You are on page 1of 16

Marine and Petroleum Geology 99 (2019) 45–60

Contents lists available at ScienceDirect

Marine and Petroleum Geology


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

Research paper

Quantitative controls on the regional geometries and heterogeneities of the T


Rayda to Shu'aiba formations (Northern Oman) using forward stratigraphic
modelling
Marya Al-Salmia,1, Cédric M. Johna,∗, Nicolas Hawieb
a
Department of Earth Science and Engineering and Qatar Carbonate and Carbon Capture Research Centre, Imperial College London, United Kingdom
b
Beicip-Franlab, Rueil-Malmaison, France

ABSTRACT

The complex geometry of carbonate systems is influenced by a multitude of physical as well as biological processes. The Lower Cretaceous carbonates of Northern
Oman are characterised by a variability of regional-scale geometries with expected vertical and lateral facies variations. The main environmental and tectonic
controls acting on the depositional processes of the Lower Cretaceous ramps and platforms through space and geological time (in 4 dimensions) are only partially
understood. In this study, we use a 4D DionisosFlow Forward Stratigraphic Modelling (FSM) approach to explore the role of: (i) eustasy; (ii) subsidence; (iii) initial
paleobathymetry, and (iv) wave energy, to generate carbonate stacking patterns and heterogeneities. Carbonate production was maintained constant through
deposition. Multi-disciplinary and multi-scale datasets were used (i.e. seismic, well and field data) to constrain the FSM input parameters and sensitivity analysis was
carried out to validate or refute some depositional model hypotheses. Results show that basement topography and eustasy have the greatest influence on the
progradational geometries and the lateral continuity of clinoform architectures during the Tithonian to Valanginian second-order super-sequence. In the Valanginian
to Aptian super-sequence, subsidence was the primary control for the observed aggradational stacking pattern. Lateral and vertical stacking of carbonate lithologies,
textures as well as facies are thus apprehended through this FSM approach, leading to a better assessment of petroleum systems elements as reservoir, seal and trap.

1. Introduction data appear to be much more complex. They are characterised by dif-
ferentiated platforms with clear breaks in the slope and relatively steep
During Jurassic and Early Cretaceous times, an extensive carbonate clinoforms (Droste, 2010). Seismic interpretation of the Early Cretac-
platform covered the eastern part of the Arabian Plate representing a eous sequence of North and Central Oman highlighted complex pro-
period of broad marine transgression (Murris, 1980). Several regional- gradation geometries (Droste and Van Steenwinkel, 2004). Moreover,
scale investigations were conducted to assess the stacking patterns and the Early Cretaceous platform margin, situated just north of the present-
lateral geometries of the Lower Cretaceous carbonates, combining both day Oman Mountains (Pratt and Smewing, 1990), present complex
outcrop and subsurface data (Masse et al., 1997; Pratt and Smewing, vertical and lateral facies mosaic and stacking patterns during the
1993; Van Buchem and Razin, 1996 and; Hillgärtner et al., 2003). The Berriasian to the Aptian interval (Glennie et al., 1974 and Hillgärtner
Lower Cretaceous reservoirs have been producing in many fields along et al., 2003). For example, the Lower Kahmah Group (Rayda, Salil and
Northern Oman and extend towards the northern part of the Ghaba Salt Habshan) is restricted to areas north of ca. 21° latitude line. South of
Basin (Fig. 1) (Pratt and Smewing, 1993 and Van Buchem and Razin, this line the Lekhwair Formation lies directly in the Jurassic or older
1996). Examples of such fields include the Yibal, Al Huwaisah, Lekh- strata (Haan et al., 1990). The Upper Kahmah (Lekhwair, Kharaib and
wair and Safah fields (Van Buchem et al., 2002), where the main pro- Shu'aiba formations) covers the southern area, but south of ca. 20° N, it
ducing reservoir units are the shallow marine, peri-reefal Shu'aiba is progressively cut out beneath the basal Wasia Group (Glennie et al.,
Formation and the Kharaib Formation (Litsey et al., 1986). 1974).
The interior part of these carbonate ramp systems is usually imaged The depositional patterns seen in the Lower Cretaceous of Oman are
as widespread shallow water sectors, where carbonates mainly aggrade. largely controlled by the interplay of subsidence, eustasy, sediment
Due to their negligible topography, deposition is generally assumed to supply and transport, but these controls cannot easily and in-
involve broad facies belts with laterally continuous, low-angle facies dependently be constrained by looking at their product (i.e. the re-
(Droste, 2010). However, ramp systems interpreted through 3D seismic sulting stratigraphic stacking pattern) (Hillgärtner et al., 2003). By


Corresponding author.
E-mail address: cedric.john@imperial.ac.uk (C.M. John).
1
Now at: Petroleum Development Oman, Oman.

https://doi.org/10.1016/j.marpetgeo.2018.09.030
Received 28 May 2018; Received in revised form 19 September 2018; Accepted 23 September 2018
Available online 26 September 2018
0264-8172/ © 2018 Elsevier Ltd. All rights reserved.
M. Al-Salmi et al. Marine and Petroleum Geology 99 (2019) 45–60

Fig. 1. A) Location map showing the study area in northern Oman (red box). B) Simplified tectonic provinces in northern Oman. Highlighted are some outcrop
locations in Jebel Akhdar and some major oil fields producing the Shu'aiba reservoir that are mentioned in the text. Modified from Glennie et al. (1974) and
Borgomano et al. (2002). (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

contrast, forward stratigraphic modelling provides a quantitative way


to generate several realisations of carbonate deposition by simulating
the sedimentological and stratigraphic processes in time and space
(several tens to hundreds of kilometres) (Shafie and Madon, 2008), as
well as, identifying key factors, like production rate, subsidence and
eustasy that control the architecture of carbonates (Hawie et al., 2015b;
Huang et al., 2015).
In this paper, we use a numerical modelling approach to: (1) in-
vestigate the impact of initial bathymetry; wave energy and eustasy
curves on the depositional architecture and clinoform geometries of the
carbonate platforms of the late Jurassic to Lower Cretaceous of Oman
(Tithonian to Early Aptian); (2) to test if the prograding to aggrading
regional trends can be reproduced using eustasy alone or if changes in
regional subsidence rates are needed; and finally (3) to explore the
value of the model to reproducing the main depositional hetero-
geneities observed in the rock record. We adopt an experimental ap-
proach that allows testing conceptual geological models by producing
multiple stratigraphic scenarios. This study enhances facies predictions
for the Lower Cretaceous of Oman, and help understanding overall
controls on the deposition of carbonates.

2. Geological setting

Fig. 2. Plate reconstruction for the studied megasequence (149-92 Ma). The
The separation of the Arabian-Iranian terrains during the Triassic to
intra-cratonic rifting between Africa/Arabia and India/Australia led to the de-
Jurassic led to relaxation and rapid subsidence of the passive con-
velopment of the passive margin deposits in the Arabian Plate (Edited from
tinental margin of the Neo-Tethys Ocean that resulted in widespread Sharland et al., 2001).
flooding and carbonate deposition on the Arabian platform (Sharland
et al., 2001) (Fig. 2). This epeiric sea was bounded to the north and east
strata (Droste and Van Steenwinkel, 2004). The Early Cretaceous evo-
by a passive continental margin with the Neo-Tethys Ocean, and in the
lution of the eastern Arabian Plate is schematically shown in a regional
southwest by the Arabian shield that shed siliciclastic sediments to the
cross-section (Fig. 3). The stratigraphy is bounded by two regional
shelf (Davies et al., 2002). At the transition between the Arabian shield
unconformities linked to major tectonic events (Sharland et al., 2001).
and the epeiric sea, a mixed carbonate siliciclastic sedimentary system
The Late Jurassic basal unconformity is linked to the final opening of
was observed, termed ‘Migratory Carbonate Supressed Belt’ (Droste,
the Indian Ocean, while the upper megasequence unconformity
2010). During high sea-level, the clastic belts retreated westwards and
(∼92 Ma) is linked to the regional doming and erosion subsequent to
the carbonate dominated the shelf interior (Davies et al., 2002; Droste,
the formation of an intra-oceanic subduction along the Afro-Arabian
2010). During lowstands, the carbonate platforms were exposed and
Plate (Béchennec et al., 1995). This megasequence comprises 3 second-
carbonate deposition was restricted to the shelf margin (Droste, 2010).
order supersequences supersequences (Sharland et al., 2001): (i) the
The Cretaceous carbonate-platform in Oman is around 1200 m thick
older (in this paper referred to as ‘supersequence 1’ is Tithonian to
and 1000 m wide and it unconformably overlies Jurassic and older

46
M. Al-Salmi et al. Marine and Petroleum Geology 99 (2019) 45–60

Fig. 3. Cross-section through the Lower Cretaceous carbonate platform in Oman, showing the lower and middle second-order supersequences (modified after Droste
and Van Steenwinkel, 2004). The location of the cross section is shown in Fig. 1.

