Geological Society, London, Special Publications published online June 27, 2012 as doi: 10.1144/SP370.6
Reservoir properties and petrophysical modelling
of carbonate sand bodies: outcrop analogue study
in an epicontinental basin (Triassic, Germany)
DENIS PALERMO1,2*, THOMAS AIGNER1, BJOERN SEYFANG1,3 & SERGIO NARDON2
1
Department of Geosciences, Centre of Applied Geosciences,
University of Tübingen, Sigwartstr. 10, D-72076 Tübingen, Germany
2
Eni S.p.A. – Exploration & Production Division, via Emilia 1,
20097 San Donato Milanese, Italy
3
Present address: Centre Scientifique et Technique Jean Feger (CSTJF),
Avenue Larribau, F-64018 Pau Cedex, France
*Corresponding author (e-mail: [email protected])
Abstract: This paper represents the second part of an integrated study that is focussed on the
development and distribution of reservoir bodies and properties in epeiric carbonate systems. It
is based on outcrop analogue data from Triassic ‘layer-cake’ carbonates in the South German
Basin, which were deposited along an epicontinental, very gently inclined carbonate ramp.
The reservoir facies consists of skeletal and oolitic carbonate grainstones (Fmax 23%, Kmax
700 mD), which are organized in a pronounced hierarchy of stratigraphic cycles. Based on outcrops, cores, gamma ray (GR) logs and thin sections, a high-resolution, geocellular 3D facies
model was generated, which covers the area of a Middle East giant gas field (25 × 36 km). The
spatial distribution of reservoir properties was systematically investigated on different scales.
The lateral distribution of reservoir properties remains in the same order of magnitude for hundreds
of metres, within in the same stratigraphic position. However, on a kilometre scale, facies bodies,
diagenetic trends and thus reservoir properties show gradual lateral changes. Vertically, in contrast,
properties change commonly on a decimetre scale and are largely controlled by stratigraphic
cycles. Petrophysical modelling enhanced the understanding of key factors and processes controlling both reservoir quality and quantity.
Reservoir heterogeneity remains a significant issue in
reservoir modelling and prediction of field performance. Outcrop analogues are ideal for establishing
the possible geometries and property distributions
at the inter-well scale. Quantitative data on carbonate
rock bodies remain scarce (e.g. Handford 1988;
Burchette et al. 1990; Grant et al. 1994; Harris &
Kowalik 1994; Borgomano et al. 2002; Grammer
et al. 2004; Kostic & Aigner 2004; Ruf & Aigner
2004; Rankey et al. 2006; Qi et al. 2007; Aigner
et al. 2007; Palermo et al. 2008). Qualitative outcrop studies of geometry (e.g. Burchette et al. 1990;
Gawthrope & Gutteridge 1990; Azerêdo 1998) and
quantitative petrophysical studies carried out on
outcrops at the appropriate scale (e.g. Kittridge
et al. 1990; Senger et al. 1991; Eisenberg et al.
1994; Cavallo & Smosna 1997; Jennings 2000;
Savary & Ferry 2004; Pranter et al. 2005, 2006)
made useful contributions to the available data.
Outcrop analogue studies of this nature are
important in order to condition subsurface reservoir
models to real data and concepts, thus significantly
improving field appraisal work and development
planning. The present work has resulted from a
joint ENI E&P –University Research Consortium
on the ‘Geometry of Carbonate Objects’. Within
this project, the Triassic Upper Muschelkalk carbonates in the South-German Basin were studied
as an analogue to the ‘layer-cake’-type reservoir
systems of the Middle East. The Upper Muschelkalk
was deposited on a gently inclined carbonate ramp,
filling an epicontinental basin, and therefore represents an analogue to an important type of ‘nonreefal’ skeletal and oolitic carbonate sand reservoirs
(e.g. Khuff, Hanifa, and Arab in the Middle East).
A series of excellent quarries and natural outcrops in Southern Germany has been used to investigate the geometries within high-energy shoal
water deposits along the margin of the Upper
Muschelkalk Basin. Close outcrop spacing allows
for lateral tracing of beds and mapping of lateral facies transitions. A hundred years of detailed
From: Garland, J., Neilson, J. E., Laubach, S. E. & Whidden, K. J. (eds) 2012. Advances in Carbonate Exploration
and Reservoir Analysis. Geological Society, London, Special Publications, 370,
http://dx.doi.org/10.1144/SP370.6 # The Geological Society of London 2012. Publishing disclaimer:
www.geolsoc.org.uk/pub_ethics
Geological Society, London, Special Publications published online June 27, 2012 as doi: 10.1144/SP370.6
D. PALERMO ET AL.
logging and mapping has resulted in a wellconstrained litho- and biostratigraphic framework
(e.g. Wagner 1913; Vollrath 1938, 1955, 1957,
1958, 1970; Skupin 1969; Bachmann 1973; Aigner
1985; Ulrichs & Mundlos 1987, 1990; Hagdorn &
Simon 1988; Ockert 1988; Geyer & Gwinner 1991;
Braun 2003). This well-established stratigraphy provides the opportunity to transform the so far largely
descriptive picture into a genetic, process-based
analysis of the sedimentary bodies and packages.
A detailed outcrop, core and well logging programme was carried out (e.g. GR logs, cores, porosity and permeability), allowing an analysis of the
sedimentology, petrophysical characteristics, internal architecture and reservoir properties, within a
3D geocellular model. Palermo et al. (2010) have
recently reported on the facies modelling aspects
of this project. This companion paper focusses on
reservoir properties and petrophysical modelling.
Geological setting
During the Triassic, rifting and disintegration of the
Pangean supercontinent caused the western extension of the Tethys Ocean. This regional crustal extension induced the subsidence of a complex network of
grabens and troughs in Western and Central Europe
(e.g. Ziegler 1990) and formed the Triassic basins
including the South German Muschelkalk Basin
(Fig. 1). This basin extended roughly from Poland
to the North Sea and from the Alpine foothills to
Denmark. During the Middle Triassic (≏240–
231 Ma), the Muschelkalk Basin was covered by an
epicontinental, semi-enclosed marginal sea that was
separated by the Vindelician/Bohemian High from
the open Tethys ocean in the SE. Temporary connections existed only through three narrow shifting seaways (Ziegler 1990; Dercourt et al. 1993). In the
early Anisian a relative sea-level rise induced the
Lower Muschelkalk transgression from the open
Tethys ocean via the Silesian–Moravian and EastCarpathian Gates into the Germanic Basin with
fully marine conditions. The open communication
became restricted during the later Anisian, resulting
in hypersaline conditions and formation of evaporites
of the Middle Muschelkalk. Finally the fully marine
carbonates of the Upper Muschelkalk were deposited
during a transgression in the uppermost Anisian, connecting the Germanic Basin with the open Tethys
realm via the three gates. The Upper Muschelkalk
succession (Middle/Late Ladinian) represents one
overall transgressive/regressive third-order sea-level
cycle (Aigner 1985). The transgressive part is represented by an overall fining-upward sequence of
crinoidal/shelly shoal water carbonates grading into
Fig. 1. Palaeogeography of Central Europe in the Middle Triassic modified after Ziegler (1990), Hagdorn (1991) and
Palermo et al. (2010) (AAPG # 2010, reprinted by permission of the AAPG, whose permission is required for further use).
Geological Society, London, Special Publications published online June 27, 2012 as doi: 10.1144/SP370.6
CARBONATE SAND BODIES IN GERMANY
muddy carbonates and marlstones. The upper regressive section is represented by an overall coarseningupward sequence, grading from the muddy sediments
around maximum transgression into shelly/oolitic
shoal carbonates and muddy backshoal sediments.
The boundary between the Muschelkalk and the
overlying, mainly siliciclastic Keuper is considered
as maximum regression (Aigner et al. 1999). Generally, five major facies belts can be distinguished
in the Upper Muschelkalk: (a) a narrow zone of
coastal siliciclastics; (b) an irregular backshoal zone
dominated by (partly dolomitized) peloidal mudand wackestones; (c) a high energy belt of skeletal/
oolitic pack- to grainstones; (d) shallow-ramp sediments dominated by skeletal storm sheets; and (e) a
deeper ramp characterized by mud- and marlstones.
The facies belts are modified by subtle structural
trends that define subtle palaeo-highs and palaeolows, which are well known in the regional geology
and can also be recognized by thickness variations
(isopach maps) in various stratigraphic levels (e.g.
Geyer & Gwinner 1991). These are probably reactivated basement blocks, inducing slight variations in
differential subsidence.
Study area, methods and database
The SW German Hohenlohe area provides numerous quarries and natural outcrops in the Upper
Muschelkalk (Fig. 2). The shoal water carbonates
represent a major target for the raw-material industry in this region. Therefore, the closest spacing of
quarries can be observed in areas where the clean
and porous skeletal/oolitic grainstone bodies show
their thickest development. Furthermore, natural
exposures along river gorges allow a lateral tracing
of individual bodies and property transitions.
The database consists of 50 measured outcrop
sections, supplemented by six cores and wireline
logs (1787 m). Thirty-one of these sections (711 m)
were integrated from previous work (Aigner 1985;
Kostic 2001; Ruf 2001; Braun 2003; Allgöwer
2006; Dmitrieva 2006; Looser 2006; Seyfang
2006). For the investigation of diagenetic and petrophysical properties 568 cylinder-shaped plugs were
extracted from rock samples and cores. The plugging focused on potential reservoir rocks, areas
with visual porosity or dolomite content. The
plugs have a diameter of four centimetres and
Fig. 2. Composite map of the study area indicating the data points and the thickness of the Upper Muschelkalk;
modified after Palermo et al. (2010).
Geological Society, London, Special Publications published online June 27, 2012 as doi: 10.1144/SP370.6
D. PALERMO ET AL.
were cut to a depth between 4.5 cm and 3 cm.
Additionally, four rock columns (1 m × 0.2 m ×
0.2 m) were derived from a block (2.1 m ×
1.1 m × 1.0 m) in order to investigate the distribution of permeability within an individual reservoir body on different scales.
