Sedimentological characterization of the mid-Cretaceous Mishrif reservoir in southern Mesopotamian Basin, Iraq

Cenomanian–Early Turonian carbonates of the Mishrif Formation in the Mesopotamian Basin form one of the world’s most prolific hydrocarbon reservoirs in many oilfields in Iraq, such as West Qurnah, Halfaya, Majnoon and Rumaila (Figure 1). Deposition occurred in an extensive shallow-water carbonate platform associated with an intra-shelf basin (the Najaf Basin of Aqrawi et al., 2010) that is similar to other correlative Middle Eastern Cretaceous carbonate reservoirs (e.g. as figured in Murris, 1980) (Figure 2). During the Cenomanian–Early Turonian, the shelf margins were the ideal locations for prolific growth of rudist biostromes. This facies forms important hydrocarbon reservoirs in various other Cretaceous Middle East basins (e.g. Harris and Frost, 1984; Burchette and Britton, 1985; Videtich et al., 1988; Burchette, 1993; van Buchem et al., 1996; Aqrawi et al., 1998).

In Iraq, the Mishrif Formation consists of shallow open-marine, organic detrital limestones, with beds of algal, rudist, coral-reef and lagoonal facies. The major hydrocarbon accumulations are found in the rudist-bearing facies (Figure 2b) deposited along the shelf margins and extend from Hamrin and Badra in the north and central Iraq towards the south (Figure 3) (Aqrawi et al., 2010). In addition, some fields have been discovered in shoal facies of inner-shelf settings. Shallow open-marine facies have minor reservoir potential, but can still be producible in some fields such as the Amara and Halfaya.

The rudistid reservoirs are characterized by high porosities (exceeding 30%) and high permeability, which locally can exceed 1,000 mD, while the net-pay thickness may exceed 150 feet or 50 m (BP, 1970). Such unique reservoir characteristics, which also have wide geographic extent, result in the Mishrif ranking as the number one reservoir in Iraq. So far the total proven reserves in the Mishrif reservoir exceeds 30% of the total Iraqi national reserves (Al-Sakini, 1992).

The main objectives of this study are: (1) to characterize the Mishrif reservoir units, primarily by using a comprehensive microfacies/facies scheme, (2) to predict reservoir unit distribution within a sequence-stratigraphic framework; and (3) to develop a porosity-permeability predictive procedure for exploration and production purposes.

The Mishrif Formation was originally described by Owen and Nasr (1958) from the Zubair-3 Well in southern Iraq. The definition was further formalized by Dunnington in van Bellen et al. (1959-2005), whilst the type section was re-described in more detail in the work of Al-Naqib (1967). Models of the Mishrif facies were developed by Gaddo (1971) working on the Zubair, Tuba and Rumaila fields; whilst Al-Khersan (1975) presented the first descriptions of the formation and its facies from the fields in the border area with Iran to the northeast. Subsequent papers, such as Reulet (1982), Aqrawi et al. (1998), Sharland et al. (2001) and Sadooni (2005), concentrated on the stratigraphic and sequence-stratigraphic organization of the formation in various parts of its platform. They also considered, to some extent, diagenesis of the formation and its importance for reservoir characterization. Of importance to this study is the observation by Aqrawi et al. (2010, p. 201) that in the numerous previous studies of the Mishrif Formation, there is no single agreed-upon microfacies/facies scheme, and that previous publications often covered detailed local, rather than regional, aspects of the formation.

Based on benthic and planktonic foraminifera, rudists, and other fauna Dunnington (in van Bellen et al., 1959-2005) suggested a Mid-Cenomanian–Early Turonian age for the Mishrif Formation. The formation conformably overlies and passes westwards into the oligosteginal carbonates of the Rumaila Formation with a gradational boundary (Al-Naqib, 1967; Aqrawi et al., 1998) (Figure 2). In central parts of the intra-shelf Najaf Basin, the upper boundary of the Mishrif and Rumaila formations are conformable with the overlying Turonian Kifl Evaporite Formation (Al-Naqib, 1967). The upper boundary is unconformable with the overlying, transgressive, basinal Upper Turonian–Coniacian Khasib Formation in most other parts of Iraq (Dunnington, in van Bellen et al., 1959-2005; Darmoian, 1975) (Figure 2). The presence of conglomerates, early cements, bored limestones and brackish lithologies with fresh-water algae at the top of the succession indicates erosion and exposure of the Mishrif carbonate platform (e.g. Al-Khersan, 1973; Al-Siddiki, 1978; Sadooni, 2005). The Mishrif-Khasib break represents the upper boundary of the middle Cretaceous Wasia Group over most of the Arabian Plate, and it resulted from a combination of tectonic uplift and a global eustatic sea-level fall during the Mid-Turonian (Sharland et al., 2001).