Valanginian in age and is capped by the Late Valanginian Un- (2004) (Fig. 3) was chosen as the reference case, as it is well accepted
conformity, (ii) the middle second-order supersequence is Valanginian by other authors and constrained by regional data that combines wells,
to Aptian in age ‘supersequence 2’ and is capped by a regional Late seismic and outcrop information. Published regional seismic lines in the
Aptian Unconformity that was caused by a relative sea-level fall and centre of Oman (Fig. 4) were used to infer the stacking patterns and
erosion in Northern Oman (Harris et al., 1984; Sharland et al., 2001); geometries observed in this model: the Lower Khamah Group is char-
(iii) the youngest supersequence ‘supersequence 3’ is Upper Aptian to acterised by a northward progradation and well-defined clinoform
Turonian in age and comprises the Nahr Umr Formation (Sharland geometries.
et al., 2001). The first two supersequences, each lasting around 25 Ma, Seismic mapping shows that the clinoform dips range from less than
can be linked to a large second-order sea-level cycles (Scott, 1990). 1° to more than 20° (Droste and Van Steenwinkel, 2004). Sections with
Supersequence 1 spans between 149 and 129 Ma, while supersequence high-angle clinoforms comprise thick sequences of platform-derived
2 spans from 129 to 113 Ma. The focus of this study is the Khamah grainstones and packstones interpreted to have formed during relative
Group that ranges from the late Berriasian to Aptian (Droste and Van sea-level rise, while the low-angle clinoforms comprises finer-grained
Steenwinkel, 2004). The Lower Khamah group shows an overall pro- sediments, formed during forced regression (Droste and Van
gradation of the Habshan-Salil-Rayda system. The platform edge pro- Steenwinkel, 2004). This clinoform interval is overlain by around
graded around 250 km to the north and northeast in approximately 700 m thick package of strong, nearly parallel, horizontal reflections
15 Ma (Droste and Van Steenwinkel, 2004). The progradational phase that contains the platform interior (Lekhwair to Natih formations)
was followed by a mainly aggradational trend, forming around 700 m (Droste and Van Steenwinkel, 2004). Well data from surrounding fields
thick sediments of the platform interior during Barremian to early are useful in obtaining a range of thicknesses for those facies and their
Turonian (Droste and Van Steenwinkel, 2004). The transgressive cycles lateral and vertical facies trends. The Lower Khamah Group section was
are dominated by gently dipping ramp depositional profile with com- penetrated by several wells in the subsurface of Oman (Haan et al.,
plex internal geometries as shown from sequence stratigraphic studies 1990). Two examples are given from two separate fields: the Yibal and
of Shu'aiba and Natih formations (van Steeninkel, 1992 and Van Al Huwaisah fields, where a subdivision was made between Habshan
Buchem et al., 2002). Table 1 summarizes the main lithology types (∼150–220 m thick) and Rayda/Salil (∼150–250 m thick) (Fig. 5)
within supersequence 1 and 2. (Haan et al., 1990). Well data point to a shallowing upward facies trend
within a clinoform belt from the pelagic Rayda Fm to slope facies of the
Salil Fm into the shallow marine bioclastic shoal of the Habshan Fm
3. Data set and methodology (Droste and Van Steenwinkel, 2004).
The Habshan Fm generally shows low gamma ray response com-
3.1. Data set pared to the underlying Salil and Rayda formations that show a serrated
signal (Haan et al., 1990). Many other wells target the upper Khamah
A conceptual numerical deterministic model is built to test several Group within the Shu'aiba (∼90–160 m) and Kharaib formations
depositional hypotheses. The model of Droste and Van Steenwinkel

Table 1
Lithology and type of rocks description.
Formation lithology Interpretation References

Rayda Formation Pelagic, argillaceous lime- mudstone and wackstone Pelagic carbonates (deposited below storm Haan et al. (1990)
wave base).
Salil Formation Argillaceous lime- mudstone and wackstone Slope deposits Pratt and Smewing (1990)
Habshan Formation Oolitic and bioclastic grainstone that marks a shallowing Platform edge deposits Haan et al. (1990)
upward trend
Lekhwair Formation Interbedded grainstone, wackstone and peloidal and oncoidal Lagoonal deposits in a broad shoal settings Alshahran and Narin (1986)
Kharaib Formation Argillaceous lime wackstones to packstones. It shows good Platform carbonates Van Buchem et al. (2002)
evidence of shoaling upward cycles.
Shu'aiba Formation Mudstone with deep water shales and redeposited grainstone Platform carbonates Droste and van Steenwinkel (2004), Masse
and packstones et al. (1998)

47
M. Al-Salmi et al. Marine and Petroleum Geology 99 (2019) 45–60

Fig. 4. Seismic section showing the interpretation of the Lower Cretaceous units. Note the clear progradation pattern within the platform edge (Habshan Formation).
Modified from Droste and Van Steenwinkel (2004). See Fig. 1 for seismic location.

(60–150 m) (e.g. wells A and B in Fig. 5) (Van Buchem et al., 2002). result of differential subsidence i.e. the block underlying Jebel Akhdar
Exposures of Lower Cretaceous strata in the Jebel Akhdar tectonic subsided less compared to the interior of Oman (Pratt and Smewing,
window, located around 100 km southwest of Muscat in the Oman 1993). Moreover, subsidence in the region could have been influenced
Mountains, allow reliable recognition of the geometry of the platform- by salt withdrawal and the shelf margin has undergone complex tec-
margin. Three representative lithological sections are shown in Fig. 6 tonic readjustments as the underlying Jurassic faults have experienced
and Table 2, located close to the depositional section of Droste and Van reactivation (Hillgärtner et al., 2003). This implies that there is gen-
Steenwinkel (2004): Wadi Bani Kharus, Wadi Sahtan, and Wadi Nakhr. erally a greater complexity and higher uncertainty in the northern part
It is worth noting that various local changes in thicknesses are visible of the Droste and Van Steenwinkel (2004) model compared to the
between different outcrops (Pratt and Smewing, 1993). This include southern part.
thinning of the Upper Kahmah unit in Jebels Akhdar and Madar as a The stratigraphic complexity needs to be considered in the context

Fig. 5. Well logs of the Lower Khamah Group from Al Huwaisah and Yibal Fields (Left). Modified after Haan et al. (1990). Log correlation through the prolific
reservoirs of Shu'aiba and Kharaib Formations (right). Modified after Van Buchem et al. (2002). See Fig. 1 for well locations.

48
M. Al-Salmi et al. Marine and Petroleum Geology 99 (2019) 45–60

Fig. 6. Outcrop sections of three Wadies, showing sedimentological observations and facies interpretations. See Fig. 1 for wadi locations. Modified after Pratt and
Smewing (1993).

Table 2 defined by a user designed subsidence or uplift maps and a eustatic sea-
Thicknesses of facies at three Wadi sections in meters (Pratt and Smewing, level curve; (2) the supply of sediment and/or carbonate in situ pro-
1993). duction; and finally (3) the transport of sediments using diffusion
Location Rayda Salil Habshan Lekhwair and Kharaib Shu'aiba equations (Granjeon, 2014; Seard et al., 2013). The depositional para-
meters (e.g. water depth, wave energy, basin slope) and resultant se-
Wadi Sahtan 83 485 75 240 55 diment/lithology proportions are quantified at each time step (Seard
Wadi Bani Kharus 74 295 70 165 125
et al., 2013).
Wadi Nakhr 37 275 70 165 125
The FSM comprises the northern region of Oman, and incorporates
the carbonate platform interior and its edge (Fig. 1, A). A sector of
of the proposed tectonostratigraphic evolution for this margin by Pratt about 300 km × 300 km has been delimited with a grid size of
and Smewing (1993) for the Early Cretaceous platform (Fig. 7). A 2 km × 2 km. The modelled Lower Cretaceous sequence spans from 149
uniform subsidence, typical of passive margins, dominated during the to 113 Ma and the time steps used for the simulations was set at 0.2 Ma.
Late Jurassic. This is followed by a major drowning event marking the This provides a good basin scale stratigraphic resolution for the model
Jurassic-Cretaceous boundary that is associated with extensive erosion whilst still allowing for a fast computational speed. To obtain a rea-
following variable uplift of up to 300 m. The uplift is believed to be a sonable representation of the various generic facies across the model,
result of the compression from oceanic crust and tilting of the eastern three different sediment types (grain sizes) were integrated in the
side of the platform (Pratt and Smewing, 1993). During the Early model: (i) coarse carbonate grains (0.5 mm), (ii) medium carbonate
Cretaceous, the area had subsided to below storm wave base, when the grains (0.1 mm), (iii) carbonate muds (0.004 mm). Facies were gener-
Rayda Formation was deposited (Pratt and Smewing, 1993). During the ated by combining the proportion of grain sizes with the various
Valanginian-Hauterivian, thermal subsidence dominated the region modelled depositional environments and physical parameters.
with no major tectonic influence. The second regional drowning event The main influencing parameters tested in this study are the initial
during the Albian was associated with a great detrital influx sourced topography, wave energy, subsidence rates and eustasy. Carbonate
from the uplifted Arabian Shield in the south and fringing exposures of production was kept constant throughout the simulated deposition, a
Palaeozoic sediments in the southwest (Droste, 2010). This has gener- reasonable assumption given that the type of carbonate producers and
ated the Nahr Umr Formation that forms a good regional seal for the climate regime did not markedly change throughout the Lower
prolific reservoirs of the Shu'aiba and Kharaib formations (Witt and Cretaceous in Oman. All assumptions are detailed and discussed in the
Gökdag, 1994). following paragraphs. For models simplicity, sediment compaction and
subsidence by sediment-load were not considered in this study.
To distribute sediments in the system, DionisosFlow solves a diffu-
3.2. Method
sion equation that links the depositional slope and transport coefficient
to local gradients. This controls sediments erosion and transport
The DionisosFlow forward simulation software (hereafter
throughout the basin (Hawie et al., 2015a). In this study, Long-Term
‘DionisosFlow’) is a process-based tool that simulates sediments pro-
Low-Energy transport parameters (LELT), were used to mimic con-
duction, transport and deposition over geological time scales in terri-
tinuous sediment transport through wave and gravity/slope at a large-
genous and carbonate settings. Four-Dimensional multi-lithology
scale in a low gradient carbonate system.
models simulating long-term (100 ky to 100 My) and large-scale
The equation for the linear slope-driven diffusion is defined as such:
(10–100 km) evolution of sedimentary basins are achieved (Granjeon,
1996; Granjeon and Joseph, 1999; Hawie et al., 2015a). Q = Kslope*S (1)
The simulation is initiated from an initial paleo-topography/
where Q is the average sediment flux [km2ky−1], Kslope is the diffusion
bathymetry, and proceeds following a set of time steps, within a spe-
coefficient and S is the slope [km2ky−1].
cified time interval. At every time step, three basic stratigraphic pro-
Wave transport depends on wave parameters, including wave
cesses are performed: (1) the addition or removal of accommodation,