Porosity and density measurements were carried
out with two devices in two separate steps:
(a) The determination of net volume (netVol) and
net density (without pores) was carried out with
a calibrated helium pycnometer.
(b) Gross volume (grossVol) of the sample was
measured with a powder pycnometer. This
device calculates the rock density and the percentage porosity (F) with the formula:
F = (grossVol − netVol) × 100/grossVol.
Permeability was measured with a gas minipermeameter. Eight measurements were carried
out on every sample, to provide the mean horizontal
and the mean vertical permeability. The device is
controlled by special automatic software, which
adjusts the contact pressure, regulates the flow rate
and calculates the permeability from the gas-flow
rate. The measuring error of the mini-permeameter
is around 0.5 mD.
For the reconstruction of the diagenetic history
and the deeper understanding of reservoir properties, 451 stained thin sections were investigated. For a systematic semi-quantitative analysis of
several parameters (e.g. amount of components,
mud content, cement generations, pore types) the
comparison charts of Flügel (2004) were applied.
The petrography and cement-types were investigated with transmission light and cathodoluminescence microscopy. The pore types were categorized
and analysed by using mainly the concepts of
Choquette and Pray (1970) and Lucia (1983). In
order to quantify the results with the porosity and
permeability measurements, the thin sections were
commonly derived from the trim ends of the plugs.
The 3D petrophysical modelling was carried out
using industry-standard software within a high resolution 3D facies model described in Palermo
et al. (2010).
Facies and diagenesis
Palermo et al. (2010) and thus do not need to be
further discussed here. The facies types are interpreted to be deposited within a shallow epeiric carbonate ramp setting, characterized by a very gentle
depositional gradient (between 0.0028 to 0.38).
The carbonate ramp can be subdivided into three
major subenvironments (cf. Burchette et al. 1990):
(1) low energy inner ramp or backshoal; (2) highenergy mid-ramp with carbonate sands in shoreline
detached shoals; and (3) low energy foreshoal
and outer ramp with mud-dominated successions
(Fig. 3). Figure 4 shows a generalized depositional
model with the main reservoir facies types and
their petrographic composition.
Diagenetic history
The diagenetic history of the Upper Muschelkalk
grainstones has been subject to several previous
studies (e.g. Bachmann 1973; Kostic 2001; Braun
2003; Seyfang 2006). Bachmann (1973) differentiated two major diagenetic stages with several
cementation and leaching phases:
(1)
(2)
Within the early diagenetic stage under
shallow burial, isopachous circum-granular
crust cement (A1 cement; Fig. 5.1) consolidated the original sedimentary grains. This
process was followed by leaching of the
aragonitic components and the precipitation
of secondary isopachous crust cement (A2
cement; Fig. 5.2) of similar appearance,
grown exclusively on the fringes inside the
dissolved pores.
The late diagenetic stage was refined by observations of Braun (2003), who recognized
an additional phase of late diagenetic dog
tooth cement (B1 cement) grown in irregular
rims of isopachous prismatic spar, which
was followed by the precipitation of equant
drusy calcite-spar (B2 cement; Fig. 5.3).
Finally, a minor amount of scattered dolomite
cement precipitated in remnant separate vug
pore space.
The present investigation of 442 thin sections
from different locations and stratigraphic intervals,
using standard petrographic, cathodoluminescence
and fluorescence microscopy, resulted in slight revisions of some aspects of the so far established late
diagenetic history (Table 1).
Facies types
Based on lithology, Dunham texture, particle size,
sorting, sedimentary structures and environmentindicative allochems such as for example ooids,
oncoids and black pebbles 16 different lithofacies
types were distinguished in the Upper Muschelkalk
of the study area. These have been described in
Discussion: controlling factors of diagenesis
Facies. Facies types and their primary composition seem to have a major impact on diagenesis
and reservoir development since both leaching
and dolomitization are selective with respect to
the mineralogy of components. For instance, the
Geological Society, London, Special Publications published online June 27, 2012 as doi: 10.1144/SP370.6
CARBONATE SAND BODIES IN GERMANY
Fig. 3. Facies types with key attributes, depositional environment and colour code.
semi-quantitative analysis of 442 thin sections
(Figs 5 & 6) shows a negative correlation between
mud content and the intensity of early diagenetic
A-type cements (correlation coefficient, 20.7; standard deviation, 0.19). However, apart from the
depositional energy and mud content, the growth
intensity of A1 cements is also controlled by the
type of associated components. For example, A1
cements are often pore filling in ooid dominated
grainstones, while they from only thin seams
around crinoid ossicles. In general, however, the
relationship between primary facies and diagenesis
is complex (see below).
Stratigraphy. The systematic semi-quantitative
investigation of thin sections in their stratigraphic context showed that diagenetic alteration
is also controlled by stratigraphic cycles. This is
Geological Society, London, Special Publications published online June 27, 2012 as doi: 10.1144/SP370.6
D. PALERMO ET AL.
Fig. 4. Facies distribution and petrographic composition, based on the semi-quantitative analysis of 442 thin sections.
particularly important for the prediction of systematic changes of reservoir properties within individual
facies bodies. An example for these variations
within the reservoir facies of one regressive
medium-scale hemicycle is depicted in Figure 6.
In particular, the mud content and primary A
cements change systematically within the cycle.
Generally, the sum of mud content and A cements
constitutes a more or less constant petrographic
volume fraction together (arithmetic mean of mud
content and A cement, 40.9 Vol%; standard deviation, 17.3; n ¼ 442 thin sections). As displayed
in the column ‘Cumulative Vol %’ of Figure 6,
upwards decreasing mud content is compensated
by a systematic increase in B-type cements. Both
A-type and B-type cement combined show a clear
upward increasing fracture towards the mediumscale regressive maximum.
Maximum values of interparticle and mouldic
porosity are commonly located around the regressive maximums of the medium-scale cycles in shoal
facies associations, following an upward decreasing
fraction of matrix mud (Fig. 7b). However, intense
late diagenetic B cements can partially plug the
pore space in the upper portion the regressive hemi
cycles, shifting the maximum porosity and permeability values towards the middle part (Fig. 6).
The upward decreasing mud content reflects the
vertically increasing depositional energy of a prograding shoal complex and seems to be a major
controlling factor in the reservoir properties. The
carbonate mud in the lower portion of the regressive
Geological Society, London, Special Publications published online June 27, 2012 as doi: 10.1144/SP370.6
CARBONATE SAND BODIES IN GERMANY
Fig. 5. Thin section photographs. (1) Thin section Crailsheim, parallel Nichols, stained: blue indicates porosity.
Pel-oolitic grainstone: the peloids are surrounded by A1 cement (A1) forming typical circum-granular crusts of fibrous
to bladed calcite crystals. Note the preserved primary interparticle porosity between the grains (Ip). The black opaque
inclusions within the peloids are probably pyrite crystals (Py). (2) Thin section Sattelorf-Bölgental, parallel Nichols,
stained: blue indicates porosity. Crinoidal grainstone: shape-preserving micritic envelope of a dissolved aragonitic
bioclast (M). A1 cement (A1) forms a more regular seam of fibrous to bladed crystals compared with the more irregular
A2 cement (A2) grown inside the biomouldic pore (Bm). (3) Thin-section Steinbächle, parallel Nichols. Bioclastic
grainstone: a mosaic of drusy, equant B2 cement (B2) overgrows the earlier generation of B1-‘dog tooth’ cement (B1).
(4) Thin-section Crailsheim, parallel Nichols, stained: blue indicates porosity. Oolitic grainstone: the dolomitization of
ooids and coated grains occurs preferentially by replacing the cortical layers with a sub- to anhedral crystal mosaic (D).
Note the undisturbed primary A1 cement (A1) surrounding the grains. Note the preserved primary porosity (Ip). (5a, b)
Thin-section Steinbächle, parallel Nichols (1); crossed Nichols (2), stained: blue indicates porosity. Oolitic grainstone:
patchy saddle dolomite cement with typical curved crystal faces (S) and sweeping extinction (E). (6) Thin section
Satteldorf Bölgental, parallel Nichols, stained: blue indicates porosity. Crinoidal grainstone: limonitic crusts of leached
saddle dolomites (LC), and partly leached dolomite crystals (LD) with rusty limonitic fractures. Remaining A1 cement
has grown in the mould of an early diagenetic leached possibly aragonitic coated grain-core (A1). The cortical layers
were dolomitized and subsequently leached during the second leaching phase (L2).
hemicycles reduced the primary porosity and the
circulation of diagenetic fluids, which enhanced
the porosity in later diagenetic stages. However, in
cases with intense B cementation at the top of the
regressive hemicycles, the best reservoir properties
tend to be located below the regressive maxima
(Fig. 6).
Reservoir properties
In order to quantify the reservoir properties, 442 thin
sections and 570 porosity and permeability plugs
from outcrop samples and cores were analysed.
Porous intervals are generally restricted to discrete
reservoir bodies, composed of high-energy shoal
and shoal fringe facies types. However, the internal
petrophysical characterization of these reservoir
bodies turned out to be complex. During diagenesis,
several cementation and leaching phases led to a
stepwise alteration of the pore-space, which resulted
in a complex relationship between reservoir properties and reservoir-facies types. Nevertheless,
owing to the selective character of the diagenetic
processes, the initial sediment composition and
Geological Society, London, Special Publications published online June 27, 2012 as doi: 10.1144/SP370.6
D. PALERMO ET AL.
Table 1. Overview of the diagenetic history and the impact on pore-space
Diagenetic stage
Early marine, shallow
burial diagenesis
Diagenetic phase
Impact on pore space
Circum granular crust
cement (A1 cement)
First leaching phase
Protection and reduction of primary porosity
Late diagenetic, deep
burial dolomitization
Circum granular crust
cement (A2 cement)
Dog tooth cement
(B1 cement)
Patchy calcite spar
cement (B2 cement)
Patchy dolomitization
and recrystallization
Uplift and meteoric alteration
Second leaching phase
Late marine, deeper
burial cementation
mineralogy remains one of the most important controlling factors. The investigated Upper Muschelkalk reservoir bodies have average reservoir
properties of 11.5% porosity (standard deviation,
4.9) and 30.8 mD permeability (standard deviation
72.4) with a correlation coefficient of 0.41.