Both tectonism and eustasy controlled deposition of the Mishrif Formation in Iraq. According to the Cretaceous sequence-stratigraphic organization of Iraq by Aqrawi et al. (2010), the Mishrif Formation belongs to the latest Albian–Early Turonian aged Supersequence IV (Figure 2a). This, in turn, is equivalent to the uppermost part of tectonostratigraphic Megasequence AP8 of Sharland et al. (2001), which was deposited on a passive margin setting. Geographically, the study area (i.e. southern Iraq) is located in the Mesopotamian structural zone, which was subdivided into the Zubair, Tigris, and Euphrates tectonic subzones (Buday and Jassim, 1987) (Figure 1). Stratigraphic evidence for this subdivision is the pronounced thickening of the Mishrif Formation along the Tigris subzone, which indicates higher subsidence rates there (Figure 1). Each of these subzones is, in turn, characterized by structural highs and lows with different trends that were formed by the deformation of the northeastern Tethyan margin of the Arabian Plate in Cenomanian–Early Turonian time (Jassim and Goff, 2006). Many structures within southern Iraq are the result of salt diapirism as indicated by the presence of negative gravity residuals below some supergiant oilfield structures such as Zubair, Rumaila and Nahr Umr (Jassim and Goff, 2006; Aqrawi et al., 2010). Some of these structures started growing during the Early Jurassic (e.g. Al-Sakini, 1992; Sadooni and Aqrawi, 2000).

This study utilized data from 25 wells distributed across southern Iraq (Figure 1). Petrographic analysis of more than 2,000 thin sections prepared from core and cutting samples was carried out using polarized and binocular microscopes. The analyzed samples were collected at 0.5–1.0 meter depth intervals. The definition of their grain types, depositional textures, and diagenetic features was based on a combination of classifications and terminology of Dunham (1962), and Embry and Klovan (1971), respectively. Information from a report by ELF (1970) aided in the foraminifera identifications. Electric log-suites of gamma-ray, sonic, density and neutron logs were used to correlate facies in non-cored intervals of the studied wells, as each facies association has some distinctive log signatures that can easily be tracked vertically and laterally between the wells. In addition, these signatures assisted us in the interpretation of depositional cycles and key sequence-stratigraphic surfaces. Two well correlation lines were selected to show the facies changes from outer-to inner-shelf settings of the Mishrif carbonate platform (Figure 1). The concepts of carbonate sequence stratigraphy of Sarg (1988), Handford and Loucks (1993) and Embry (1993, 2002) are used to construct the sequence-stratigraphic framework based on the correlation of depositional cycles and key sequence-stratigraphic surfaces between the studied wells.

The petrographic study of the Mishrif Formation in 25 wells across the study area has led to the identification of 21 microfacies using the standard microfacies approach of Flügel (1982) and the standard facies belts of Wilson (1975). Also, examination of cores was helpful in defining facies, especially for the coarse-grained rudistid facies (Figure 4). Defined microfacies are then grouped into what we term “facies associations” (although strictly speaking, data origin means that these are often “microfacies associations”) in terms of their common depositional environments (Table 1). Also, a typical vertical stacking of these facies associations is shown for the Dujaila-1 Well (Du-1) in Figure 5. Trends in facies association (Figures 2b and 5) represent a complete carbonate shelf succession covering various depositional environments from deep-marine to inner-shelf (Figure 6). The vertical stacking of the facies associations reflects relative changes in sea level, representing various depositional cycles. Two main regressive cycles were distinguished across the stratigraphic succession of the Mishrif (Figures 2b and 5) consisting of smaller accommodation cycles, which will be further discussed in later sections. The key characteristics of facies associations are summarized below.