49
M. Al-Salmi et al. Marine and Petroleum Geology 99 (2019) 45–60

Fig. 7. Schematic sections of the Arab Peninsula, presenting the main tectonostratigraphic events. Modified after Pratt and Smewing (1993).

energy and wave action depth. The equation for the wave-driven dif- time and water depth, and can be constrained to grow at a certain wave
fusion is: energy range. Carbonate production is further modulated in
DionisosFlow using a bathymetry coefficient ranging between 0 and 1
Q = kwave*Ewave*S (2)
that varies with water depths. Carbonate production at any given age is
where Q is sediment flux [km2y−1], kw is the diffusion coefficient for a result of the multiplication of the several parameters:
wave energy, Ewave is the wave energy, S is the local gradient of the
Pi = Pref (t) . Pb.i . Pw.i . Pe.i (3)
basin slope, [km2y−1] (Granjeon and Joseph, 1999).
Diffusive coefficients are directly related to the behaviour of sedi- Where Pi is the production rate (in m/My) for each sediment type, Pref
ments (Granjeon, 1996; Euzen et al., 2004; Hawie et al., 2015a,b), is the production of reference (in m/My); Pw.i is the wave influence (in
where finer grain sediments are more diffusive than coarse grain sedi- m/My); Pb.i is the bathymetric influence (in m) and Pe.i is the ecology
ments. For linear diffusion coefficient, a wide range of values have been influence.
used in the literature, ranging from < 0.01 km2 ka−1 (Gawthorpe et al.,
2003) to more than 1.69 107 km2 ka−1 (Burgess et al., 2006). For ex- 3.3. Assumptions
ample, the diffusion coefficient is estimated to be 560 km2ky−1 in the
Mississippi delta (Kenyon and Turcotte, 1985) and 0.007 km2ky−1 in The carbonate production and transport coefficients were assumed
pelagic strata on the flanks of the Galapagos spreading centre (Mitchell, to be constant throughout sediment deposition, and this allows us to
1996). test the influence of initial topography and choice of eustasy curves.
The sediment supply in the model mainly corresponds to in situ Initially, an average of 325 m/Ma for coarse and medium grains vs.
carbonate production. The combination of the various grain sizes (i.e. 250 m/Ma for muds were used as the reference production rates. The
carbonate mud, medium- and coarse grain-sized carbonates) is a proxy reason for this choice is that it led to a good fit between the expected
for the various carbonate producers within the depositional model. The facies types in the model and the modelling results. Fig. 8 shows sedi-
production of each type of carbonate grains is defined as a function of mentation rates of the carbonate factories by Schlager et al. (1998) that

50
M. Al-Salmi et al. Marine and Petroleum Geology 99 (2019) 45–60

In order to simulate supersequence 1, the basin floor geometry prior


to deposition must be defined. The initial topography here corresponds
to the unconformity between the older Jurassic deposits and the first
Cretaceous deposits (Droste and Van Steenwinkel, 2004). The bathy-
metry was assumed to range between zero meters, where the Habshan
slope edge has a southern limit at around 21° latitude, to a maximum
depth of 450 m in the basin where the pelagic Rayda Formation was
deposited. This is based on available regional data that suggests that the
depocenter of the Rayda Formation was deposited in few hundreds of
meters (Droste, 2010). The facies and microfossils found in the Salil and
Rayda formations support this interpreted water depth range as radi-
olarian are commonly found in deep water settings.
DionisosFlow is capable of accounting for sediment compaction is
an appropriate compaction law is supplied. However, in this study we
decided to not implement sediment compaction. This decision was
motivated by the fact that (1) finding an appropriate compaction law
for carbonates is complicated by early cementation, which reduce me-
chanical compaction and can occur at different times and with different
degree post-deposition for different lithology. This is thus a poorly
Fig. 8. Sedimentation rate of carbonate T-factory. Calculated based on thick- constrained effect in carbonates, (2) accounting for sediment compac-
ness and stratigraphic ages of ancient deposit versus the length of the time of tion greatly complicates calibration of the model, as sediment com-
observation (Schlager et al., 1998). paction could occur up to present day and thus the model should strictly
be run from basement up to modern sediment cover, and (3) also spe-
serves as a proxy for the tropical factory which is representative of the cific to carbonate is the importance of pressure-solution that can result
Jurassic and Cretaceous period along the Arabian Plate. The values used in decrease of sediment volume after mechanical compaction. Taken
in this study plot within an intermediate range in the graph (below the together, these complicating factors imply that accounting for sediment
upper limit) and are similar to measured values of similar rock types in compaction is likely to reduce, not improve, uncertainties in the model.
previous studies (Table 3). Given that the Arabian Peninsula was lo- As a consequence of not accounting for sediment compaction, volumes
cated in warm tropical water (i.e. 8° latitude, Murris, 1980), producers of sediments used in this study represent fully compacted sediment
are mainly autotrophic organisms such as rudists and algal limestones. loads.
Therefore, the sedimentation rates used in our model consider that
maximum production of carbonates occurs in the shallow subtidal zone
4. Results
and drops with increasing water depth. Two different production versus
depth profiles were used (Fig. 9): (i) for coarse and medium carbonate
4.1. Lower second-order supersequence (Tithonian -Valanginian)
grains, the highest production is in the upper 20 m of the water column
and rapidly decreases between 20 and 40 m, as light penetration drops,
4.1.1. Influence of initial bathymetry and wave energy on clinoformal
(ii) for mud grains, the production is assigned 120 m/Ma in the upper
geometries
10 m and drops to 15 m/Ma below 60 m to represent the influence of
Two end models for the initial bathymetry were tested: (1) a gentle
light penetration on pelagic sediments. Models can generally be more
topographic slope, with an average of 2 m/km (Fig. 10, Model A), and
complex than the reference case and to answer some important ques-
(2) steeper topographic slope, with an average of 10 m/km, that passes
tions related to depositional environment, variable laws of production
to flat bottomed basin ∼450 m (Fig. 10, Model B). Note that both
can be added. In this study, the carbonate production in all models were
models have the same total bathymetric change of 450 m and all the
constrained by wave energies that results in maximum production
other simulation parameters are similar (Table 6).
being focused vertically at shallow water depth, and laterally at various
The carbonate grains distribution and proportions is generally si-
locations depending on the energy setting of the platform (Table 4).
milar in Models A and B. Carbonate coarse-grains dominate the plat-
The diffusion coefficients used in the model runs are shown in
form edge (> 60%) that then passes to medium-fine grained slope fa-
Table 5: coarser, more resistant lithologies such as rudist reefs are re-
cies, and finally to basinal muds (> 60% mud grains). The inclination
presented by lower coefficients compared to finer unconsolidated car-
of the clinoforms is around 8 m/km at the edge and drops to 5-3 m/km
bonate muds. Another crucial factor in shallow waters (i.e. less than
at the slope. The slope shows a clear intercalation between fine,
40 m) influencing both carbonate production and transport is wave
medium and coarse carbonate grains. The thickness of the slope in
energy. In shallow-water carbonates, wave energy is high and wave
Model A increases gradually within 2 km from the southwest (∼50 m)
reworking is intense, leading to greater heterogeneity in the facies re-
to the northeast (350 m) that then passes to the basin (200 m). Coarse-
cord (Purkis et al., 2015). A constant wave base of around 20 m was
carbonate grains dominates the first ∼10 km of the prograding clino-
chosen for the studied period, which is a general average value for
forms. Whilst in Model B, there is relatively sudden break of slope and
shallow water environments (Coe et al., 2003). We chose a wave energy
thinner deep basin deposits (∼80 m). Moreover, results show that the
of 250 kW/m2, and a wave direction of 360° taking into consideration
steepness of the initial bathymetry has an impact on the geometry of the
the orientation and position of the basin.
carbonates. Where a gentle slope is used, individual units can be

Table 3
Examples of carbonate production rates for rocks of similar age and type.
Carbonate production (m/Ma) Age of the sample/s Location Reference

155 Early Cretaceous Shu'aiba Fm (Middle East) Sarg, 1988


Vary between 70 and 290 Early Cretaceous packstones and mudstones (Jura mountains in France) Strasser and Samankassou, 2003
Vary between 80 and 100 Cretaceous sediments Pelagic and deep water (North Sea, Italy, USA range) Enos, 1974

51
M. Al-Salmi et al. Marine and Petroleum Geology 99 (2019) 45–60

Fig. 9. (A) Carbonate production with depth, typical to that of autotrophic organisms that mainly depend on light penetration (B) wave energy versus depth.