Pore-types
Lucia (1983) developed commonly used concepts
for reservoir characterization of carbonate rocks by
subdividing the pore-space into the three major pore-
Creation of biomouldic porosity owing to
selective leaching of aragonite
Slight reduction of complete plugging
of biomouldic porosity
Slight reduction of complete plugging
of biomouldic and primary porosity
Slight reduction of complete plugging
of biomouldic and primary porosity
Slight reduction of complete plugging
of biomouldic and primary porosity,
dolomitization of ooids, coated grains
and crinoidal columnar plates
Creation of Oo- and crino-mouldic porosity
type classes: (a) touching vug porosity; (b) separate
vug porosity; and (c) interparticle porosity. The
Upper Muschelkalk is dominated by separate vug
porosity (SV, 60%). Touching vug (TV) and interparticle porosity (IP, 20%) constitute the minor
part, but are important factors controlling permeability. However, the touching vugs observed in
the Upper Muschelkalk are smaller (up to several
millimetres; Fig. 8) than those described by Lucia
(1983).
Using the fabric-selective concept for the classification of carbonate pores after Choquette and Pray
Fig. 6. Diagenesis and the resulting petrography are controlled by stratigraphic cycles. Generally, mud content and
A cements constitute a more or less constant petrographic volume fraction, as displayed in the column ‘Cumulative
Vol %’; upwards decreasing mud content is compensated by a systematic increase of A cement.
Geological Society, London, Special Publications published online June 27, 2012 as doi: 10.1144/SP370.6
CARBONATE SAND BODIES IN GERMANY
Fig. 7. (a) Cross section through two reservoir bodies (yellow/orange) along the depositional gradient, highlighting
facies distribution, sedimentology and reservoir properties. Within the individual facies bodies, the lateral porosity and
permeability distribution is mostly marked by gradational transitions. (b) Cross section highlighting the qualitative
distribution of cement types and mud content. The vertical sections depict the cumulative petrographic composition
of the shoal bodies, combined with petrophysical properties. Both properties show laterally gradational transitions and
are strongly influenced vertically by stratigraphic cycles.
(1970), biomouldic porosity comprises the largest
pore-fraction. Most of the investigated samples
belong to one of the following combined pore-types:
(1)
(2)
(3)
biomouldic (BM, 27%);
biomouldic and interparticle (BM + IP, 42%);
biomouldic, interparticle and Oomouldic
(BM + IP + OM, 23%);
(4)
biomouldic, and oomouldic (BM + IP + OM,
8%).
Porosity and permeability relationships
Pore types after Choquette and Pray (1970) v.
porosity and permeability (Fig. 9a). Generally, the
samples with pure mouldic porosity show less
Geological Society, London, Special Publications published online June 27, 2012 as doi: 10.1144/SP370.6
D. PALERMO ET AL.
Fig. 8. Thin section photographs, stained: blue indicates porosity, parallel Nichols: typical pore-types in the Upper
Muschelkalk reservoir bodies. Pore-types after Lucia (1983): touching vugs, TV; separate vugs, SV; interparticle
porosity, IP. Pore-types after Choquette & Pray (1970): biomouldic porosity, BM; oomouldic porosity, OM. (a) Oolitic
grainstone with preserved primary interparticle porosity. (b) Bioclastic grainstone showing two pore generations of
biomouldic porosity. Upper left corner: leached dolomite recrystallization. Middle right corner: early diagenetic
biomouldic pore with several cement generations. (c) Bioclastic pack- to grainstone with touching vugs owing to
leached dolo-cement. The dissolution enlarged touching vugs of the Upper Muschelkalk reservoir bodies are much
smaller (several millimetres) than those described by Lucia (1983). Furthermore, remnants of collapsed biomouldic
porosity are visible in the upper middle part. (d) Oolitic grainstone. Typical oomouldic separate vug porosity, enhanced
by a large touching vug of leached dolo-cement and small amounts of interparticle porosity. (e) Bioclastic grainstone:
oomouldic porosity of leached dolomite recrystallizations combined with early diagenetic biomouldic separate vug
porosity and large interparticle pores. (f) Bioclastic grainstone: early diagenetic biomouldic separate vug porosity.
permeability than samples where both, mouldic- and
interparticle porosity are combined. Regarding the
pore type distribution within the Lucia classes,
samples with biomouldic, oomouldic and interparticle porosity tend to class 2, while samples with biomouldic and interparticle porosity plot slightly more
towards Lucia’s class 1.
Reservoir rock types and classes using Lucia (1983)
classification (Fig. 9b). Several studies on the
Upper Muschelkalk carbonates have shown that
porosity – permeability relationships show strong
scattering (e.g. Braun 2003; Kostic & Aigner
2004; Ruf & Aigner 2004; Dmitrieva 2006;
Seyfang 2006). However, Lucia (1999) was able
to categorize porosity – permeability relationships
in discrete petrophysical classes by the definition
of rock-fabric types, which are a combination of
both Dunham texture and pore-type. His rockfabric petrophysical classes were mainly developed
for non-vuggy carbonates with interparticle porosity. As the fraction of vuggy porosity within the
investigated reservoir bodies amounted to 80%,
the Lucia classes could only be applied to a
certain extent. Most of the porosity and permeability values plot in class 2 of Lucia, followed by
class 1. The values above 3 mD permeability
show a relatively linear trend within a certain
range of scatter. The values plot relatively close
together between Lucia classes 1 and 2. As
Geological Society, London, Special Publications published online June 27, 2012 as doi: 10.1144/SP370.6
CARBONATE SAND BODIES IN GERMANY
Fig. 9. Porosity and permeability relationships keyed to (a) pore type according to (a) Choquette & Pray (1970),
(b) Lucia (1983), (c) approximated pore-throat size measured in thin sections and (d) pore size.
expected, samples with only separate vugs have
generally lower permeability. When touching-vug
porosity is present, permeability is higher. The
highest porosity and permeability values occur
when all pore types are present (Table 2).
However, biomouldic pores compose the largest
part of porosity and occur, independent of Dunham
texture, in all reservoir rock types. This fact implies
that the distribution of biomouldic porosity could
be an additional key controlling factor for the
reservoir properties.
Pore-throat size v. poro-perm (Fig. 9c). The porethroat diameters presented in this study have been
measured in 2D thin sections and represent therefore just a limited approximation to the real 3D
pore-throat distribution. Nevertheless, Figure 9c
depicts a clear increase of both permeability and
porosity with larger pore-throats (Table 4). The
remaining data scatter could be due to the fact that
pore-throat diameters were measured in 2D.
Pore size v. porosity and permeability (Fig. 9d).
Despite considerable scattering, permeability generally increases with an increasing pore-size (Table 2).
Relationship facies association: reservoir
properties
As can be observed in many carbonate systems, facies and reservoir properties of the Upper
Muschelkalk do not show a direct relationship.
This is mainly due to the susceptibility of carbonates
to diagenetic alteration, but also to the higher
complexity of primary sedimentary parameters in
addition to measuring inaccuracies. Nevertheless,
the primary lithofacies plays a fundamental role
for the distribution of reservoir properties in the
Upper Muschelkalk, since porous shoal bodies are
exclusively formed of high energy shoal and shoalfringe facies types. Furthermore, the diagenetic
leaching and dolomite cementation phases proceeded very selectively according to the mineralogy
of the components, a parameter that is in turn mainly
facies controlled.
Therefore, despite considerable data scattering, a
certain relationship between facies association and
reservoir properties can be observed.
Shoal-fringe facies association (pel-ooidal packto grainstone, oncoidal pack- to grainstone, intraclastic packstone). The facies types forming the
shoal-fringe facies association show a strong scattering, in particular within the range below 3 mD
permeability and 3% porosity (Fig. 10a, Table 3). A
possible explanation for the strong data scatter
could be a scale-dependent measuring inaccuracy,
which occurs specifically with the mini-permeameter.
Vugs and large bioclastic moulds may act as
direct flow-connection through the plug-sample.
Intraclastic packstones have low porosities, predominantly in the form of large vugs. This facies
Geological Society, London, Special Publications published online June 27, 2012 as doi: 10.1144/SP370.6
D. PALERMO ET AL.
Table 2. Correlation coefficients, standard deviation and arithmetic mean between porosity/permeability and
selected parameters (pore size, pore throat size and pore types), based on the semi-quantitative analysis of
thin sections
Correlation coefficient
Poro/micro
n ¼ 222
Mean
Standard deviation
Perm H/meso
n ¼ 222
Mean
Standard deviation
Perm H/macro
n ¼ 222
Mean
Standard deviation
Poro/TV
n ¼ 222
Mean
Standard deviation
Poro/SV
n ¼ 222
Mean
Standard deviation
Poro/IP
n ¼ 222
Mean
Standard deviation
Perm H/TV
n ¼ 222
Mean
Standard deviation
Perm H/SV
n ¼ 222
Mean
Standard deviation
Perm H/IP
n ¼ 222
Mean
Standard deviation
Poro/pore-throats
n ¼ 168
Mean
Standard deviation
Perm H/pore-throats
n ¼ 168
Mean
Standard deviation
Poro (%)
12.0
4.6
Perm H
32.5
58.1
Perm H
32.5
58.1
Poro (%)
12.0
4.6
Poro (%)
12.0
4.6
Poro (%)
12.0
4.6
Perm H (mD)
32.5
58.1
Perm H (mD)
32.5
58.1
Perm H (mD)
32.5
58.1
Poro (%)
12.0
4.6
Perm H (mD)
32.5
58.1
type shows the most extreme scatter combined
with the poorest reservoir properties. In contrast,
peloidal- and oncoidal pack- to grainstones have
additionally small amounts of regularly distributed
primary porosity and commonly smaller vugs.
The limited reservoir properties (Fmax ¼ 13%;
Kmax ¼ 12 mD) of the shoal-fringe facies types
could be explained by the combination of a high
mud content, occluding the primary interparticle
pore space and the dominance of mostly calcitic,
relatively leaching-resistant components (e.g. oncoids, peloids, intraclasts), and preventing the creation of vuggy porosity (Fig. 10a).
