) and intermediate (RILM)
28 A.E. Raef et al. / Marine and Petroleum Geology 73 (2016) 21e35
Fig. 10. Conrath wireline logs; target interval where density porosity (DPOR) has been plotted indicate low permeability zone as evidenced by small- to no-difference between the
deep induction (RILD) and intermediate induction (RILM) resistivity logs.
induction resistivity logs for both Wyman 1e11 and Conrath 1e10,
respectively. These logs illustrate that while there are zones 10e15
feet thick with sufficient porosity to be considered potential res-
ervoirs, the gamma ray signatures for these zones are not partic-
ularly clean, suggesting a significant amount of shale. It is worth
noting that trace amounts of oil spotting could be seen in DSTs
collected for the Dinkel-Victor 1, and Dinkel 1 wells further to the
north. It was thought by the operator that these locations, as well as
the Conrath 1e10 were perhaps on the edge a channel, with only
partial sand development.
2.2. Modeling, seismic horizons interpretation, and attributes
extraction and analysis
2.2.1. Seismic modeling and its implications
Formation top data from each well log was uploaded into an
industry-standard software (Kingdom HIS), using values taken
from completion reports available on the Kansas Geological Survey
Website for both the Conrath and Wyman dry holes. These tops are
namely; tops of Cherokee and Mississippian Groups. The seismic
modeling objectives are to identify seismic events (peak/trough)
associated with the stratigraphic tops and determine parameters
relevant to evaluating seismic resolution limits; mainly dominant
wavelet breadth and interval velocity as well as zero-phaseness of
seismic data. An average seismic wavelet and a reflectivity series
based on sonic logs were convolved to satisfy the convolutional
model and produce synthetic seismic trace that was matched with
the real seismic data at locations of Wyman and Conrath wells;
example of the synthetic to seismic match, time-depth table
established, representative wavelet, and sonic log is shown in
Fig. 11. The interval velocity in the basal Cherokee zone directly
above the Mississippian is 13,200 ft/s (4025 m/s) (Fig. 11). The
average normalized correlation coefficient of the synthetic-to-
seismic match is 0.58 for all traces in a circle of 91 m around the
well location and above 0.6 for an average trace. The mismatch is
partly related to the level of signal to noise ratio and also to devi-
ation of seismic data phase from phase of wavelet used (Henry,
2000; Ziolkowski et al., 1998). The tops (corresponding to surface
of unconformities) of both the Cherokee and Mississippian Groups
were associated with two peak-amplitude reflection events at
0.877 and 0.899 ms, respectively, at Wyman 1e11.
The limit of temporal seismic resolution of this dataset accord-
ing to criteria based on average dominant-wavelet breadth (e.g.,
Widess, 1973; Kallweit and Wood, 1982) is 7 ms (about 28 m at
interval velocity of 4025 m/s) at best with amplitude tuning effects
starting at about 22 ms (289 ft; 88 m of thickness), (Fig. 7c). Seismic
amplitude changes due to tuning effects in case of thin layers, as
defined by quarter of a wavelength (Widess, 1973), may play key
role in the detectability of thin palaeochannels and fluvial
lithofacies.
This study focuses on the top of the Mississippian formation, as
the Cherokee targets typically appear roughly 10 ms above that
formation top in the time domain (Philip, 2011). The top of the
Mississippian formation was picked (i.e. manually traced) hori-
zontally in each 3-D seismic inline and cross-line so that it could be
automatically tracked throughout the entire seismic block. Simi-
larly, the top of Cherokee formation was also picked and tracked. A
seismic cross section is displayed in Fig. 10, which depicts the
 A.E. Raef et al. / Marine and Petroleum Geology 73 (2016) 21e35 29

Fig. 11. Wyman-well 1-D synthetic-to-seismic match and tops of Heebner shale (dotted-dashed red) Cherokee (dotted-dashed blue) and Mississippian (dotted dashed green) and,
Mississippian horizon was tracked to a peak event through the 3D seismic data (For interpretation of the references to color in this figure legend, the reader is referred to the web
version of this article).