Table 4 Reducing the diffusion coefficient values (i.e. without wave-driven


Wave energy constraints. diffusion) to model supersequence 1 results in gentle clinoformal geo-
Sediment type Min constraint (kW/m) Max constraint (kW/m) metry that dip by maximum of 5 m/km and comprises mainly of in-
tercalated medium and mud carbonate grains that can be correlated for
Carbo-coarse grains 30 230 (Max) great distances. (Fig. 11, Model C). This is because the sedimentation
Carbonate medium -grains 15 110
rate is lower within the system and less coarse grain carbonates are
Carbo mud Min (0) 50
transported basinward as the wave energy is reduced, as shown in
Table 7.
Table 5 Model E was simulated using both wave and gravity diffusion, but
Diffusion coefficient values. with lower values compared to models A and B, as shown in Table 8.
The carbonate coarse grains dominate the composition of the slope,
Sediment type Gravity-driven diffusion Wave-driven diffusion
forming a steep gradient of around 15 m/km that gradually drops to
(km2/kyr) (km2/kyr)
6 m/km, then to a flat bottomed basin (Model D, Fig. 11). There is re-
Carbo-coarse grains 0.3 0.3 latively less grain-size intercalations within the clinoforms as the car-
Carbo-medium grains 0.65 0.6 bonate coarse grains are transported short distances away from the
Carbonate mud 0.95 0.9
platform edge (i.e. flat-topped, relatively steep margined platform with
an abrupt break of slope) (see Table 9).
correlated over a greater distance. For example, some inclined packages An increase in the wave base by 5 m causes the platform edge to be
transported further northward by at least 20 km, forming relatively
will downlap the base Cretaceous unconformity over a horizontal dis-
tance of 300 km within 0.5 myr in Model A, versus only 250 km in steeper clinoform (up to 9 m/km), compared to Model E (Fig. 11). High
value of wave-energy (> 250 kW/m2) induces intense reworking of
Model B.
The initial bathymetric surface of Model B is considered the base sediments above the storm wave base and causes the sediments to
prograde and fill in the basin. The slope is dominated by coarse-car-
case in our study for the following reasons: (1) the edge of the platform
in supersequence 1 had prograded around 250 km during the 20 Ma of bonate grains, with negligible carbonate muds. Varying the azimuth of
the wave propagation angle does not seem to have an impact on the
supersequence 1 based on seismic data (Droste and Van Steenwinkel
(2004). This is similar to findings in Model B where the progradation development or morphology of the carbonate system and that wave
energy plays an important role in distributing the sediments in the re-
rate is around 10 km/Myr, and contradictory to Model A where the
gion.
platform edge extends further northwards. This has generated a lat-
erally extensive back-reef lithology (∼2.5 km) in Model A compared to
Model B. (2) the water depths in front of the platform were estimated to 4.1.2. Influence of eustatic curve
be of a few hundreds of meters, where the Rayda Formation was mainly The depositional model (Droste and Van Steenwinkel, 2004) shows
deposited (Droste and Van Steenwinkel, 2004). This deep water setting that the shelf-edge trajectory remained more or less constant at a re-
is reached sooner in Model B, which agrees with the southward extent gional scale in the first ∼ 20 Ma of deposition (Fig. 12), indicating: (i)
of the Rayda Formation available from outcrop data of Oman Moun- base level remained relatively constant, with tectonic subsidence and
tains (Fig. 6); (3) The inclination of the clinoforms of the Lower eustasy balancing each other out, and (ii) carbonate production and
Khamah Group is reported to range between 1° and 20°, as interpreted transportation had always exceeded subsidence rate. Therefore, a low
from seismic data and best captured in Model B; (4) Finally, the overall thermal subsidence rate was assigned for this interval of around 2 m/
thickness of the sequence (a combination of initial bathymetry, sub- Ma, a value calculated to compensate for the eustatic changes seen in
sidence and sedimentation) matches observation when using Model B. the Haq, (2014) eustatic curve.
(e.g. the thickness of the Rayda Formation increases laterally from the Three different eustatic curves have been tested, whilst keeping
south ∼70 m to the north ∼150 m as the bathymetry deepens). other environmental parameters constant. The eustatic curve from
Therefore, all other models presented below are based on the initial Miller et al. (2005) was used in the first scenario (Model F, Fig. 12). The
bathymetry of Model B. Miller et al. (2005) curve was established based on data from the

52
M. Al-Salmi et al. Marine and Petroleum Geology 99 (2019) 45–60

Fig. 10. Model A shows the geometry of carbonates when a gentle slope is used, while Model B represents the impact of using steep slope passing to flat bottomed
basin. Rayda Formation is predicted where the carbonate mud percentage exceeds 50% and it has been described in Oman Mountain (Table 2).

eastern margin of North America (New Jersey margin) and proposes 4.2. Second second-order supersequence (Valanginian – Early Aptian)
changes of 100 ± 50 m during the Cretaceous. Model G (Fig. 11) uses
the Haq and Bersma, (1998) long-term sea-level curve. Despite un- Several models were designed to investigate the relationship be-
certainties, especially during the Cretaceous-Cenozoic, the Haq and tween accommodation and stacking pattern in Oman. Four tests have
Bersma, (1998) curve has been used extensively in basin modelling, been conducted to examine subsidence role in carbonate geometries:
often used as a reference for contrasting global versus regional sea-level Model I. Simulations using eustatic curves (e.g. Haq, 2014; Miller
changes (Nystuen, 1998). Finally, the Haq (2014) short-term eustatic et al., 2005) with a similar subsidence rate as supersequence 1 (i.e. 2 m/
curve was used to simulate Model H (Fig. 12). This curve is an updated Ma) resulted in a continuous progradation of the clinoforms without
version of the Mesozoic eustatic curve of Haq et al. (1987), but the platform aggradation, as illustrated in Model I (Fig. 13). Results thus
recent work considers changes around the Arabian Peninsula and the confirm that eustasy alone cannot explain the aggradational stacking
global eustatic curve. pattern during supersequence 2, as the eustatic curves show no sig-
Simulations show that the shelf-edge trajectory will differ based on nificant increase throughout this period (129-113 Ma) (Model I,
the sea-level curve used: Model F shows a low angle, slowly ascending Fig. 13). Hence, a change in regional subsidence is needed to create
shelf-edge trajectory, while Model G shows that the shelf-edge remains sufficient accommodation for the ∼300 m thick back-reef sediments.
at the same level in the initial simulation time steps and rapidly rises Model J. This model was simulated using uniform subsidence rates
between 135 and 129 Ma. Model H shows that the first 20 km of the across the whole platform (Model J, Fig. 13). A subsidence rate of 21 m/
shelf edge is partially eroded as a result of sea-level drop and the se- Ma was used between 129 and 122 Ma and a rate of 50 m/Ma was used
diments were transported downslope. As the sea-level increases the between 122 and 113 m/Ma. This has generated a high progradation to
back-reef strata (< 40% carbonate coarse grains) onlap the shelf edge aggradation pattern in the first ∼20 km of the second second-order
as shown by the white arrows in Model H (Fig. 12). Moreover, carbo- supersequence. Between 122 and 113 Ma, a high aggradation to pro-
nate systems are highly sensitive to short-term eustatic amplitudes. In gradation pattern developed, where a barrier platform edge formed at
particular, the position of the shelf-edge (rudist complex), with low- the highest energy part of the depositional system (160–230 kW/m),
stand packages intercalated with highstand package representing comprising carbonate-coarse grains (100-80%(. The wave energy lat-
shorter-term sea-level fluctuation, being best captured using the Haq erally drops towards the back-reef (< 140 kW/m). Variable energy
(2014) short-term curve. This is because it includes higher-frequency belts developed vertically at the back-reef: the energy at the lower part
changes and it is based on more recent data, in part derived from the of the back-reef is less (< 40 kW/m) compared to the upper part
Arabian platform In contrast, less grain size intercalations are devel- (40–140 kW/m), where the aggradation rate is higher. Areas with high
oped within the clinoforms where long-term Haq sea-level or Miller wave energy in the back-reef comprises > 40% carbonate-coarse
et al. (2005) curves are used i.e. the platform edge is mainly composed grains, whilst areas with low wave energy are dominated by medium
of coarse carbonate grains and hence the models underestimate the and mud-carbonate grains (30–40%). These models show no significant
heterogeneity within the slope. lateral variations in the carbonate grains proportion.