Mid-shoal facies association (crinoidal pack- to
grainstone, poorly sorted pack- to grainstone, amalgamated packstone). These facies types constitute
the largest part of the reservoir bodies and show relatively similar porosities and permeability trends
Within this group, amalgamated packstones have
Micro-pores (%)
2.2
1.8
Meso-pores (%)
6.5
3.8
Macro-pores (%)
3.2
3.8
TV (%)
2.6
2.7
SV (%)
7.1
3.0
IP (%)
2.2
3.1
TV (%)
2.6
2.7
SV (%)
7.1
3.0
IP (%)
2.2
3.1
Pore-throats (mm)
177.8
94.6
Pore-throats (mm)
177.8
94.6
20.07
0.17
0.55
0.63
0.47
0.50
0.58
0.00
0.30
0.57
0.52
the best correlation between porosity and permeability (Fig. 10b, Table 4). Amalgamated packstones
are dominated by biomouldic porosity; the high
mud content plugs the potential interparticle porespace. Therefore, permeability is probably due to the
occurrence of solution-enlarged touching vugs. Consequently, diagenetic leaching is the most important
controlling factors on pore space within this facies
type. In contrast, crinoidal pack- to grainstones have
additionally substantial amounts of interparticle porosity. They show the overall best reservoir properties
(Fmax ¼ 25%; kmax ¼ 710 mD), but also a strong
scattering of data points. The scattering within the
porosity and permeability relationship is thus mainly
controlled by the interplay of different pore-systems
(interparticle and vuggy). Nevertheless, also scaledependent measuring inaccuracy (see above) can be
observed within the poorly sorted pack- to grainstone
facies, where vugs of large shells are common, dispersing the values towards higher permeability.
Geological Society, London, Special Publications published online June 27, 2012 as doi: 10.1144/SP370.6
CARBONATE SAND BODIES IN GERMANY
Fig. 10. (a) Porosity–permeability cross-plot depicting the reservoir properties of shoal fringe facies types. Strong
scattering towards low permeability is mainly interpreted as scale-dependent measuring inaccuracy, since vugs of large
shells may act as direct flow connection through the plug sample. (b) Porosity– permeability cross-plot depicting the
reservoir properties of mid-shoal facies types. Strong scattering towards low permeability within the poorly sorted packto grainstone facies is interpreted as scale-dependent measuring inaccuracy. Amalgamated packstones are dominated by
vuggy porosity and show an excellent fit into the main cluster, whereas crinoidal pack- to grainstones tend to a stronger
scattering, but overall better reservoir properties. The stronger scattering can be explained by two different interacting
pore-systems owing to a combination of interparticle and vuggy porosity. (c) Porosity–permeability cross-plot
depicting the reservoir properties of inner-shoal facies types. Facies types of the mid-shoal facies association have a
higher fraction of oomouldic porosity and fewer solution enlarged touching vugs, which results in a scattering towards
higher porosities.
Inner-shoal facies association (oolitic grainstone,
well-sorted pack- to grainstone). These facies
types, in particular oolitic grainstones, show a
tendency towards higher porosities (Fig. 10c,
Table 5) and comparably low permeability. Vugs
and larger bioclastic moulds which interconnect
Geological Society, London, Special Publications published online June 27, 2012 as doi: 10.1144/SP370.6
D. PALERMO ET AL.
Table 3. Correlation coefficients, arithmetic mean and standard deviations of shoal-fringe facies types
Facies type
Oncoidal WP
n ¼ 14
Pelo-ooidal P
n¼7
Intraclastic P
n ¼ 12
Mean
Standard deviation
Mean
Standard deviation
Mean
Standard deviation
Porosity
Permeability
7.6
3.8
5.5
3.1
3.9
2.9
4.4
6.6
7.4
11.1
0.6
0.8
the pores are rare in this grain-dominated, commonly well sorted facies association. Thus, this
facies association is characterized by relatively
high amounts of oomouldic porosity, which is,
however, not accompanied by higher permeability
values. If biomouldic pores are present, they are
commonly surround by thick cement crusts. These
cements are commonly not affected by selective
diagenetic alteration and are an additional reason
for reduced pore connectivity. As observed in thin
sections, solution-enlarged touching vug porosity
is present but not very abundant, which could be
the result of the more stable pore-framework preventing the circulation of late diagenetic fluids and
associated porosity creation.
Distribution of reservoir properties
on different scales
The final aim of a reservoir characterization is to
predict the spatial distribution of reservoir properties on a field scale. In this outcrop study, the geometries of the facies bodies show a remarkable lateral
continuity on a kilometre-scale (for details see
Palermo et al. 2010). In contrast to one-dimensional
well data, the outcrops have the advantage that the
individual facies bodies and the presence of porosity
can be traced laterally in a qualitative way, since
millimetre- to centimetre-sized mouldic macroand meso-pores can be observed easily with a
hand lens. In order to calibrate these observations,
plug measurements and thin sections were taken
Correlation coefficient
0.34
0.14
20.25
along vertical sections. Different scales of reservoir
heterogeneities in one selected shoal body were
investigated in a hierarchical way (for details see
Seyfang 2006).
Decimetre scale (tens of centimetres)
In order to investigate the distribution of permeability on a decimetre scale, four rock columns
from the central part of a major shoal body were
cut from a 2.1 × 1.1 m-sized block from the
quarry Neidenfels. The investigated shoal portion
consists of the poorly sorted pack- to grainstone
facies and shows a variable mud content. The twodimensional permeability distribution was measured with a mini-permeameter by covering each
rock column (1 × 0.2 m) with a grid of 7 × 50
data points spaced at 2 cm intervals (Fig. 11). The
porosity values of the investigated samples range
between 15 and 23%. The values were interpolated
with the software Surfer TM using a kriging algorithm and a logarithmic colour scheme. Sedimentary
structures are not recognizable at this scale (Fig.
12) and the facies appeared as a moderately bioturbated, massive unit; therefore the interpolation
was carried out without preferential direction. The
resulting permeability is rather patchily distributed,
which can be partly related to unidirectional algorithm. Nevertheless, the values show a clear upward
increasing trend in all four rock columns. The values
follow a small-scale regressive hemicycle with an
upward decreasing fraction of matrix mud. The permeability ranges in the order of 10 mD (green) in the
Table 4. Correlation coefficients, arithmetic mean and standard deviations of mid-shoal facies types
Facies type
Crinoidal PG
n ¼ 55
Poorly sorted PG
n ¼ 103
Amalgamated P
n ¼ 63
Mean
Standard deviation
Mean
Standard deviation
Mean
Standard deviation
Porosity
Permeability
Correlation coefficient
14.2
4.0
10.5
5.1
11.8
3.4
63.2
123.2
28.2
56.0
38.7
77.7
0.26
0.51
0.60
Geological Society, London, Special Publications published online June 27, 2012 as doi: 10.1144/SP370.6
CARBONATE SAND BODIES IN GERMANY
Table 5. Correlation coefficients, arithmetic mean and standard deviations of inner-shoal facies types
Facies type
Oolitic G
n ¼ 33
Well sorted PG
n ¼ 38
Mean
Standard deviation
Mean
Standard deviation
Porosity
Permeability
Correlation coefficient
12.9
3.7
13.1
4.7
7.9
13.2
25.9
49.6
0.45
lower part and reaches up to several 100 mD (orange
to red) in the top part (Fig. 12). Thin sections document that this trend is combined with an upwards
increase in interparticle porosity.
Metre scale
For the determination of the metre-scale heterogeneities in permeability, a 2.1 × 1.1 m-sized block from
the quarry Neidenfels was investigated for sedimentary structures. The positions of the above-described
rock columns within the block are marked in
Figure 13. The individual rock columns appeared as
massive units, whereas the entire block reveals
some subtle, irregular low-angle cross-beds. They
consist of crudely graded and shell-rich laminae
above an erosive base and are intercalated into moderately bioturbated poorly sorted pack- to grainstones.
Generally, slight differences in texture, sorting and
mud content seem to follow the orientation of the
cross bedding on this larger scale. The photograph
of the block (Fig. 13) depicts the predominant inclination of the sets to the left, traced with yellow lines.
The heterogeneities in permeability are mainly
due to variations in the amount of mouldic porosity
and the shape of the leached components. Portions
with leached shells make a better contribution
to permeability than those with ooids and crinoidal
columnar plates. The shells are often connected by
solution-enlarged touching vugs and build an interconnected framework with primary interparticle
0.42
pores. Based on the observed sedimentary structures, the interpolation of permeability values
between the four measured columns was carried
out with a directional kriging algorithm, characterized by an elevated horizontal range. The resulting
permeability distribution is displayed as an overlay with logarithmic colour gradient (Fig. 13). An
important result is the systematic upward increase
in permeability that follows a small-scale regressive
hemicycle and the upward decreasing mud content.
This systematic vertical trend results in a laterally
continuous highly permeability zone in the upper
portion of the block. Note also that the lowpermeability lower part is laterally continuous, as
well as some high permeability streaks in the
middle part, which seem to be controlled by the
sedimentary structures. The permeability distribution shows that the highly permeable areas are
often vertically interconnected.
Tens of metres scale. For the determination of porosity and permeability trends within the shoal body on
a tens of metre scale, two vertical sections were
investigated along a 25 m-long outcrop wall that
was previously mapped for facies. The outcrop
wall panel, log positions and the corresponding
porosity and permeability histograms are depicted
in Figure 14. The reservoir body can be subdivided
into four layers (A –D) consisting of alternating well-sorted pack- to grainstone (Lft 2e) and
poorly sorted bioclastic pack- to grainstone (Lft 2e)
Fig. 11. Measurements of the two-dimensional permeability distribution with the mini-permeameter. Each rock
column is covered (1× 0.2 m) with a grid of 7× 50 data points.