relationship between the Wyman 1e11 synthetic seismogram, the
seismic reflection series, the formation top data.
2.2.2. Seismic horizons interpretation and time structural maps
Upon determining the seismic events associated with the
Cherokee and Mississippian stratigraphic unconformities based an
acceptable match of the synthetic and real seismic data at Wyman
and Conrath, a 3-D horizon seismic horizon tracking were under-
taken. After manual editing of horizons to safeguard against in-
consistencies of automatic picking due low signal-to-noise ratio or
structural faults, we compiled time structural maps of Cherokee
and Mississippian unconformities (Fig. 12 a&b). An isochrone map
(time thickness map) of the Cherokee interval (Fig. 12c) was also
calculated based on the picked Cherokee and Mississippian time
horizons. The Mississippian time horizon, though it is essential in
its own right in attempting to map basal Cherokee palaeo channels/
valleys, represents our main guide for extraction of amplitude time
slices and other relevant stratal slice attributes within 10 ms above
or below it.
Lower values, outlines by dashed lines, on the two-way time of
the Mississippian has been interpreted as a result of incision valleys
eother supporting evidence will be examined in subsection 2.2.3.
Those valleys/palaeochannels agree with the conclusions of a study
by Archer et al. (1995) that presented results supporting develop-
ment of incision valleys of a river system in late Mississippian early
Pennsylvanian as a post-regression consequence of Hugoton
embayment regression. The thickening in the Cherokee interval
(Fig. 12c), which is obvious correlation with the Mississippian time
lows, is a corroborating evidence especially in the light of having no
time anomalies “highs” at the top of the Cherokee that may explain
the thickening.
2.2.3. 3-D seismic attributes and interpreted palaeo-river channels
and valley
Because the main objective of this study is to provide a syner-
gistic approach towards understanding and outlining the fluvial
elements that may serve as the main controls on the basal sand
lithofacis distribution of the Cherokee Group, we extracted a set of
3-D seismic attributes, which included stratal amplitude (ampli-
tude time slice extracted after Mississippian-horizon flattening)
30 A.E. Raef et al. / Marine and Petroleum Geology 73 (2016) 21e35
Fig. 12. Two-way time structural maps of the Mississippian (a) and the Cherokee (b),
and a Cherokee isochrone map (c) with interpreted incised valley overlaid (dashed
lines), all well locations are either at the periphery of or away from the thickening of
the Cherokee interval.
Fig. 13. Amplitude time slice within 6 ms above the Mississippian horizon, (a)
amplitude slice without horizon, (b) amplitude slice with horizon.
above the Mississippian; amplitude slices (amplitude along con-
stant time; no horizon-flattening); relative acoustic impedance (is
calculated by continuous integration of the original seismic trace
with the subsequent application of low cut filter; e.g. Latimer et al.,
2000), spectral-component amplitudes (amplitude constituents
spectral sub-bands); and maximum negative curvature (volumetric
estimates of curvature from volumetric estimates of reflector dip
and azimuth) attributes (e.g. Chopra and Marfurt, 2010; Al-Dossary
and Marfurt, 2006), with extraction time/time-window to maxi-
mize capturing the potential effects of channels incision into and
channel lithofacies variation above the Mississippian unconformity.