53
M. Al-Salmi et al. Marine and Petroleum Geology 99 (2019) 45–60

Models K and L. A second set of scenarios were simulated to test

basin
basin
basin
basin
basin
basin
basin
basin
basin

Steep slope passing to flat bottomed basin


Steep slope passing to flat bottomed basin
the influence of differential subsidence in the system. In Model K, a
maximum subsidence rate of 27 m/Ma was used at the landward end of
bottomed
bottomed
bottomed
bottomed
bottomed
bottomed
bottomed
bottomed
bottomed
the model that drops to 16 m/My at the basinward end of the grid, as
shown in Fig. 14. Note that the subsidence maps are kept simple to
flat
flat
flat
flat
flat
flat
flat
flat
flat
represent the general subsidence trend at a regional scale and to satisfy
Steep slope passing to
Steep slope passing to
Steep slope passing to
Steep slope passing to
Steep slope passing to
Steep slope passing to
Steep slope passing to
Steep slope passing to
Steep slope passing to
the thickness of the system (section 3.1). The high subsidence rate that
Gentle slope 2 m/km

was induced in the platform could be due to loading on the platform


Initial topography

interior related to the uplift of the Haushi-Huqf High and the Arabian
Shield. The subsidence decreases away from the load, towards the
basin. This has the impact of deepening the shallow water portion of the
profile in the back-reef. Therefore, the main carbonate production is
forced to be closer to the platform margin and decreases towards the
back-reef. As a result, lateral variation in carbonate grains proportion
and thicknesses can be seen in Model K between 129 and 124 Ma that
Haq and Boersma, (1998)

are also likely controlled by the energy of the back-reef. For Model L,
two differential subsidence maps were used for ages between 129-
Miller et al. (2005)

124 Ma and 124-113 Ma, as shown in Fig. 14. Both models show very
Sea-level curve

similar back-reef composition, but with relatively more lateral hetero-


Haq (2014)
Haq (2014)
Haq (2014)
Haq (2014)
Haq (2014)

Haq (2014)
Haq (2014)
Haq (2014)

Haq (2014)
Haq (2014)

geneity in Model L between 129 and 124 Ma. Moreover, Model L shows
a high progradation:aggradation ratio as the system progrades north-
wards and the back-reef thickness increases from 300 m basinward to
500 m landwards.
Sediment Transport

5. Interpretation

5.1. Controlling parameters


5
5
7
8
5
5
5
5
5
5

Table 5
Table 5
Tabel
Table
Table
Table
Table
Table
Table
Table
Table
Table

The bathymetric gradient, or initial topography, plays a crucial role


when modelling carbonates because of its role in controlling the loci of
carbonate production, alongside wave energy distribution: the gentle
slope of Model A would imply a much larger area of intense carbonate
Max production rate (m/Ma)

production in shallow-waters, compared to Model B. This caused the


platform edge to prograde further in Model A. Outcrop and subsurface
observations favour Model B. An interesting result is that the gradient
of the initial topography is almost inherited by the gradient observed in
the modelled geometry in Model B. This is because of the reworking and
transport of sediments that prevents significant steepening of the plat-
form edge, and thus the maximum gradient of the slope does not exceed
325
325
325
325
325
325
325
325
325
325

325
325

the basement gradient value. Hence, initial topography plays a crucial


role in defining carbonate clinoform geometries.
Results indicate the importance of using a high transport system
combining wave-driven and gravity-driven coefficients, as shown in
Wave Base (m)

Model B. The high transport coefficients used in this study produces a


low-gradient ramp and slope system where most coarse sediments are
largely dispersed by normal marine processes and accumulate along the
20
20
20
20
25
20
20
20
20
20

20
20

slope. Such a transport-dominated system implies that carbonate fac-


tories produced sediments that are easily transported. This modelling
result is supported by empirical data from the Lower Cretaceous of
Subsidence rate (m/Ma)

Oman that show high energy, ooidal and reefal or bioclastic deposits of
Summary table of the input parameters in all models.

Salil and Habshan deposits of Jebel Akhdar and Madar outcrops (Pratt
and Smewing, 1993). Studies within the Quishn and Jurf Formations in
Differential subsidence (Map K)
Differential subsidence (Map L)

the Haushi-Huqf High (i.e. equivalent to the Kharaib are Shu'aiba


Formations) confirm that storm and wave reworking were the main
sedimentary processes influencing the tidal flats (Sena and John, 2013).
21
50
2
2
2
2
2
2
2
2
2

Also, results suggest that high production rates influence topographic


gradient by tending to steeper slopes, especially when transportation
rate is low, as shown in Model D. Hence, the interplay between in-situ
Period (m)

production and transport is crucial in determining the architecture of


149–129
149–129
149–129
149–129
149–129
149–129
149–129
149–129
149–113
129–122
149–114

sediments.

5.2. Relationship between accommodation and stacking pattern in Oman


H

Results of forward modelling confirm that the accommodation


D

G
A

Model K
C
B

E
F

Model L
J
I
Table 6

Model
Model
Model
Model
Model
Model
Model
Model
Model
Model

during the first second-order supersequence can be filled by the pro-


grading sediments using the specified production rates (Fig. 13) and

54
M. Al-Salmi et al. Marine and Petroleum Geology 99 (2019) 45–60

Fig. 11. Model C shows the impact of using gravity-driven diffusion alone, while Model D test the influence of using less diffusion coefficient values for both gravity
and wave-driven diffusion. Model E shows the result of increasing the wave base by 5 m.

Table 7 clinoforms develop during lowstand by forced regression. Thus, using


Diffusion coefficient values for Model C. higher-resolution eustasy curve could have great implications in iden-
Sediment type Gravity-driven diffusion Wave-driven diffusion
tifying the distribution of various depositional environments such as
(km2/kyr) (km2/kyr) lagoon, slope/open-shelf. Model A shows ∼40 m shallow water deposit,
developed during sea-level rise, being truncated towards the south
Carbo-coarse grains 0.3 0 (Model H, Fig. 12). These deposits comprise < 40% carbonate coarse
Carbo medium -grains 0.65 0
Carbonate mud 0.95 0
grains and high proportion of fine carbonates.
Droste and Van Steenwinkel (2004) suggest that the change from
progradational to overall aggradational pattern along the platform
Table 8 during supersequence 2, is related to a eustatic sea-level rise. This is
Diffusion coefficient values for Model C. supported by a regional rise in sea-level that was recorded in different
areas bordering the Tethys Ocean (e.g. Hillgärtner et al., 2003). How-
Sediment type Gravity-driven diffusion Wave-driven diffusion
(km2/kyr) (km2/kyr) ever, we argue that the evolution of the platform margin records an
increase in subsidence during the second super-sequence. The platform
Carbo-coarse grains 0.1 0.15 edge acts as a wave-resistant barrier that generates a shelter to the
Carbo medium -grains 0.4 0.3
back-reef and the wave energy propagation decreases towards the
Carbonate mud 0.6 0.6
platform interior as shown in Model L (Fig. 14). The system, in turn,
distributes the grain size considering both; water depth and energy
transport coefficients (Table 2), and also reproduces short-term shifts in regime, which has caused the development of muddy and medium-
the position and geometries of the reef tract resulting from changes in grained back-reef facies (Model L). This highlights the importance of
eustasy alone (Fig. 11). This result support previous studies by Connally considering the hydrodynamic setting in reconstructing the platform
and Scott (1985) and Scott (1990) suggesting that the Early Cretaceous geometry and that over simplifying depositional models would under-
period witnessed eustatic rise in sea level, and to results from Pratt and estimate the heterogeneity in shallow water systems.
Smewing (1993) who suggest that the regional subsidence rate was very Using uniform subsidence results in laterally oversimplified back-
low during the Berriasian-Hauterivian with no significant tectonic load reef lithology, likely related to the diffusion-based smoothing for the
(Fig. 7). This is evidenced by the lack of change in water depth during distribution of sediments used in the modelling. Differential subsidence
deposition of the Rayda Formation and age equivalent Mayhah and causes some lateral energy variations that induces a laterally variable
Guweyza limestone units (Pratt and Smewing, 1993). Eustasy greatly carbonate grain-size distribution. Differential subsidence affecting the
impacts the clinoform geometries, their composition, lateral continuity Arabian platform margin were discussed by many authors: for example,
and stacking patterns. Model B confirms that the geometries suggested Alsharhan and Nairn (1986) and Pratt and Smewing (1993) mention
by Droste and Van Steenwinkel (2004): high-angle clinoforms develop that tilting and erosion of the western part of the Arabian Platform
during aggradation phase of highstand system tract and low-angle could have induced differential subsidence and instability in the region

Table 9
Facies interpretation in the model mainly based on grain size cut-offs and bathymetry.
Facies type Carbo- medium Carbo-medium Carbo-mud Carbo- mud Carbo-coarse Carbo-coarse Bathymetry (min) Bathymetry (max)
grains (min) grains (max) (min) (max) grains (min) grains (max)

Grainy back- 0 1 0 0.5 0.5 1 <0m 2m


reef
Muddy back- 0 1 0.5 0 0 0.5 <0m 2m
reef
Reef Facies 0 1 0 1 0.45 1 2m 80 m
Slope Facies 0.2 0.7 0 1 0 1 100 m 531 m
Basinal Mud 0 1 0.3 1 0 0.5 300 m 455 m

55
M. Al-Salmi et al. Marine and Petroleum Geology 99 (2019) 45–60

Fig. 12. Three scenarios showing the impact of eustasy curve choice: Model F uses Miller et al. (2005), Model G uses Long term Haq and Boersma, (1998) and Model
H uses Short-term Haq (2014). All other parameters were kept constant.