Geological Society, London, Special Publications published online June 27, 2012 as doi: 10.1144/SP370.6
D. PALERMO ET AL.
Fig. 12. Rock-columns from the central part of a major shoal body were derived for a 2.1× 1.1 m sized slab which was
previously mapped for sedimentary structures. Permeability shows a patchy distribution but also a clear upward
increasing trend in all four rock columns.
facies types. The well-sorted pack and pack- to
grainstones (A) of the bottom part have about
10% porosity and low permeability, which can be
explained by comparably high mud content, occluding parts of potential primary interparticle porespace. In contrast, the upper well-sorted grainstone
sheet (C) has lower mud content and higher
amounts of interparticle porosity, resulting in better
porosity and permeability values. The poorly sorted
bioclastic pack- to grainstones also show textural
differences, but are additionally controlled by the
amount of biomodic porosity of leached shells.
Generally, the porosities in the investigated poorly
sorted pack- to grainstone beds (B, D) range
between 11.3 and 18.2%. The moderate permeability
in bed B ranges between 7.6 and 32.9 mD, while
the average permeability in the upper bed (D)
varies between 90.1 and 202.0 mD. Thus, lateral
changes in porosity and permeability within an individual facies body remain in the same order of
magnitude within a similar stratigraphic position.
These changes seem to correspond generally to the
vertical textural changes that seem to follow the
observed stratigraphic cycles. Moreover, also the
lateral differences in porosity and permeability
within the individual layers seem to be accompanied
by slight lateral variations in the Dunham texture.
The general occurrence of laterally continuous reservoir bodies is commonly controlled by
medium-scale cycles, since porous units occur
preferentially around their regressive maxima. In
contrast, internal porosity and permeability variations are often controlled by small-scale cycles.
The best reservoir properties are found around
small-scale regressive maximum. This effect is
most likely the result of stratigraphically controlled
variations in mud content, and additional complex
diagenetic overprints of cementation and selective
leaching (see previous section).
Regarding the entire flow unit, the lateral pattern
of the permeable layers along the same stratigraphic
cycles indicates that the observed centimetre-scale
heterogeneities owing to sedimentary structures
play a subordinate role.
Geological Society, London, Special Publications published online June 27, 2012 as doi: 10.1144/SP370.6
CARBONATE SAND BODIES IN GERMANY
Fig. 13. A 2.1×1.1 m sized block sampled from a major shoal-body in quarry Neidenfels with subtle low-angle cross
bedding (1) and the resulting permeability distribution (2). Permeability was interpolated from the rock columns along
the major sedimentary structures. Note the systematic upward increase of permeability following a small-scale
regressive hemicycle and the resulting lateral continuity of the highly permeable zone in the upper portion of the block.
Kilometre scale
Three sections through two shoal bodies, with a
lateral spacing of 1 km each, were investigated for
the distribution of facies, diagenetic changes and petrophysical properties along the depositional gradient.
The cross-section in Figure 7a displays the geometries of facies-bodies combined with porosity and
permeability values of medium- and small- scale
cycles. Within the individual facies bodies, the
porosity and permeability distribution shows gradual lateral changes. The cross-section of Figure 7b
depicts the gradational lateral changes of mud
content and cement-types, which are held as key
indicators for the relationship between the diagenetic
overprint and the resulting reservoir properties. The
overall distribution of petrographic trends and reservoir properties shows similar patterns. Both properties show laterally gradational transitions and are
strongly influenced by stratigraphic cycles. The vertical distribution of reservoir properties is mainly
controlled by medium-scale cycles, whereas the
smaller internal variations seem to be controlled by
small-scale cycles.
Petrophysical modelling
Three-dimensional geological modelling software
offers a wide range of different possibilities for the
distribution of petrophysical properties. However,
in contrast to the subsurface, outcrop analogue
studies allow a direct determination of the spatial
distributions of reservoir properties. Furthermore,
outcrop data can be used to constrain different conditioning factors and algorithms that are applied
in the modelling process. As documented above,
the lateral distribution of reservoir properties
Geological Society, London, Special Publications published online June 27, 2012 as doi: 10.1144/SP370.6
D. PALERMO ET AL.
Fig. 14. Outcrop wall panel with stratigraphic cycles, sedimentology and the reservoir properties. Lateral changes in porosity and permeability within the individual facies bodies
remain in the same order of magnitude within a similar stratigraphic position. The red arrows indicate a common trend in the Dunham texture (mud content) and the reservoir property
values, following the stratigraphic cycles. (Dunham Textures: G, grainstone; PG, pack- to grainstone; P, packstone; W, wackestone; M, mudstone. Particle size: L, lutite; S, siltite;
A, arenite; R, rudite.)
Geological Society, London, Special Publications published online June 27, 2012 as doi: 10.1144/SP370.6
CARBONATE SAND BODIES IN GERMANY
Fig. 15. Fence diagram of the 3D facies model (200× vertical exaggeration), which was used as the trend for the spatial
distribution of the reservoir properties.
within a facies body, in the same stratigraphic position, remains in the same order of magnitude for
hundreds of metres. On a kilometre scale, facies
bodies, reservoir properties and diagenetic trends
show gradual lateral changes, whereas the mostly
stratigraphic cycle-controlled vertical differences
vary commonly on a decimetre scale.
These observations within this particular setting were considered for the construction of the
geological model. The detailed, deterministic geomodel of Palermo et al. (2010) was used as a main
input for geostatistical data analysis and the distribution of reservoir properties. It covers an area
on the scale of a Middle East giant gas-field
(25 × 36 km) and provides both (a) high resolution
sequence stratigraphic reservoir layering and (b) a
detailed facies distribution (Fig. 15). It is composed
of 619 layers and 3.5 million cells. This high vertical
resolution provides the possibility to model detailed
variations of reservoir properties within individual
reservoir bodies. Furthermore, these observations
and the data distribution suggest a deterministic
approach for modelling reservoir properties. With
respect to the palaeogeographic positions and the
resulting differences in facies distribution, the
well spacing in this study is commonly similar or
smaller than the minimum dimension (commonly
dip width) of the major reservoir bodies (cf. Kerans
& Tinker 1997).
Input data
The investigation of thin sections showed that preservation and creation of pore space is restricted to
high-energy shoal and shoal-fringe facies types
and is modified by the selective diagenetic history.
Therefore, the primary facies types can be clearly
subdivided into reservoir and non-reservoir facies.
Furthermore, structural influences (e.g. fractures)
on the reservoir properties are generally negligible.
These observations and the software-dependent
necessity for a constant sampling rate led to the following steps in preparing the dataset for 3D modelling: after the insertion of the plug measurements as
point values, all portions of reservoir facies without
corresponding plug-data were set to ‘undefined’
(2999.2). Subsequently, the porosity and permeability values of all non-reservoir facies types were
set to zero. The step of data preparation was furthermore used for a quality control of the dataset. The
upscaling process transfers porosity and permeability plug measurements to the grid cells neighbouring
Geological Society, London, Special Publications published online June 27, 2012 as doi: 10.1144/SP370.6
D. PALERMO ET AL.
Table 6. Results from the sample-variogram analysis for porosity and permeability, which are the main input
parameters for petrophysical distribution algorithms (sequential Gaussian simulation and kriging)
Results from sample variograms
Porosity (F)
Variogram type
Nugget (random values)
Vertical range
Major direction
Range
Bandwidth
Number of lags
Search radius
Minor direction
Range
Bandwidth
Number of lags
Search radius
Permeability (K)
Variogram type
Nugget (random values)
Vertical range
Major direction
Range
Bandwidth
Number of lags
Search radius
Minor direction
Range
Bandwidth
Number of lags
Search radius
Shoal fringe
Mid-shoal
Inner shoal
Gaussian
27%
0.6 m
2938
3991 m
2211 m
10
9000 m
2038
2029 m
1624 m
10
6000 m
Gaussian
27%
1.5 m
2888
4702 m
2612 m
15
10230 m
1988
2843 m
15
6292 m
Gaussian
29%
1.8 m
2158
1930 m
1101 m
15
6580 m
1258
1732 m
1004 m
15
5070 m
Gaussian
53%
1.1 m
1168
920.7 m
1121.7 m
15
4254 m
268
728 m
1121 m
25
4096
Gaussian
24%
0.8 m
2238
4552.3 m
1975 m
20
12844.4 m
1258
1732 m
1610 m
15
8310.4 m
Spherical
36%
3.2 m
1498
4721 m
2585
20
14070
598
2568
1858
17
8000
the wells. In contrast to porosity, permeability tends
to change exponentially and requires a different averaging method for the upscaling process:
(1) Porosity: arithmetic average of point data
upscaled on neighbour cells.
(2) Permeability: geometric average (e.g. Warren
& Price 1961; Jensen 1991) of point data
upscaled on neighbour cells.
However, owing to the extraordinary high vertical
resolution (average 0.1 m) of the 3D grid, which
corresponds to the sampling rate of the plug
measurements, the upscaling had no considerable
impact on the quality of the model.
Geostatistical data analysis
Variograms are geostatistical models for the 3D distribution of reservoir heterogeneity and fundamental
input parameters for the most common modelling
algorithms, for example sequential Gaussian simulations and kriging. Variograms are defined as the
semi-variance of the difference between field values
at two locations across realizations of the field
(Cressie 1993). In order to establish the necessary
input parameters, a geostatistical variogram analysis
(Gringarten & Deutsch 2001) was carried out on the
complete dataset with the original log data spaced
between 10 m and a few kilometres. Based on the
observations described above, the reservoir properties were analysed for each facies type and each
association with various lag-sizes, ranges, and so
on. A reasonable variogram model could only be
found for the facies associations, by using relatively
large bin sizes and large ranges that are in the order
of the observed facies changes (Palermo et al.
2010). The results and the used parameters are summarized in Table 6. The vertical semi-variance was
determined within decimetre-sized lags and the
resulting ranges for porosity result in 0.6 m for shoalfringe, 1.6 m for mid-shoal and 1.8 m for inner-shoal
facies associations. The lateral semi-variance of porosity results in Gaussian variogram models for all
three facies associations with high random values
(Nugget effect, after Eisenberg et al. 1994) slightly
below 30% in all three cases. The lateral ranges are
remarkably high and reach values in the major direction of 1930 m for the inner-shoal, 3991 m for shoal
Geological Society, London, Special Publications published online June 27, 2012 as doi: 10.1144/SP370.6
CARBONATE SAND BODIES IN GERMANY
fringe and 4702 m for mid-shoal facies associations.