Because the main objective of this study is to provide a synergistic
approach towards understanding and outlining the fluvial ele-
ments that may serve as the main controls on the basal sand lith-
ofacies distribution of the Cherokee Group, we extracted a set of 3-
D seismic attributes, which included stratal amplitude above the
Mississippian; amplitude slices; relative acoustic impedance,
spectral-component amplitudes; and maximum negative curvature
attributes, with extraction time/time-window to maximize
capturing the potential effects of channels incision into and channel
lithofacies variation above the Mississippian unconformity. The set
 A.E. Raef et al. / Marine and Petroleum Geology 73 (2016) 21e35
Fig. 14. (a) Mississippian horizon amplitude and (b) TWTT of Mississippian horizon
with Mississippian horizon amplitude greater than 4909 overlaid as contour, higher
amplitudes greatly coincide with the interpreted incised valley, except in parts where
tuning effects (ellipse), as evidenced by isochrones map (Fig. 12c) has interfered.
of attributes targeted in our selection combine both geometrical
and lithological attributes is discussed further in section 2.2.3. For
history, classification and applications of seismic attributes, we
refer the reader to Brown (1996); Chopra and Marfurt (2007 &
2008(b)).
An amplitude time slice, extracted at 0.896 s (Fig. 13a), shows
amplitude change pattern that is in notable agreement with a
concaving downward TWTT of the Mississippian time horizon
(Fig. 13b). Areas outlined in Fig. 12a associated with amplitude
anomalies, sinuous in shape (Fig. 13b), and interpreted as incised
valley of a palaeo-river system, are also associated with higher
amplitudes (Fig. 14 a&b). The higher amplitudes are likely the
response of greater thickness of a fluvial lithofacies with lower
velocities than the underlying Mississippian carbonates. One may
not rule out the effects of thickness variation of the Cherokee on the
amplitudes of the Mississippian horizon. However, upon examining
the isochrone map (Fig. 12c), it is acceptable to relate the amplitude
“highs” (Fig. 14a) to fluvial lithofacies effects. The spatial correlation
between those amplitude highs is also in agreement with the
31
Fig. 15. Time slice at 0.896 s of Envelope of SD-components (17 and 62 Hz), (a) smooth
variation for the 17 Hz component and (b) anomalous (red) areas in conformity with
the palaeovalley (For interpretation of the references to color in this figure legend, the
reader is referred to the web version of this article).
higher impedance (lower sonic log transient time, shown by a
cross-plot in Fig. 19) of the sand log facies.
Amplitude variation of spectral decomposition (SD) compo-
nents can be of great value in outlining fluid and/or lithofacies;
thickness variation can also lead to variation in amplitudes of SD
(Giroldi and Alegria, 2005). Comparing the envelope of the 17 Hz
and 62 Hz of the SD, one can observe an agreement of the higher
anomalous regions of the 62 Hz component with the significant
part of the interpreted palaeovalley (dashed) (Fig. 15 a&b). In the
light of those anomalies the area to the north of Conrath seems to
be the most promising; likely effect of both hydrocarbon and
reservoir-grade fluvial lithofacies that resulted in higher energy
levels for the 62 Hz component.
Relative acoustic impedance (Lancaster and Whitcombe, 2000)
sinuous feature (Fig. 16 a&b) on stratal slices extracted at 3 and
5 ms above the flattened Mississippian corroborate the rock facies
signature change introduced by the interpreted fluvial lithofacies.
North of Conrath is an area with the largest areal continuity with
lower impedance and in agreement with the SD results shown in
32 A.E. Raef et al. / Marine and Petroleum Geology 73 (2016) 21e35
Fig. 16. Relative acoustic impedance stratal slices extracted at (a) 3 and (b) 5 ms above
the flattened Mississippian horizon, fluvial lithofacies is associated with high values,
north of Conrath the relative impedance high correlate with SD-component anomaly
of Fig. 15b.
Fig. 15b.
The well cuttings description of the Cherokee from the two dry
holes, Conrath 1e10 and Wyman support the geophysical inter-
pretation. The description of the Conrath hole, which is on the edge
of the channel as seen in the previous figures is as follows:
“Sandstone, white-cream-light gray w/fine-medium-trace coarse
clear-opaque-pink subangular-sub rounded quartz grains poorly
sorted … faint-poor odor fair show of dark brown free oil … .”