Fig. 13. Model I shows the result of using a sub-


sidence of 2 mMa−1 between 149 and 113 Ma.
Uniform subsidence rate of 21 m/Ma (129-122 Ma)
and 50 m/Ma (122-113 Ma) are used in Model J and
three models are shown (from top to bottom): (i)
coarse-grains proportion, (ii) medium-grains pro-
portion and (iii) mud-grains proportion. The dashed
line separates Supersequence 1 (S1) from
Supersequence 2 (S2).

56
M. Al-Salmi et al. Marine and Petroleum Geology 99 (2019) 45–60

Fig. 14. Models K shows the result of using single differential subsidence map between 129 and 113 Ma. Model L uses two subsidence maps: 129-124 Ma (left) and
124-113 Ma (right). The dashed line separates Supersequence 1 (S1) from Supersequence 2 (S2).

due to periodic uplift of the Arabian Shield. However, further studies slope facies with some alternations between coarse-grained and fine-
are needed to constrain the timing, magnitude and spatial variations of grained carbonates as a response of relative sea-level fluctuations. The
the subsidence. Using differential subsidence generates differentiated back reef facies are characterised by a relatively lower wave energy
bathymetry that impact wave energy propagation and hence laterally compared to the platform edge and largely comprise of coarse to
variable carbonate grains distribution are generated. medium-grained carbonates. During aggradation phase in super-
sequence 1 mud-carbonate grains are developed. This unit likely cor-
5.3. General implication of the modelling results responds to the lower part of the lagoonal Lekhwair Formation in the
Droste and Van Steenwinkel (2004) model. It is also associated with the
As presented earlier, the modelling work helps understanding the slower sedimentation rate compared to the fast aggradation stacking
potential interactions between different parameters, such as sea-level pattern dominating supersequence 2. Generally, the model shows
fluctuations and subsidence, in controlling the stratal architecture and coarse-carbonate grains dominating the back-reef where the aggrada-
depositional patterns of carbonates. It is also considered a powerful tool tion rate is high (Fig. 15). The model also shows streaks of relatively
in generating some of the depositional heterogeneities observed in higher mud content that correspond to periods of low energy setting in
subsurface and outcrop data, such as lithological variations within the the system. These can be correlated for at least 150 km and could en-
back-reef and the slope deposits. To illustrate this, the facies were de- hance the heterogeneity of the back-reef reservoirs as it may form
fined based on sediment proportions and ecological parameters (e.g. barriers and baffles of vertical flow. The gentle slope in the carbonate
bathymetry, wave energy, slope … etc.) (Table 8 and Fig. 15). Model k ramp means that flooding surfaces are generally diachronous. Due to
(Fig. 13) was chosen for this purpose, as it provides satisfactory geo- the fluctuations in relative sea-level, extensively exposed surfaces can
metries and reasonable thicknesses: Back-reef facies (300–400 m), slope be detected from the bathymetry model (Fig. 16). Differential sub-
facies (400–500 m), platform edge facies (60–220 m), and Basinal Fa- sidence, like Model L, causes differential exposure times of these sur-
cies (50–180 m), compared to the published data (refer back to Figs. 6 faces and control their lateral continuity. It is expected that the het-
and 5). The model shows a regional transition from grainy to muddy erogeneity induced by the fluctuations in sea-level and the thicknesses

57
M. Al-Salmi et al. Marine and Petroleum Geology 99 (2019) 45–60

Fig. 15. A model showing facies interpretation that were distributed based on Table 8. Five main facies were assigned to describe the geometries and heterogeneities
within the carbonate system.

of these surfaces enhances reservoir heterogeneity within the back-reef Shu'aiba Formation is highly complex and comprises several deposi-
facies (e.g. karstification, fractures, dolomisation/cementation). Ex- tional environments including open-marine settings, shallow marine
posure surfaces have been described at the Haushi-Huqf High, central settings (rudist-rich), shoals and lagoon (semi-restricted setting) (Jeong
Oman (e.g. Sattler et al., 2005; Immenhauser et al., 2004). Differential et al., 2017). The best reservoir quality is found within the inner ramp
subsidence causes differential exposure times of the surfaces and con- and is dominated by Lithocodium/Bacinella to backshoal grainstones.
trol their lateral continuity as seen in Models K and L. At outcrops, they Reservoir quality decreases with increasing micritic matrix and clay
show evidence of meteoric diagenesis, subaerial exposure and boring, content, typically with a decrease in quality in a north-westwards di-
and could be traced laterally at least for 20–60 km and include laterally rection (Jeong et al., 2017). Several facies schemes have been proposed
extensive erosion surfaces, omission surfaces (hard and firm grounds), for the Shu'aiba Formation to understand this complexity. Nevertheless,
and flooding surfaces (Hillgärtner et al., 2003). The model allows development of reservoir framework remains a challenging step in
simple extraction of facies belts and surfaces, as shown in Fig. 15, B. constraining a 3-D geological model, as it requires mapping of in-
This shows that the platform margin is the most sensitive recorder of dividual clinoform surfaces that holds great complexity (Yose et al.,
relative sea-level changes. Enhanced progradation during Berriasian to 2006). Clinoform geometries have proven to have big impact on con-
Hauterivian times results from slow subsidence and continuous sedi- nectivity between injectors and producers in Shuabia reservoirs. Hence,
ment supply. This changes to higher aggradation rates in the Late seismic-scale flow units and barriers are often mapped to visualise in-
Hauterivian to Barremian as the subsidence rate increased. Moreover, jector-producer well-connectivity (Yose et al., 2006). Generating high-
the 3-D geometry of the platform edge allows a good visualisation of the resolution, detailed stratigraphic models could capture facies change
vertical and horizontal stacking of the coarse grained carbonates during down the profile of each single clinoform (mainly due to water depth
specified time intervals, which could, for example, enhance reservoir and wave energy change) that could significantly enhance the char-
productivity. acterisation of individual flow units. Moreover, there is a considerable
diagenetic overprint in Shu'aiba reservoirs. Diagenesis is mostly facies
selective and is related to sequence-stratigraphic surfaces (e.g. exposure
5.4. Potential future direction to extract more reservoir implications for the
surfaces) (Yose et al., 2006). Combining both depositional facies and
subsurface of Oman
sequence-stratigraphic framework in forward modelling could improve
reservoir characterisation and management (i.e. placing injectors and
One of the main producing reservoirs in Lower Cretaceous strata in
producers), as it predicts key exposure surfaces and their thicknesses.
the region is the Shu'aiba Formation. The depositional setting of the

Fig. 16. Bathymetry model highlighting exposed surfaces. All surfaces were exposed between 0.1 and 1 Ma.