The semi-variance of the permeability is generally
characterized by higher random values in the shoal
fringe (53%) and the inner shoal (36%) facies types,
whereas the random factor of the mid-shoal (24%)
is comparably low. Compared with the porosity, the
computed permeability ranges are significantly
lower for the shoal fringe, similar for the mid-shoal
and higher for the inner-shoal facies associations.
The results of the variogram analysis for both porosity and permeability are summarized in Table 6.
Effect of input data
As described in the sections above, the major portion
of the effective porosity within the system was
created at an early stage and is mostly linked to
either primary porosity or cycle-controlled early
diagenetic leaching. In the upper described qualitative and quantitative outcrop observations, two major
scales of lateral heterogeneities could be observed:
(a) small-scale heterogeneities on centimetre to decimetre scale; and (b) large-scale, smooth transitions
in the order of the lateral facies association extensions. A reasonable variogram model could only be
found for the largest scale. The intermediate scale,
tens to hundreds of metres and the centimetre scale
did not provide enough data to make a sufficient
variogram analysis. According to the previous
observations, the remarkably high lateral porosity
ranges can be related to the subtle gradational transitions of mud content along the depositional slope,
reflecting the depositional energy of a flat and very
gently inclined epeiric carbonate ramp.
Owing to the continuous character of the porosity that can be traced visually along the outcrop
walls, the predominantly Gaussian semi-variogram
models and the resulting ranges were considered
as the most probable scenario. However, the possibility of other scenarios cannot be generally
excluded. Follow-up sampling as performed by,
for example, Pranter et al. (2005) is planned to
better quantify the small-scale heterogeneities.
Conditioning and modelling algorithms
In order to compare the impact of stratigraphy and
facies (Palermo et al. 2010) on the distribution of
reservoir properties, three different approaches
were used for the petrophysical modelling.
(1) Unconditioned. The grid used here is a relatively simple 3D cube that was subdivided
into 180 conformable layers. Only the top
and base of the model were defined by stratigraphic surfaces. Unrelated to facies or stratigraphy, the modelled reservoir properties
were directly interpolated between the data
points using different algorithms.
(2)
Conditioned to stratigraphic cycles. In this
approach, the petrophysical properties were
conditioned to the zones based on stratigraphic cycles. Each reservoir bearing zone
was subdivided into eight layers and the
average vertical resolution was around 10 cm.
Therefore, vertical variations of reservoir properties within single reservoir units could also
be displayed with a high level of detail.
(3) Conditioned to stratigraphic cycles and facies
associations. In this scenario the entire information from the geological model (Fig. 17)
was used to model the spatial distribution of
reservoir properties. Each reservoir facies
association was modelled individually, and
tight facies types were set to zero. To avoid
unrealistically sharp contacts between the
lateral facies boundaries the properties were
slightly smoothened.
Three different modelling algorithms were tested
for the distribution of the reservoir properties in
the upper described scenarios and compared with
the outcrop observations (Table 7):
(1) Moving average. This interpolation algorithm
is based on the average values of the input data
and weights the distribution according to the
distance from data points without a spatial
preference (Chou 1975).
(2) Kriging. This is a deterministic interpolation
method developed by Matheron (1963) and
D. G. Krige. The algorithm distributes a property with a spatial preference that has to be
defined by an input variogram. The estimated
variables are then calculated by the linear
combination of the input data points. In contrast to the moving average, the kriging algorithm uses a weighting function and honours
both distance and direction.
(3) Sequential Gaussian simulation (SGS). This
variogram based stochastic simulation technique is based on kriging and allows multiple equiprobable realizations of a property
(Alabert 1987; Deutsch & Journel 1992).
Kriging calculates the weighted mean and
the standard deviation at each estimated
point. SGS, however, represents the variable
as a random deviate from the Gaussian
normal distribution, which represents the
input data and trends better than most of the
kriging algorithms.
Porosity distribution with sequential
Gaussian simulation (Fig. 16)
Porosity – unconditioned. According to the randomity of the algorithm, the distribution of porosity
appears very patchy and the general geological
Porosity
Permeability
Sequential Gaussian simulation
Porosity
Permeability
Kriging
Porosity
Permeability
Vertical range
1m
1m
Point weighting
Exponent: 4
Exponent: 4
Variogram settings
Parameters from variogram analysis
Parameters from variogram analysis
Variogram settings
Parameters from variogram analysis
Parameters from variogram analysis
Direction
Distribution
Trends
Follow layers
Follow layers
Transformations
Normal score
Normal score
Transformations
Normal score
Normal score
normal
Logarithmic
Distribution
Linear
Logarithmic
Distribution
Linear
Logarithmic
Facies
Facies, porosity
Trends
Facies
Facies, porosity
Trends
Facies
Facies, porosity
D. PALERMO ET AL.
Moving average
Geological Society, London, Special Publications published online June 27, 2012 as doi: 10.1144/SP370.6
Table 7. Parameters and algorithms used for the spatial distribution of reservoir properties
Geological Society, London, Special Publications published online June 27, 2012 as doi: 10.1144/SP370.6
CARBONATE SAND BODIES IN GERMANY
Fig. 16. Different porosity distributions with the Sequential Gaussian simulation algorithms conditioned to: (a) a
simple layering – the porosity values are vertically scattered and form laterally simple bulls eye patterns; (b)
stratigraphic cycles – vertical distribution matches better with the outcrop observations, although unconnected bulls
eyes are still present; (c) stratigraphic cycles and facies – the smoothened distribution shows the best match with the
outcrop continuities.
pattern is not recognizable. Additionally, the reservoir bodies are relatively small with low lateral continuity. Both kriging and moving average provided
better results than SGS in this scenario.
Porosity – conditioned to stratigraphic cycles. The
resulting reservoir bodies follow the stratigraphic
trends and have acceptable dimensions. However,
compared with the results produced with the
moving average algorithm, many artefacts are
visible, especially in the tight middle part.
Porosity – conditioned to cycles and facies associations. The modelled porosities follow the stratigraphic trends and the outline of the facies bodies.
Compared with the results with the moving average
Fig. 17. Different permeability distributions with the Sequential Gaussian simulation algorithms conditioned to (a) a
simple layering – the porosity values are vertically scattered and laterally isolated flow units, (b) stratigraphic cycles –
more realistic vertical connectivity, but the flow units are laterally still isolated, and (c) stratigraphic cycles and facies –
the smoothened distribution shows a good match with the outcrop observations.
Geological Society, London, Special Publications published online June 27, 2012 as doi: 10.1144/SP370.6
D. PALERMO ET AL.
algorithm and kriging, the reservoir bodies modelled
with a slightly smoothed sequential Gaussian Simulation show a better connectivity and more reasonable dimensions. Analysing this final scenario, the
resulting net/gross ratio of the modelled Upper
Muschelkalk area amounts to 7%, with a theoretical
storage capacity of 419 490 000 m3.
Permeability distribution with sequential
Gaussian simulation (Fig. 17, Fig. 18)
Permeability – conditioned to porosity. Like the
porosity, the distribution of permeability appears
patchy and the general geological pattern is not
recognizable. The reservoir bodies are comparably
small and highly compartmentalized.
Permeability – conditioned to stratigraphic cycles
and porosity. Comparable to the distribution of porosity, the resulting flow units follow the general
stratigraphic trends. Although there are no artefacts
in the middle part visible, the distribution of permeability remains patchy, especially in the upper part.
Permeability – conditioned to cycles, facies associations and porosity. The modelled permeability
follows the stratigraphic trends and the outline of
the facies bodies. Both dimensions and connectivity
of the flow units seem to be more reasonable in
this case.
Modelling results
The outcrop analogue studies of Palermo et al.
(2010) could show that the shoal facies associations
show remarkable lateral extensions of up to several
tens of kilometres whereas vertical facies changes
range within the order of decimetres. Another part
of the study shows that lateral facies changes are
characterized by very subtle and gradual transitions
within the range of up to a few kilometres. Therefore, the Palermo et al. (2010) study suggests
Fig. 18. Filtered permeable reservoir bodies (cut-off 1 mD) distributed with sequential Gaussian simulation,
conditioned to stratigraphic cycles and facies and processed with a smoothing algorithm. All three hierarchies of
stratigraphic cycles had an impact on quality and presence of the flow units: (a) large scale cycle – controls the lateral
extension, retro- and progradation, (b) medium scale cycles – control the stratigraphic presence of a flow unit, and
(c) small-scale cycles – control the body internal vertical heterogeneities. Furthermore the distribution of palaeohighs
and -lows is important for the general localization of the reservoir bodies.
Geological Society, London, Special Publications published online June 27, 2012 as doi: 10.1144/SP370.6
CARBONATE SAND BODIES IN GERMANY
encoding facies as a continuous rather than discrete
property. The thin section observations and petrophysical measurements confirm theses observations.
Mud content in the matrix changes laterally and vertically in a gradual way and reflects the depositional
energy of the epeiric system with a very gentle
depositional gradient. The outcrop observations
above show that mud content has a significant impact
on the presence of primary porosity. However, early
diagenetic mouldic porosity also seems to increase
with increasing matrix mud content. The petrophysical investigations on multiple scales showed two
major scales of reservoir heterogeneities:
(1)
(2)
Small-scale heterogeneities on centimetre to
several decimetre scales mostly controlled by
sedimentary structures and diverse mouldic
pore types from different components were
observed. Similar small-scale heterogeneities
have been described by several studies (e.g.
Jennings et al. 1998; Pranter et al. 2005)
However, the quantity of samples in this
study was insufficient to perform a solid variogram analysis that could quantify the 3D lateral
ranges of the small-scale heterogeneities.
On a large scale, relatively smooth transitions on the order of a few kilometres were
observed and could be confirmed by the variogram analysis. However, the resulting lateral
long range features are significantly higher
compared with the dolomitized platform
carbonates investigated by Jennings et al.