The deep-marine facies association includes two microfacies types: (1) pelagic mudstone-wackestone (MFT-1); and (2) microbioclastic calcisiltite (MFT-2). In both microfacies types, lime mud and silt-size skeletal grains (derived from the breakup of rudists, mollusks, echinoids and brachiopods) are the dominant components, and they lack in-situ shallow-marine macrofaunas.

Pelagic organisms (mostly Hedbergella and “oligosteginids”) with sponge spicules are the major components of MFT-1 (Figure 7a). This microfacies is mud-supported, and characterized by high gamma-ray due to the presence of significant amounts of clay or organic matter. In a deep-marine succession, these facies can grade into shale beds, which are usually bounded by two MFT-1 units (Figure 5). These shale beds are mappable in several fields (e.g. Ratawi and Luhais). Their log responses show high gamma-ray, neutron, transit time, and they exhibit low density. They can be used as good correlation transgressive markers (i.e. MFSs) when using electric logs without core data. These fine-grained limestones and shaly beds can reach up to 10 meters in thickness in the Ratawi Field.

The microbioclastic calcisiltite facies (MFT-2) is dominated by microbioclastic debris with fine peloids, which are embedded in mud-supported packstone (Figure 7b). The major components of this silt-grade debris are derived from rudists, mollusks and echiniods. Some rare larger skeletal grains may also be present and include rudist fragments, brachiopods, and crinoids.

The main evidence for deep-marine depositional environment in both MFT-1 and MFT-2 is the predominance of pelagic foraminifera, microbioclastic debris, and the absence of in-situ shallow-marine fauna. In addition, the presence of shale units also suggests a low-energy depositional environment and could represent the maximum flooding surfaces above transgressive systems tracts. In addition, the abundance of microbioclastic debris in MFT-2 may indicate the influx of hemipelagic sediment derived from shallower depths during storms. The deep-marine facies association of the Mishrif Formation is equivalent to basin-and-lower slope facies belts of Wilson (1975), where the facies are deposited below normal wave base and under low-energy hydrodynamic conditions (Wilson, 1975; Flügel, 1982). Similar facies have been reported by Burchette and Britton (1985) for the Mishrif Formation in the United Arab Emirates (UAE), and by Razin et al. (2010) in the equivalent Sarvak Formation in Iran. The paleobathymetry of these facies is estimated by Razin et al. (2010) to be between 50Ð70 m.

This is the most abundant facies association in many wells. It consists of bioclastic-dominated microfacies, which are either low-energy, mud-supported microfacies (MFT-3 to MFT-7), or high-energy, grain-supported microfacies (MFT-8). These microfacies can be intercalated in some studied successions (Figure 4a). The skeletal grains of this facies association are diverse and abundant with different sizes (silt–pebble), and have various degrees of sorting (Figure 4a). In some cases, diagenetic processes such as dolomitization, neomorphism, micritization and dissolution affect microfacies components, which hamper classification.

In many studied successions, shallow open-marine facies can be dominated by certain types of skeletal grains such as echinoderms, molluscs, or dasycladaceans (green algae) such as in MFT-3 or MFT-4 (Figure 7c). Small and large benthonic and planktonic foraminifera may occur in these shallow open-marine facies; however the benthic foraminifera are more abundant. Their occurrence is quite evident in the foraminiferal-bioclastic wackestone-packstone facies, MFT-5 (Figure 7d), where they are associated with other bioclastic debris (Table 1). Coral-algal boundstone facies (MFT-6) is found in only a few of the studied wells, such as Nahr Umr-1, Ratawi-1 and Dujaila-1. These are interpreted as buildup microfacies and have framestone and bafflestone texture (Figure 7e), and are intercalated with other shallow open-marine facies, especially MFT-7 and MFT-8. Rudist bioclasts and fragments are a dominant component of this facies association and can form rudistid packstone (Figure 7f) and floatstone microfacies. Most of these bioclasts are abraded; their semi-rounded shapes suggest reworking (Figure 7f).