(http://www.kgs.ku.edu/PRS/Scans/Elog4/W/16S22/198626.zip:
accessed; 19/01/2016) The description from the geological report of
the Wyman well, significantly removed from the interpreted
channel is as follows: “Chert, vari-colored fresh; yellow shale and
influx of fine-medium and occasional coarse subangular qtz grains,
No Shows” (http://www.kgs.ku.edu/PRS/Scans/Elog4/W/16S22/
146578.zip: accessed; 19/01/2016). The increased volume of sand,
complete with oil shows, from the Conrath well is consistent with
Fig. 17. (a) a seismic amplitude time-section (inline 27) revealing asymmetry of
channel bed at the crest of a loop; this is a typical aspect of the hydraulic geometry of a
sinuous channel, (b) relative impedance (warmer color for higher) stratal slice with
location of the seismic section, dashed arrowed lines show the boundary of the incised
valley (For interpretation of the references to color in this figure legend, the reader is
referred to the web version of this article).
proximity to a channel, while the increased shale and lack of shows
from the Wyman would be expected in an area away from the sand
channel.
2.2.4. Comparison with a modern river-channel hydraulic geometry
aspects
Geomorphic aspects of channels at loop crests and post-
confluence have been well documented from studying ancient
and modern river channel systems (e.g. Best, 1987; Gamboa et al.,
2012). In this study, we examine the interpreted channel geome-
try utilizing seismic sections and the seismic horizon of the
Mississippian in order to verify our interpretation. A seismic
amplitude time-section (Fig. 17a) reveals asymmetry of channel
bed at the crest of a loop, which is a typical response of the hy-
draulic geometry aspects of sinuous channel. The location of this
seismic section is plotted on the relative acoustic impedance stratal
slice (Fig. 17b) with dashed arrowed lines show the boundary of the
incised valley.
In Fig. 18, post-confluence hydraulic geometry aspects of
scouring and cross-sectional asymmetry is evident and represent
compelling evidence supporting our interpretation and indicating a
flow direction to the south west toward a palaeoshoreline that was
outlined by Raef et al. (2015). Broadening and deepening of chan-
nels in post-confluences provide an opportunity a more pronounce
effect on seismic data.
 A.E. Raef et al. / Marine and Petroleum Geology 73 (2016) 21e35 33

Fig. 18. Relative acoustic impedance stratal slice 5 ms above the Mississippian horizon after flattening and locations of seismic sections (in-line 37&cross-line 42), arrowed dashed
lines and thick short arrows mark boundaries of the interpreted channel on cross-line 42 and in-line 37, respectively; channel geometry reflects scouring after a confluence (cross-
line; right to left) and steepening on the eastern side (marked an arrow pointing upward on the in-line section).

Fig. 19. Well-log facies (left) in DT-DPOR with point colored with deep induction resistivity log (RILD) and well logs for Wyman with well-log lithofacies overlay (gray ¼ sand,
green ¼ shale, red ¼ carbonates) (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article).
34 A.E. Raef et al. / Marine and Petroleum Geology 73 (2016) 21e35
3. Discussion and implications for pre-drill hydrocarbon
prospect evaluation
The locations of the five dry holes within this study evidence
that the well-placement (prior to this study) by the operator
focused on structural highs, which is consistent with the traditional
exploration model for this area, according to Matthew Totten's
personal communications. Drilling results of those five wells
echoed a message dedicating a reconsideration of the prospect
evaluation targeting Cherokee sands in the study area. Outlining
fluvial facies of the early Pennsylvanian Cherokee and examining
SD and amplitude anomalies of the interpreted fluvial facies offer
promising prospective sand targets north of the Conrath location.