(1998), which can reach up to 800 m. Apart
from the different depositional settings and
diagenetic histories, which make a direct
comparison difficult, the possibility of additional nested heterogeneities as observed and
described by Pranter et al. (2005) cannot be
excluded. At least one smaller heterogeneity
scale is most likely hidden within the high
random effect and could be an interesting
subject for further investigations. Extensive
lateral sampling is foreseen for a future
study to investigate the presence of additional
meso-scale heterogeneities.
with subtle gradational transitions. The main part
of the porosity in the shoal bodies has been
created at an early stage and is mostly linked to
either cycle-controlled (a) primary porosity triggered by a systematically changing mud content or
(b) early diagenetic leaching around maximum
regressions.
Two major scales of lateral heterogeneities could
be observed: (a) small-scale heterogeneities on
centimetre to decimetre scale, mainly related to
sedimentary structures and different mouldic pore
types which are represented by high random
values (30%) within the geostatistical variogram
model; and (b) large-scale heterogeneities of up to
4700 m lateral range. The remarkably high lateral
porosity and permeability ranges can be partly
related to subtle gradational transitions of mud
content along the depositional slope, which reflect
the depositional energy of a flat and very gently
inclined epeiric carbonate ramp.
Different petrophysical modelling scenarios
could show the impact and necessity of geological
constraints for the geostatistical distribution of porosity and permeability. The following factors help
in predicting and modelling the spatial distribution
of reservoir properties within the Upper Muschelkalk, which may be transferred to similar epeiric
carbonate reservoirs (Fig. 15):
(1)
(2)
Conclusions
This outcrop analogue study documents the reservoir properties in the Upper Muschelkalk formation
as an outcrop analogue for epeiric carbonates.
Average porosity is 9–23%, and average permeability is 21 –700 mD. The net/gross ratio of
the Upper Muschelkalk amounts to 7%, with a
theoretical storage capacity of 419 490 000 m3.
Laterally, units with high porosity and permeability show a kilometre-scale lateral continuity
(3)
Facies associations. The reservoir properties
show a close relationship to the lithofacies
and the associated matrix mud content.
Porous facies types are restricted to the highenergy shoal facies, whereas both muddy
inner and outer ramp facies are commonly
tight since the muddy matrix protects the sediment from diagenetic fluids and associated
mouldic porosity creation. Reservoir heterogeneities within the shoal bodies can be
mainly related to the mud content and early
diagenesis.
Statigraphic cycles. All three hierarchies of
stratigraphic cycles have an impact on quality
and presence of the flow units: (a) large-scale
cycle – controls the lateral extent, as well as
retro- and pro- gradation of the reservoir
bodies; (b) medium-scale cycles – control the
stratigraphic presence of the reservoir facies
associations; and (c) small-scale cycles –
control the mud content and early mouldic
porosity creation in facies types with primary
porosity. which are considered as key drivers
for systematic vertical changes in porosity
and permeability within the reservoir bodies.
Palaeo-relief. Gross volume and dimensions
of the reservoir bodies seem to be mainly controlled by the combination of both stratigraphic cycles and a subtle palaeo-relief,
Geological Society, London, Special Publications published online June 27, 2012 as doi: 10.1144/SP370.6
D. PALERMO ET AL.
induced by slight differential subsidence of
inherited structural grains. In particular,
palaeo-highs are important for the presence
of reservoir facies.
Authors
Denis Palermo studied geology at the University of
Tübingen and completed his MSc thesis in 2004 on
carbonate reservoir sedimentology at the Nederlandse Aardolje Matschapij (NAM) in the Netherlands. He then joined the Sedimentary Geology
Group at the University of Tübingen to do a Ph.D.
on outcrop analogue modelling and reservoir characterization in cooperation with Eni E&P. In
2007 he joined Eni E&P as sedimentologist and
member of the Eni E&P carbonate research team.
Thomas Aigner studied geology at the universities of Stuttgart, Tübingen, Reading and Miami.
After working for 6 years with Shell Research, he
has been a professor and head of the sedimentary
geology group at the University of Tübingen since
1991. His research group focuses on reservoir
geology, carbonate reservoir characterization and
3D modelling. He was an AAPG European Distinguished Lecturer in 1996. In 2007–2008 he spent
a sabbatical with Petroleum Development Oman
and Qatar Shell.
Björn Seyfang studied geology at the University
of Tübingen and finished his M.Sc. on outcrop analogue studies for carbonate shoal reservoirs in 2006.
In 2009, he completed his Ph.D. on facies and reservoir modelling at the Sedimentary Geology Group
of Tübingen University, initiated by GDF-SUEZ
and conducted in cooperation with ExxonMobil,
Wintershall and RWE-Dea. In September 2009, he
joined Total as an operations geologist.
Sergio Nardon graduated in geology at the University of Trieste, Italy, in 1982 and joined Eni
(former Agip) in 1984 as clastic sedimentologist.
His experience spans from fractured reservoir
characterization to carbonate sedimentology and
he has worked on several projects in Middle East,
Europe and North Africa. His main interests are
the integration of carbonate architecture, outcrop
data and geological concepts into the 3D environment. He is currently technical leader of the research
and development team at the geology department of
Eni E&P.
This work is part of an integrated research and development project funded by Eni E&P. We want to express our
thanks to Eni Management for their support and the permission to publish. We are grateful for discussions with
many colleagues working on the German Muschelkalk,
notably H. Hagdorn, T. Simon, M. Urlichs and
R. Borkhataria. Assistance was provided by members
of the Sedimentary Geology group at the University of
Tübingen (A. Allgöwer, E. Dmitrieva, M. Looser, A.
Schmid-Röhl, C. Schneider, M. Zeller) and A. Satterley,
B. White and K. Mair from Eni E&P. P. Jeiseke produced the thin sections. Several quarry companies, for
example Schön & Hippelein, Hohenloher Schotterwerke,
Schotterwerk Schuhmann, J. Heumann and SHB Schotterwerke Hohenlohe generously allowed access to their quarries. The Geological Survey of Baden-Württemberg and
Bayern supplied us with borehole cores and well data.
We wish to thank A. Glocke and Schlumberger for
the access to Petrel (TM of Schlumberger) and to
A. Henriette (ALT) for access to WellCAD.
References
Aigner, T. 1985. Storm Depositional Systems. Dynamic
Stratigraphy in Modern and Ancient Shallow Marine
Sequences. Lecture Notes in Earth Sciences, Vol. 3,
Springer, Berlin.
Aigner, T. 1999. Dynamische Stratigraphie des Oberen
Muschelkalks am Beispiel Südwestdeutschlands,
115–128.
Aigner, T., Braun, S., Palermo, D. & Blendinger, W.
2007. 3D geological modelling of a carbonate shoal
complex: reservoir analogue study using outcrop
data. First Break, 25, 65–72.
Alabert, F. 1987. The practice of fast conditional simulations through the LU decomposition of the covariance matrix. Mathematical Geology, 19, 369–386.
Allgöwer, A. 2006. Stratigraphy, petrophysics and
facies analysis of epeiric carbonates in the Upper Muschelkalk. An outcrop analogue study for skeletal/
oolitic carbonate sand bodies in the Middle East
(Upper Muschelkalk, South German Basin). Diploma
thesis, University of Tübingen.
Azerêdo, A. C. 1998. Geometry and facies dynamics
of Middle Jurassic carbonate ramp sandbodies. In:
Wright, V. P. & Burchette, T. P. (eds) Carbonate
Ramps. Geological Society, London, Special Publications, 149, 281– 314.
Bachmann, G. H . 1973. Die karbonatischen Bestandteile
des Oberen Muschelkalks (Mittlere Trias) in SüdwestDeutschland und ihre Diagenese. Arbeiten des Instituts für Geologie und Paläontologie der Universität
Stuttgart, 68, 1 –99.
Borgomano, J. R. F., Masse, J.-P. & Al Maskiry, S.
2002. The lower Aptian Shuaiba carbonate outcrops
in Jebel Akhdar, northern Oman: Impact on static
modelling for Shuaiba petroleum reservoirs. AAPG
Bulletin, 86, 1513–1529.
Braun, S., 2003. Quantitative analysis of carbonate sandbodies: outcrop analogue study from an epicontinental
basin (Triassic Germany). PhD thesis, University of
Tübingen.
Burchette, T. P., Wright, V. P. & Faulkner, T. J.
1990. Oolitic sandbody depositional models and
geometries, Mississippian of southwest Britain: implications for petroleum exploration in carbonate ramp
settings. Sedimentary Geology, 68, 87–115.
Cavallo, L. J. & Smosna, R. 1997. Predicting porosity
distribution within oolitic tidal bars. In: Kupecz,
J. A., Gluyas, J. & Bloch, S. (eds) Reservoir
Quality Prediction in Sandstones and Carbonates.
Geological Society, London, Special Publications published online June 27, 2012 as doi: 10.1144/SP370.6
CARBONATE SAND BODIES IN GERMANY
American Association of Petroleum Geologists, Tulsa,
OK, Memoirs, 69, 211–229.
Choquette, P. W. & Pray, L. C . 1970. Geologic nomenclature and classification of porosity in carbonates.
Bulletin AAPG, 54, 107–250.
Chou, Y. 1975. Statistical Analysis with Business and
Economic Applications. Elsevier, Amsterdam.
Cressie, N. 1993. Statistics for Spatial Data. Wiley,
New York.
Dercourt, J., Ricou, L. E. & Vrielynck, B. (eds) 1993.
Atlas Tethys Paleoenvironmental Maps. GauthierVillars, Paris.
Deutsch, C. V. & Journel, A. G. 1992. GSLIB: Geostatistical Software Library and user’s Guide. Oxford
University Press, New York.
Dmitrieva, E. 2006. Facies distribution, sequence stratigraphy and poroperm analysis of carbonate shoal
bodies in the Upper Muschelkalk (middle Jagst
valley, South German Basin). An outcrop analogue
study for carbonate reservoirs in the Middle East.
Diploma thesis, University of Tübingen.
Eisenberg, R. A., Harris, P. M., Grant, C. W., Goggin,
D. J. & Conner, F. J. 1994. Modeling reservoir heterogeneity within outer ramp carbonate facies using an
outcrop analog, San Andres formation of the Permian
Basin. AAPG Bulletin, 78, 1337–1359.
Flügel, E. 2004. Microfacies of Carbonate Rocks. Analysis, Interpretation, Application. Springer, Heidelberg.