58
M. Al-Salmi et al. Marine and Petroleum Geology 99 (2019) 45–60

Studies of Shu'aiba reservoir reveals a very ordered system, where 6. Conclusion


changes in stratal geometries and facies distribution are closely related
to long-term accommodation history (Yose et al., 2006). Other factors Carbonate platform geometries are a result of multiple controls,
such as relative sea-level changes, local tectonics and energy flux cause including sediment production, sea-level fluctuations and differential
significant carbonate platform asymmetry between the northern and subsidence. Forward modelling results of the Lower Cretaceous in
southern margins of the Shuabia platform (Yose et al., 2006). The Oman showed that the controlling factors of the carbonate architecture
northern margin developed a broad grain-dominated sand belt, fol- varied through time: (1) eustasy played a significant role in the car-
lowed by strong progradation system. On the other hand, the southern bonate development during the first second-order super-sequence; (2)
margin is characterised by high-energy rudist shoals that cover a broad the clinoform geometries and their lateral continuity are mainly con-
area of the margin (Yose et al., 2006). Producing reservoir-scale 4D trolled by the magnitude of the transport coefficients and the bathy-
models could provide valuable insights into carbonate platform asym- metry, (3) subsidence was the primary control for the observed ag-
metry, as it enables the integration of a wide range of depositional gradational stacking patterns during the Hauterivian-Aptian and (4)
controls and testing their relationships. differential subsidence and wave energy play a crucial role in facies
We propose that applying Dionisos 4D forward stratigraphic mod- distribution. Finally, testing the proposed model against published data
elling to a broader set of subsurface data in Oman could be used to: (1) shows that it successfully predicts the regional facies stacking patterns
complement stochatic modelling and (2) to enhance the predictability and the depositional geometries as suggested, for example, by Droste
and confidence level on reservoir models (Hawie et al., 2015b). This is and Van Steenwinkel (2004), and Pratt and Smewing (1993). Moreover,
because complex carbonate sedimentary geometries (i.e. rudist shoals, regional-scale observations and interpretations from the model can be
algal mounds, development of clinoforms, aggradation and prograda- extrapolated in a useful way to the subsurface (e.g. exposure surfaces,
tion of carbonate platforms) require a broad multi-disciplinary dataset grain-size distribution).
integration and understanding of various geological processes. For ex-
ample, a calibrated forward stratigraphic model was generated for the Acknowledgements
three Lower Cretaceous reservoirs located on onshore oil field in Abu
Dhabi (Hawie et al., 2015b). A 200 × 200 m grid size and a 50 kyrs This study was supported by data and previous fieldwork from the
time step were used to simulate these reservoirs, supported with sen- Qatar Carbonate and Carbon Storage Research Centre, funded by Qatar
sitivity analysis (Hawie et al., 2015b). These models consider in situ Petroleum, Shell, and the Qatar Science and Technology Park. We thank
production and transport by waves, gravity and water influx (Hawie the associate editor Istvan Csato, as well as Rainer Zühlke and an
et al., 2015b). A maximum of 35 wells were used to calibrate the model, anonymous reviewer for their comments that help improve this
which has resulted in the generation of 14 carbonate textures for the manuscript.
different reservoirs (Hawie et al., 2015b). For example, the model
predicts patchy rudist shoals developing on gentle inner ramp paleo- References
topographic highs. At the shelf edge, a more continuous belt is ex-
pected, as the energy is the strongest, whereby mudstone/wackstone Alsharhan, A.S., Nairn, A.E.M., 1986. A review of the Cretaceous formations in the
deposits dominate the basin (Hawie et al., 2015b). The Hawie et al., Arabian Peninsula and Gulf: Part I. Lower cretaceous (Thamama group) stratigraphy
and paleogeography. J. Petrol. Geol. 9 (4), 365–391.
2015b study forms a good analogue to Shu'aiba and Kharaib reservoirs Béchennec, F., Le Métour, J., Platel, J.P., Roger, J., 1995. Doming and downwarping of
in North Oman, where mapping clinoforms remains a challenging step, the Arabian platform in Oman in relation to EoAlpine tectonics. In: In: Al-Husseini,
yet very critical to understand producers-injectors connectivity in re- M.I. (Ed.), Middle East Petroleum Geosciences, vol. 1. Gulf PetroLink, Bahrain, pp.
167–178 GEO’94.
servoirs (Yose et al., 2006). Generating high-resolution, detailed stra- Borgomano, J., Masse, J.P., Al Maskiry, S., 2002. The Lower Aptian Shu’aiba carbonate
tigraphic models could capture facies change down the profile of each outcrops in Jebel Akhdar, northern Oman: impact on static modeling for Shu’aiba
single clinoform (mainly due to water depth and wave energy change) petroleum reservoirs. AAPG (Am. Assoc. Pet. Geol.) Bull. 86 (9), 1513–1529.
Burgess, P.M., Lammers, H., van Oosterhout, C., Granjeon, D., 2006. Multivariate se-
that could significantly enhance the characterisation of individual flow quence stratigraphy: Tackling complexity and uncertainty with stratigraphic forward
units, as shown by Hawie et al. (2015b) study. modeling, multiple scenarios, and conditional frequency maps. AAPG Bull. 90 (12),
Another forward stratigraphic modelling study of the Lower 1883–1901.
Cantrell, D.L., Griffiths, C.M., Hughes, G.W., 2015. New tools and approaches in carbo-
Cretaceous carbonates of Saudi Arabia was presented by Cantrell et al.
nate reservoir quality prediction: a case history from the Shu'aiba Formation, Saudi
(2015). The Cantrell et al. (2015) study, unlike our study and the work Arabia. Geol. Soc. Lond. Spec. Publ. 406 (1), 401–425.
of Hawie et al., 2015, is based on applying hydrolic-process equations Coe, Angela L., Bosence, Dam W.J., Church, Kevin D., Flint, Stephen S., Howell, John A.,
where both the hydrolic flow and sediment supply are explicitly cal- Wilson, R., Chris, L., 2003. The Sedimentary Record of Sea-level Change. Cambridge
University Press and the Open University, Cambridge, UK.
culated (using the package SedSim). The study area (127 × 187 km), Connally, T.C., Scott, R.W., 1985. Carbonate sediment-fill of an oceanic shelf, lower
modelled using a 1 × 1 km cartersian grid and a simulation timestep of Cretaceous, Arabian Peninsula. In: In: Crevello, T.C., Harris, P.M. (Eds.), Deep-water
20 kyr from 125 to 116 Ma, encompasses Shu'aiba Formation sediments Carbonates: Buildups, Turbidites, Debris Flows and Chalks; A Core Workshop. Society
of Economic Paleontologists and Mineralogists, Core Workshop, vol. 6. pp. 266–302.
from the extensive back-ramp carbonates in Saudi Arabia; the en- Davies, R.B., Casey, D.M., Horbury, A.D., Sharland, P.R., Simmons, M.D., 2002. Early to
vironment of deposition modelled in Cantrell et al. (2015) is thus more mid-cretaceous mixed carbonate-siliciclastic shelfal systems: examples, issues and
proximal to the shoreline than the shelf-edge environment of deposition models from the Arabian Plate. GeoArabia 7 (3), 541–598.
Droste, H., 2010. High-resolution seismic stratigraphy of the Shu'aiba and Natih forma-
modelled in our study. Four main environment of deposition were tions in the Sultanate of Oman: implications for Cretaceous epeiric carbonate plat-
modelled in the Cantrell et al. (2015) study, with maximum carbonate form systems. Geol. Soc. Lond. Spec. Publ. 329 (1), 145–162.
production rates of 200 m/Ma (compared to our production rates of Droste, H., Van Steenwinkel, M., 2004. Stratal geometries and patterns of platform car-
bonates: the cretaceous of Oman. In: In: Eberli, G., Massaferro, J.L., Sarg, J.F.R.
250–325 m/Ma for the shelf edge). A detailed comparison of the two
(Eds.), Seismic Imaging of Carbonate Reservoirs and Systems, vol. 81. American
models is beyond the scope of this contribution, in part because the Association of Petroleum Geologists, pp. 185–206.
environments of depositions modelled are different. We note however Enos, P., 1974. Surface sediment facies of the Florida-Bahamas Plateau. Geological
Society of America 5, 4.
that both the diffusive and the hydrodynamic-based models lead to the
Euzen, T., Joseph, P., Du Fornel, E., Lesur, S., Granjeon, D., Guillocheau, F., 2004. Three-
modelling of prograding clinoforms, and to realistic thickness maps (see dimensional stratigraphic modelling of the Grès d'Annot system, Eocene-Oligocene,
Cantrell et al., 2015). This reinforces the point that forward modelling SE France. Geol. Soc. Lond. Spec. Publ. 221 (1), 161–180.
of carbonate sequences can be a valuable addition to the geologist's Gawthorpe, R.L., Hardy, S., Ritchie, B., 2003. Numerical modelling of depositional se-
quences in half-graben rift basins. Sedimentology 50, 169–185.
toolkit.