The suggested prospect is strongly recommended for testing ac-
cording to the following criteria;
 Higher amplitude due to low acoustic impedance compared to
the underlying carbonates of the Mississippian
 Thickening of Cherokee
 The suggested prospect is part of the interpreted fluvial facies
 SD 62-Hz component high amplitude anomaly of likely associ-
ation of facies and/or fluid content cause
 The suggested prospect conform with regional structural high,
though the sealing is due to spatial facies change to shale
dominated facies as in Conrath and Wyman
Analysis of an isochron map for the Cherokee formation, and
time structure maps for both the Mississippian horizon and the
Cherokee horizons (Fig. 12aec), reveals that placement of the test
wells within this study area largely corresponds to Cherokee/
Mississippian topographic highs and indications of structural
closure. However, the log signatures and well-log facies, exempli-
fied by Wyman 1e11 (Fig. 19) indicated that the disappointing
drilling results from these wells could be attributed to a lack of sand
development. The gamma ray and porosity curves shown in Figs. 9
and 10 indicate that the sands with adequate porosity in these lo-
cations are thin, and inter-bedded with shale. However, the
appearance of these sands, nearby oil shows from the Cherokee
group, and the fact that drill stem tests were run in the first place
are all encouraging indications that a thicker, properly developed
sand may be all that is necessary for a productive well. A promising
exploration also stem from the high probability of a reservoir grade
zones in the basal parts of the Cherokee group, which is supported
by Archer et al. (1995) “palaeorivers were relatively small in distal
locations relative to the depositional axis of the early Pennsylva-
nian Hugoton embayment, the bedrock confinement, coupled with
seasonal flood cycles, created episodic, energetic flow; thus the
initial valley fills of such incised valleys can be conglomeratic even
within the stable shelf setting”.
As shown in Fig. 16, relative acoustic impedance values “low
contrast (high impedance) with the underlying carbonates of
Mississippian) coincides with the placement of the Dinkel-Victor,
Young, and Conrath dry holes, and because such an increase
could potentially represent a loss of reservoir quality rock, this in-
formation would have weighed against placing these wells in their
current locations. Conversely, a north to south trending feature of
relative acoustic impedance anomaly (high contrast “low imped-
ance” with the underlying carbonates of Mississippian) located
north of Conrath 1e10 and south west Dinkel-Victor is of high
potential as further justified above.
A basic paleodepositional interpretation can be made based on
an examination of the Cherokee/Mississippian time-structure maps
and the identified channel sand body. The relative structural highs
to north, suggest that this location may be a north-eastern exten-
sion of the transitional continental-marine paleo-shoreline
identified in Raef et al. (2015). The structure observed in these maps
may post-date the deposition of the Cherokee group, however, as
stratal amplitude time slices above the flattened Mississippian
horizon do not indicate a shoreline nearly as clearly. This suggests
that the channel outlined in this study may be further inland, which
is in agreement with the meandering, fluvial channel geometries
seen here. Regardless, the interpreted Cherokee channel sand body
identified in this study has a general NEeSW trend, which is in
agreement with the system identified in Raef et al. (2015).
4. Conclusions
In conclusion, the results of this study indicate the presence of a
NEeSW trending channel sand facies, in the basal part of the
Cherokee Group, located in the southeast/central portion of the
study area. Isochron thickening, TWTT of Mississippian, time slices
of amplitude and stratal relative acoustic impedance and SD 62-Hz
anomalies, strongly supports this conclusion. The general trend
(NEeSW) of this channel system, its time-depth location, and its
proximity to what has been interpreted as a paleo-shoreline (Raef
et al., 2015), for the nearby Wierman field in addition to confor-
mity hydraulic geometry aspects of modern analogs, corroborate
our seismic attributes based interpretation of a fluvial facies and
channel system.
Finally, it is recommended that future exploration operations
consider the channel feature outlined in this study when making
well placement decisions. Some of the most significant concerns for
the prospects identified within this channel would be the potential
for lack of structural development/closure (i.e. an adequate trap-
ping mechanism), and the possibility of reservoir compartmental-
ization. Regardless, utilization of the 3D seismic attributes outlined
in this study is highly recommended as part of an integrated
workflow for future petroleum exploration operations in this area
of Ness County.