Gawthorpe, R. L. & Gutteridge, P. 1990. Geometry
and Evolution of Platform-Margin Bioclastic Shoals,
Late Dinantian (Mississippian). International Association of Sedimentologists, Derbyshire, Special Publications, 9, 39–54.
Geyer, O. F. & Gwinner, M. P. 1991. Geologie von
Baden-Württemberg. E. Schweizerbart’sche Verlagsbuchhandlung, Stuttgart
Grammer, G. M., Harris, P. M. & Eberli, G. P. 2004.
Integration of outcrop and modern analogs in reservoir modeling. American Association of Petroleum
Geologists, Tulsa, OK, Memoirs, 80.
Grant, C. W., Goggin, D. J. & Harris, P. M. 1994.
Outcrop analog for cyclic-shelf reservoirs, San
Andres formation of Permian Basin: stratigraphic framework, permeability distribution, geostatistics, and
fluid-flow modeling. AAPG Bulletin, 78, 23– 54.
Gringarten, E. & Deutsch, C. V. 2001. Variogram
interpretation and modeling. Mathematical Geology,
33, 507–534.
Hagdorn, H. 1991. The Muschelkalk in Germany – An
introduction. In: Hagdorn, H. (ed.) Muschelkalk, A
Field Guide. Goldschneck, Korb.
Hagdorn, H. & Simon, T. 1988. Geologie und Landschaft
des Hohenloher Landes. Thorbecke, Sigmaringen.
Handford, R. C. 1988. Review of carbonate sand-belt
deposition of ooid grainstones and application to
Mississippian Reservoir, Damme Field, Southwestern
Kansas. AAPG Bulletin, 72, 1184–1199.
Harris, P. M. & Kowalik, W. S. 1994. Satellite images
of carbonate depostional settings, examples of
reservoir- and exploration-scale geologic facies variation. AAPG Methods in Exploration Series, 11, 147.
Jennings, W. J. Jr. 2000. Spatial Statistics of Permeability
Data from Carbonate Outcrops of West Texas and New
Mexico: implications for Improved Reservoir
Modelling. Bureau of Economic Geology. Report of
Investigations no. 258. University of Texas.
Jennings, J. W., Ruppel, S. C. & Ward, W. B. 1998.
Geostatistical analysis of petrophysical data and
modelling of fluid-flow effects in carbonate outcrops.
Society of Petroleum Engineers Annual Technical
Conference and Exhibition, New Orleans, 27–30
September, SPE Paper 49025.
Jensen, J. L. 1991. Use of the geometric average for effective permeability estimation. Mathematical Geology,
23, 833 –840.
Kerans, C. & Tinker, S. W. 1997. Sequence Stratigraphy
and Characterization of Carbonate Reservoirs.
OSEPM Short Course Notes, 40.
Kittridge, M. G., Lake, L. W., Lucia, F. J. & Fogg, G. E.
1990. Outcrop/subsurface comparison of heterogeneity in the San Andres Formation. Society of Petroleum
Engineers Formation Evaluation, 5, 233– 240.
Kostic, B. 2001. Sedimentäre Strukturen, Fazies und
Poroperm-Eigenschaften in ausgewählten ‘Karbonatsanden’: Quaderkalk, Oberer Muschelkalk. Diploma
thesis, University of Tübingen.
Kostic, B. & Aigner, T. 2004. Sedimentary and poroperm
anatomy of shoal-water carbonates (Muschelkalk,
South German Basin): an outcrop-analogue study of
interwell spacing scale. Facies, 50, 113 –131.
Looser, M. 2006. Facies and poroperm parameters of
epeiric carbonates in the Upper Muschelkalk of the
middle Kocher valley. Diploma thesis, University of
Tübingen.
Lucia, F. J. 1983. Petrophysical parameters estimated
from visual description of carbonate rocks: a field
classification of carbonate pore space. Journal of
Petroleum Technology, 3, 626–637.
Lucia, F. J. 1999. Carbonate Reservoir Characterization.
Springer, Berlin.
Matheron, G. 1963. Principles of Geostatistics. Economic Geology, 58, 1266–1963.
Ockert, W. 1988. Lithostratigraphie und Fossilführung
des Trochitenkalks (Unterer Hauptmuschelkalk, mo1)
im Raum Hohenlohe. In: Hagdorn, H. (ed.) Neue
Forschungen zur Erdgeschichte von Crailsheim. Sonderbände der Gesellschaft für Naturkunde in Württemberg, Goldschneck, Korb, 1, 43–69.
Palermo, D., Aigner, T., Geluk, M., Poppelreiter, M.
& Pipping, K. 2008. Reservoir potential of a lacustrine
mixed carbonate/siliciclastic gas reservoir: The Lower
Triassic Rogenstein in the Netherlands. Journal of
Petroleum Geology, 31, 61–96.
Palermo, D., Aigner, T., Nardon, S. & Blendinger, W.
2010. 3D facies modelling of carbonate sand bodies:
outcrop analogue study in an epicontinental basin (Triassic, SW Germany). AAPG Bulletin, 94, 475–512.
Pranter, M. J., Hirstius, C. B. & Budd, D. A. 2005.
Scales of lateral petrophysical heterogeneity within
dolomite lithofacies as determined from outcrop
analogs: Implications for 3-D reservoir modelling.
AAPG Bulletin, 89, 645 –662.
Pranter, M. J., Reza, Z. A. & Budd, D. A. 2006.
Reservoir-scale characterization and multiphase fluidflow modelling of lateral petrophysical heterogeneity
within dolomite facies of the Madison Formation,
Sheep Canyon and Lysite Mountain, Wyoming,
USA. Petroleum Geoscience, 12, 29–40.
Geological Society, London, Special Publications published online June 27, 2012 as doi: 10.1144/SP370.6
D. PALERMO ET AL.
Qi, L., Carr, T. R. & Goldstein, R. H. 2007. Geostatistical three-dimensional modelling of oolite shoals,
St. Louis Limestone; southwest Kansas. AAPG Bulletin, 91, 69–96.
Rankey, E. C., Riegl, B. & Steffen, K. 2006. Form,
function, and feedbacks in a tidally dominated ooid
shoal, Bahamas. Sedimentology, 53, 1191–1210.
Ruf, M. 2001. Facies distribution, petrophysics and
mapping of selected carbonate sand bodies in the
Upper Muschelkalk, South German Basin: a reservoir
analogue investigation. Diploma thesis, University of
Tübingen.
Ruf, M. & Aigner, T. 2004. Facies and poroperm characteristics of a carbonate shoal (Muschelkalk, South
German Basin): A reservoir analogue investigation.
Journal of Petroleum Geology, 27, 215– 239.
Savary, B. & Ferry, S. 2004. Geometry and petrophysical parameters of a calcarenitic turbidite lobe
(Barremian-Aptian, Pas-de-la-Cluse, France). Sedimentary Geology, 168, 281– 304.
Senger, R. K., Lucia, F. J., Kerans, C. & Ferris, M. A.
1991. Dominant control on reservoir-flow behavior in
carbonate reservoirs as determined from outcrops
studies. In: Burchfield, T. E. & Wesson, T. C.
(eds) Third International Reservoir Characterization
Technical Conference. National Institute for Petroleum
and Energy Research and US Department of Energy,
Tulsa, OK, l, 107–150.
Seyfang, B. 2006. Sedimentary and Poroperm Heterogeneities of Carbonate Shoals from Centimetre- to
Kilometre-Scales. An Outcrop Analogue Study For
Skeletal/Oolitic Carbonate Sand Reservoirs in the
Middle East. Diploma thesis, University of Tübingen.
Skupin, K. 1969. Lithostratigraphische Profile aus dem
Trochitenkalk des Neckar-Jagst-Kocher-Gebietes.
Jahresberichte und Mitteilungen des Oberrheinischen
Geologischen Vereins, 63, 1– 173.
Ulrichs, M. & Mundlos, R. 1987. Revision der Gattung
Ceratites de Haan 1825 (Ammonoidea, Mitteltrias),
I.Stuttgarter Beiträge zur Naturkunde, 128, 36.
Ulrichs, M. & Mundlos, R. 1990. Zur CeratitenStratigraphie im Oberen Muschelkalk (Mitteltrias)
Nordwürttembergs. Jahreshefte der Gesellschaft für
Naturkunde in Württemberg, 145, 59–74.
Vollrath, A. 1938. Zur Stratigraphie und Bildung des
Oberen Hauptmuschelkalks in Mittel- und Westwürttemberg. Mitteilungen aus dem MineralogischGeologischen Institut der TH Stuttgart, 33, 69–80.
Vollrath, A. 1955. Zur Stratigraphie des Hauptmuschelkalks in Württemberg. Jahreshefte des
Geologischen Landesamtes Baden-Württemberg, 1,
79–168.
Vollrath, A. 1957. Zur Entwicklung des Trochitenkalks
zwischen Rheintal und Hohenloher Ebene. Jahreshefte
des Geologischen Landesamtes Baden-Württemberg,
2, 119–134.
Vollrath, A. 1958. Beiträge zur Paläogeographie des
Trochitenkalks in Baden-Württemberg. Jahreshefte
des Geologischen Landesamtes Baden-Württemberg,
3, 181–194.
Vollrath, A. 1970. Ein vollständiges Profil des oberen
Muschelkalks und ein neues Mineralwasser bei
Ummenhofen, Gemeinde Untersontheim, Landkreis
Schwäbisch Hall. Jahresberichte und Mitteilungen des
Oberrheinischen Geologischen Vereins, 52, 133–148.
Wagner, G. 1913. Beiträge zur Stratigraphie und Bildungsgeschichte des Oberen Hauptmuschelkalks und
der Unteren Lettenkohle in Franken. Geologische
und Paläontologische Abhandlungen, 2, 31–180.
Warren, J. E. & Price, H. S. 1961. Flow in heterogeneous
porous media. SPE Journal, 1, 153–169.
Ziegler, P. A. 1990. Geological Atlas of Western and
Central Europe, Shell International Petroleum
Maatschappij. 2nd edn. Elsevier, Amsterdam.