59
M. Al-Salmi et al. Marine and Petroleum Geology 99 (2019) 45–60

Glennie, K.W., Boeuf, M.G.A., Hughes-Clarke, M.W., Moody-Stuart, M., Pilaar, W.F.H., Sugarman, P.J., Cramer, B.S., Christie-Blick, N., Pekar, S.F., 2005. The Phanerozoic
Reinhardt, B.M., 1974. Geology of the Oman Mountains (parts 1, 2 and 3). The record of global sea-level change. Science 310, 1293–1298.
Hague, Martinus Nijhoff. In: Verhandelingen Koninklijk Nederlands Geologie en Murris, R.J., 1980. Middle East: stratigraphic evolution and oil habitat. AAPG (Am. Assoc.
Mijnbouw Genootschap 31. Pet. Geol.) Bull. 64, 597–618.
Granjeon, D., 1996. Doctoral dissertation. Modélisation stratigraphique déterministe: Nystuen, J.P., 1998. History and development of sequence stratigraphy. Sequence stra-
Conception et applications d'un modèle diffusif 3D multilithologique, vol. 1 tigraphy—concepts and application. Norwegian Petr. Soc. Spec. Pub. 8, 31–116.
Université Rennes. Pratt, B.R., Smewing, J.D., 1990. Jurassic and Early Cretaceous platform margin config-
Granjeon, D., Joseph, P., 1999. Concepts and applications of a 3-D multiple lithology, uration and evolution, central Oman Mountains. Geol. Soc. Lond. Spec. Publ. 49 (1),
diffusive model. In: Stratigraphic Modelling, vol. 62. Society for Sedimententary 69–88.
Geology, SEPM, pp. 197–210. Pratt, B.R., Smewing, J.D., 1993. Early Cretaceous platform-margin configuration and
Granjeon, D., 2014. 3D forward modelling of the impact of sediment transport and base evolution in the central Oman Mountains. Arabian Peninsula: AAPG (Am. Assoc. Pet.
level cycles on continental margins and incised valleys. Int. Assoc. Sedimentol. Spec. Geol.) Bull. 77 (2), 225–244.
Publ. 46, 453–472. Purkis, S.J., Rowlands, G.P., Kerr, J.M., 2015. Unravelling the influence of water depth
Haan, E.A., Corbin, S.G., Clarke, M.H., Mabillard, J.E., 1990. The Lower Kahmah Group of and wave energy on the facies diversity of shelf carbonates. Sedimentology 62 (2),
Oman: the carbonate fill of a marginal shelf basin. Geol. Soc. Lond. Spec. Publ. 49 (1), 541–565.
109–125. Sarg, J.F., 1988. Carbonate sequence stratigraphy. In: Wilgus, C.K., Hastings, B.S.,
Harris, P.M., Frost, S.H., Seiglie, G.A., Schneidermann, N., 1984. Regional unconformities Kendall, C.G., Posamentier, H. w., Ross, C.A., Wagoner, J.C. (Eds.), Sea-level Change
and depositional cycles, cretaceous of the Arabian Peninsula. In: In: Schlee, J.S. (Ed.), – an Integrated Approach. vol. 42. Society of Economic Palaeontologists and
Interregional Unconformities and Hydrocarbon Accumulation, vol. 36. American Mineralogists, Special Publication, pp. 155–181.
Association of Petroleum Geologists, Memoir, pp. 67–80. Sattler, U., Immenhauser, A., Hillgärtner, H., Esteban, M., 2005. Characterization, lateral
Haq, B.U., 2014. Cretaceous eustasy revisited. Global Planet. Change 113, 44–58. variability and lateral extent of discontinuity surfaces on a carbonate platform
Haq, B.U., Hardenbol, J., Vail, P.R., 1987. Chronology of fluctuating sea levels since the (Barremian to Lower Aptian, Oman). Sedi-mentology 52, 339–361.
Triassic. Science 235, 1156e1167. Seard, C., Borgomano, J., Granjeon, D., Camoin, G., 2013. Impact of environmental
Haq, B.U., Boersma, A. (Eds.), 1998. Introduction to Marine Micropaleontology. Elsevier. parameters on coral reef development and drowning: forward modeling of the last
Hawie, N., Deschamps, R., Granjeon, D., Nader, F.H., Gorini, C., Muller, C., Montadert, L., deglacial reefs from Tahiti (French Polynesia;IODP Expedition #310).
Baudin, F., 2015 a a. Multi-scale constraints of sediment source to sink systems in Sedimentology. https://doi.org/10.1111/sed.12030.
frontier basins: a forward stratigraphic modeling case study of the Levant region. Sena, C.M., John, C.M., 2013. Impact of dynamic sedimentation on facies heterogeneities
Basin Res. https://doi.org/10.1111/bre.12156. in Lower Cretaceous peritidal deposits of central east Oman. Sedimentology 60 (5),
Hawie, N., Barrois, A., Marfisi, E., Murat, B., Hall, J., El-Wazir, Z., Al-Madani, N., Aillud, 1156–1183.
G., 2015 b b. November. Forward stratigraphic modelling, deterministic approach to Schlager, W., Marsal, D., Van Der Geest, P.A.G., Sprenger, A., 1998. Sedimentation rates,
improve carbonate heterogeneity prediction; lower cretaceous, Abu Dhabi. In: Abu observation span, and the problem of spurious correlation. Math. Geol. 30 (5),
Dhabi International Petroleum Exhibition and Conference. Society of Petroleum 547–556.
Engineers. Scott, R.W., 1990. Chronostratigraphy of the Cretaceous carbonate shelf, southeastern
Hillgärtner, H., van Buchem, F.S., Gaumet, F., Razin, P., Pittet, B., Grötsch, J., Droste, H., Arabia. In: In: Robertson, A.H.F., Searle, M.P., Ries, A.C. (Eds.), The Geology and
2003. The Barremian-Aptian evolution of the eastern Arabian carbonate platform Tectonics of the Oman Region: Journal of the Geological Society of London, vol. 49.
margin (northern Oman). J. Sediment. Res. 73 (5), 756–773. pp. 89–108 Special Publication.
Huang, X., Griffiths, C.M., Liu, J., 2015. Recent development in stratigraphic forward Strasser, A., Samankassou, E., 2003. Carbonate sedimentation rates today and in the past:
modelling and its application in petroleum exploration. Aust. J. Earth Sci. 62 (8), Holocene of Florida Bay, Bahamas, and Bermuda vs. Upper Jurassic and lower cre-
903–919. taceous of the Jura Mountains (Switzerland and France). Geol. Croat. 56 (1), 1–18.
Immenhauser, A., Hillgärtner, H., Sattler, U., Bertotti, G., Schoepfer, P., Homewood, P., Shafie, K.R., Madon, M., 2008. A review of stratigraphic simulation techniques and their
Vahrenkamp, V., Steuber, T., Masse, J.-P., Droste, H.J., Koppen, J. v., Kooij, B. v/d., applications in sequence stratigraphy and basin analysis. Bull. Geol. Soc. Malays. 54,
Bentum, E.C. v., Verwer, K., Hoogerduijn-Strating, E., Swinkels, W., Peters, J., 81–89.
Immenhauser- Potthast, I., Al Maskery, S.A.J., 2004. Barremian-Lower Aptian Qishn Sharland, P.R., Archer, R., Casey, D.M., Davies, R.B., Hall, S.H., Heward, A.P., Horbury,
Formation, Haushi-Huqf area, Oman: a new outcrop analogue for the Kharaib/ A.D., Simmons, M.D., 2001. Arabian Plate Sequence Stratigraphy: Geoarabia Special
Shu’aiba reservoirs. GeoArabia 9 (1), 153–194. Publication 2, vol. 371 Gulf Petrolink, Bahrain.
Jeong, J., Al-Ali, A.A., Jung, H., Abdelrahman, A., Dhafra, A., Shebl, H.T., Kang, J., Bonin, Van Buchem, F.S., Razin, P., Homewood, P.W., Oterdoom, W.H., Philip, J., 2002.
A., de Perriere, M.D., Foote, A., 2017. November. Controls on reservoir quality and Stratigraphic organization of carbonate ramps and organic-rich intrashelf basins:
reservoir architecture of early cretaceous carbonates in an Abu Dhabi onshore field Natih Formation (middle Cretaceous) of northern Oman. AAPG Bull. 86, 21–53.
Lekhwair, Kharaib and lower Shu’aiba formations. In: SPE Abu Dhabi International Van Buchem, F.S.P., Razin, P., et al., 1996. High resolution sequence stratigraphy of the
Petroleum Exhibition & Conference. Society of Petroleum Engineers. Natih formation (Cenomanian/Turonian) in northern Oman: distribution of source
Kenyon, P.M., Turcotte, D.L., 1985. Morphology of a delta prograding by bulk sediment rocks and reservoir facies. GeoArabia 1, 65–91.
transport. Geol. Soc. Am. Bull. 96 (11), 1457–1465. Van Steeninkel, M., 1992. Sequence stratigraphy of the Shuaiba Formation, Oman:
Litsey, L.R., MacBride Jr., W.L., Al-Hinai, K.M., Dismukes, N.B., 1986. Shu’aiba reservoir Tentative correlations: Shell International Internal Report. pp. 11.
geological study, Yibal Field, Oman. J. Petrol. Technol. 38 (6), 651–666 . Witt, W., Gökdag, H., 1994. Orbitolinid Biostratigraphy of the Shu’aiba Formation
Masse, J.P., Borgomano, J., Al Maskiry, S., 1997. Stratigraphy and tectono sedimentary (Aptian), Oman. Implications for Reservoir Development: Micropalaeontology and
evolution of a late Aptian–Albian carbonate margin: the northeastern Jebel Akhdar Hydrocarbon Exploration in the Middle East: London. Chapman and Hall, pp.
(Sultanate of Oman). Sediment. Geol. 113, 269–280. 221–234.
Masse, J.P., Borgomano, J., Al Maskiry, S., 1998. A platform-to-basin transition for lower Yose, L.A., Ruf, A.S., Strohmenger, C.J., Schuelke, J.S., Gombos, A., Al-Hosani, I., Al-
Aptian carbonates (Shu’aiba Formation) of the northeastern Jebel Akhdar (Sultanate Maskary, S., Bloch, G., Al-Mehairi, Y., Johnson, I.G., 2006. Three-dimensional
of Oman. Sediment. Geol. 119 (3), 297–309. characterization of a heterogeneous carbonate reservoir, lower cretaceous, Abu
Mitchell, N.C., 1996. Creep in pelagic sediments and potential for morphologic dating of Dhabi (United Arab Emirates). In: Harris, P.M., Weber, L.J. (Eds.), Giant Hydrocarbon
marine fault scarps. Geophys. Res. Lett. 23 (5), 483–486. Reservoirs of the World: from Rocks to Reservoir Characterization and Modeling.
Miller, K.G., Kominz, M.A., Browning, J.V., Wright, J.D., Mountain, G.S., Katz, M.E., AAPG Memoir 88/SEPM Special Publication, pp. 173–212.

60