Acknowledgments
We are thankful to David Doyel of Murfin Drilling, Wichita,
Kansas, for donating the 3D seismic data and other information in
support this research project. We also thank IHS-kingdom support
staff and management for granting Kanas state University, Geology
Department, an academic software license for Kingdom Suite
seismic interpretation software.
References
Abbas, M.Y., 2009. Incorporating Seismic Attribute Variation into the Pre-well
Placement Workflow: a Case Study from Ness County, Kansas, USA. Master's
Thesis. Kansas State University.
Al-Dossary, S., Marfurt, K.J., 2006. 3D volumetric multispectral estimates of reflector
curvature and rotation. Geophysics 71 (5), P41eP51.
Archer, Allen, W., Greb, Stephen F., 1995. An amazon-scale drainage system in the
early Pennsylvanian of central North America. J. Geol. 611e627.
Bahorich, M.S., Farmer, S.L., 1994. 3-D seismic discontinuity for faults and strati-
graphic features: the coherence cube. In: Society of Exploration Geophysicists
International Exposition and 65th Annual Meeting. Society of Exploration
Geophysicists, Tulsa, Oklahoma, pp. 93e96.
Bass, N.W., 1936. Origin of the Shoestring Sands of Greenwood and Butler Counties,
Kansas. Kansas Geological Survey Bulletin 23.
Best, J.L., 1987. Flow dynamics at river channel confluences: implications for sedi-
ment transport and bed morphology. In: Ethridge, F.G., Flores, R.M.,
Harvey, M.D. (Eds.), Recent Developments in Fluvial Sedimentology. Soc. Sedi-
ment. Geol. SEPM 39, pp. 27e35.
Brenner, R.L., 1989. Stratigraphy, Petrology, and Paleogeography of the Upper
Portion of the Cherokee Group (Middle Pennsylvanian), Eastern Kansas and
Northeastern Oklahoma (Kansas Geological Survey).
Brown, A.R., 1996. Seismic attributes and their classification. Lead. edge 15 (10),
1090.
Chen, Q., Sidney, S., 1997. “Seismic attribute technology for reservoir forecasting and
monitoring”. Seismic attribute technology for reservoir forecasting and
 A.E. Raef et al. / Marine and Petroleum Geology 73 (2016) 21e35
monitoring. Lead. Edge 16 (5), 445e448. http://dx.doi.org/10.1190/1.1437657.
Chopra, S., Marfurt, K.J., 2007. Seismic attributes for Prospect identification and
reservoir characterization. Society for Exploration Geophysicists, Tulsa, OK,
USA.
Chopra, S., Marfurt, K.J., 2008. Seismic attributes for stratigraphic feature charac-
terization. In: 78th Annual International Meeting, SEG Technical Program
Expanded Abstracts, vol. 27, pp. 1590e1594 no. 1.
Chopra, S., Marfurt, K.J., 2008b. Emerging and future trends in seismic attributes.
Lead. Edge 27 (3), 298e318.
Chopra, S., Marfurt, K.J., 2010. Integration of coherence and volumetric curvature
images. Lead. Edge 29 (9), 1092e1107.
Clark, S.L., 1987. Seismic Stratigraphy of Early Pennsylvanian Morrowan Sandstones,
Minneola Complex, Ford and Clark Counties, Kansas. AAPG Bulletin 71, no. 11,
pp. 1329e1341.
Cuzella, J.J., 1991. Depositional Environments and Facies Analysis of the Cherokee
Group in West-central Kansas. AAPG bulletin, 75, no. 3, p. 560.
Doveton, J.H., 2002. Dipmeter navigation of the location and orientation of a
Cherokee Sandstone reservoir: a Kansas case study. In: Boyd, D.T. (Ed.), Finding
and Producing Cherokee Reservoirs in the Southern Midcontinent, pp. 37e41,
2002 symposium: Oklahoma Geological Survey Circular 108.
Gamboa, D., Alves, T.M., Cartwright, J., 2012. A submarine channel confluence
classification for topographically confined slopes. Mar. Petroleum Geol. 35 (1),
176e189.
Giroldi, L., Alegria, F., 2005. Using spectral decomposition to identify and charac-
terize glacial valleys and fluvial channels within the Carboniferous section in
Bolivia. Lead. Edge 24 (11), 1152e1159.
Henry, S., 2000. Pitfalls in synthetics. Lead. Edge 19 (6), 604e606.
Kallweit, R.S., Wood, L.C., 1982. The limits of resolution of zero-phase wavelets.
Geophys. 47, 1035e1046.
Lancaster, S., Whitcombe, D., 2000. Fast-track ‘colored’ Inversion, 70th SEG Meeting,
Calgary, Canada, Expanded Abstracts, pp. 1572e1575.
Latimer, R.B., Davidson, R., Van Riel, P., 2000. An interpreter's guide to under-
standing and working with seismic-derived acoustic impedance data. Lead.
edge 19 (3), 242e256.
Merriam, D.F., 1963. The Geologic History of Kansas. Kansas Geological Survey
35
Bulletin 162.
Mourning, R.C., 2014. Seismic Attribute Analysis of the Mississippian Limestone:
Ness County, Kansas. Master's Thesis. Kansas State University.
Nodine-Zeller, D.E., 1981. Karst-derived Early Pennsylvanian Conglomerate in Ness
County, Kansas: Subsurface Mississippian-pennsylvanian Boundary Delineated
in Well Core (Kansas Geological Survey).
Philip, C.C., 2011. 3-D Seismic Attributes Analysis to Outline Channel Facies and
Reveal Heterogeneous Reservoir Stratigraphy; Wierman Field, Ness County,
Kansas, USA. Master's Thesis. Kansas State University.
Raef, A.E., Mattern, F., Philip, C., Totten, M.W., 2015. 3D Seismic attributes and well-
log facies analysis for prospect identification and evaluation: interpreted
palaeoshorline implication, Weirman Field, Kansas. USA. Petroleum Sci. Eng.
133, 40e51.
Ramaker, B.J., 2009. Influence of Mississippian Karst Topography on Deposition of
the Cherokee Group: Ness County, Kansas. Master's thesis. University of Kansas,
183 pp.
Suarez, Y., Marfurt, K.J., Falk, M., 2008. Seismic Attribute-assisted Interpretation of
Channel Geometries and Infill Litology: a Case Study of Anadarko Basin Red
Fork Channels, 78th Annual International Meeting, SEG, pp. 963e967.
Verma, A.K., Pereira, Metilda, Bharali, B.R., Khanna, A.K., Dasgupta, R., 2009. Use of
Spectral Decomposition in Seismic Interpretation for Finding Out Fluvial
Channel Sand Body: a Case Study from Upper Assam Shelf Basin, India. SEG
Annual Meeting. Society of Exploration Geophysicists.
Walters, Robert F., Gutru, Robert J., James III, Alfred, 1979. Channel Sandstone Oil
Reservoirs of Pennsylvanian Age in Northwestern Ness County, Kansas,
pp. 313e326.
Widess, M.B., 1973. How thin is a thin bed. Geophys. 38, 1176e1180.
Ziolkowski, A., Underhill, J.R., Johnston, R.G., 1998. Wavelets, well ties, and the
search for subtle stratigraphic traps. Geophys. 63 (1), 297e313.
Web references
http://www.kgs.ku.edu/PRS/Scans/Elog4/W/16S22/198626.zip, (accessed 19.01.16.).
http://www.kgs.ku.edu/PRS/Scans/Elog4/W/16S22/146578.zip, (accessed 19.01.16.).