The characteristics of the above-described facies association (FA-II) indicate deposition in a shallow open-marine environment with a higher energy than facies association FA-I. It can be considered as a transitional environment between the deep-marine FA-I and rudist-biostrome FA-III environments. The abundance of rudist debris with the absence of intact rudists in cored intervals of MFT-7 and MFT-8 indicate that they may have been re-deposited on a slope in front of rudist biostromes at the shelf margin. In the same slope environment, coral-algal buildups are believed to have coexisted with rudist-related facies, which indicates more open-marine conditions than true rudist buildups (Gili et al., 1995; Skelton et al., 1997). These facies (i.e. MFT-4 and MFT-5) may also grade downslope into the muddy deep-marine facies association, where deep-marine influence is marked by the increased downhole presence of pelagic foraminifera.

The shallow open-marine facies association is comparable with the platform slope environment recognized in the Mishrif Formation of the UAE (Burchette and Britton, 1985). However, the lower-, middle-, and upper-slope units of the Mishrif carbonate platform are not widely recognizable in the studied fields and have not been used in this study. The depositional environment is estimated to be between fair-weather wave base and storm wave base.

The rudist biostrome facies association consists mainly of rudist rudstone microfacies (MFT-9) with radiolitid rudists as its major component (Figures 4b and 7g). The rudists occur as large unbroken shells, which can be in their life position in core. Their internal structure can be well preserved (Figures 4c and 7h), bored (Figure 7g), or completely neomorphosed. Rudists are also embedded in packstone or grainstone matrix that consist mostly of rudist fragments and bioclasts (Figure 7g), or peloids with rudist bioclasts grains (Figure 7h). Other faunal associations can include ostereids (Chondrodonta) (Figure 4b), echinoderms, and benthic foraminifera (e.g. praealveolinids, miliolids, and Chrysalidina). It is worth noting that these facies are interbedded with the rudistid packstone-grainstone microfacies (MFT-10) that was deposited on the shoal environment. In addition, the percentage of fragmental rudists and bioclasts are larger than the percentage of complete preserved shells.

The above characteristics of rudstone facies indicate a rudist biostrome environment that occupied the shelf margin (or edge) setting. This environment is equivalent to organic buildup facies belt (Facies Belt 5 of Wilson, 1975, and Flügel, 1982). The presence of sufficient amounts of lime mud with rudstone facies shows that rudists were not always developed in high-energy conditions. Moderate energy conditions allowed trapping of lime mud among rudist congregations. However, the occurrence of packstone-grainstone matrix between rudists indicates higher energy currents that were responsible for the removal of lime mud and finer fragments of rudist shells. This mechanism is also evident through the interbedding of rudist biostrome facies with shoal facies. According to the paleobathymetric profile of typical Cretaceous shelf margins (Scott, 1995), the paleobathymetry of this facies is estimated to be approximately between 2–10 meters.

This facies association is composed of pack- to grainstones (Table 1) that contain rudist bioclasts (Figure 8a), peloids, and coated grains. Other components include echinoderm, algae, molluscs, and intraclasts which are represented by MFT-10 to MFT-13. The grains have different sizes (1–20 mm) and sorting, and can be totally dissolved, or replaced by cement within the micritic envelopes of the grains. In addition, the packstone matrix is typically characterized by less mud than cement (Figure 8b).

The shoal facies association is equivalent to winnowed platform edge sands (Facies Belt 6) of Wilson (1975) and Flügel (1982). The characteristics of this facies association indicate a very shallow, high-energy and current-swept environment. This is typically suggested by the presence of coated bioclasts, micritized grains, cemented grains, which are set in a packstone-grainstone texture. In addition, this facies sometimes shows dissolution vugs (Figures 4d and 8a) and fresh-water cementation (Figure 8b) resulting from exposure.

This facies association usually overlies the rudist biostrome facies association. It could have originated from wash-over transport by currents and waves of rudist bioclasts and fragments that formed a debris belt surrounding the rudist biostromes.

The back-shoal facies association has been recognized in the uppermost part of the shoal facies successions of studied wells, but is separated from the shoal facies by a sharp contact. It includes microfacies types MFT-14, to MFT-16, represented by Chondrodonta floatstone (Figure 8c), foraminiferal-Chondrodonta wackestone-packstone (Figure 8d), and bioclastic-foraminiferal wackestone-packstone facies, respectively (Table 1). These facies are mud-supported and the major fauna include benthic foraminifera and Chondrodonta. These sediments in core are strongly bioturbated and compacted (Figures 4e and 8d). The abundance of rudist fragments and benthic foraminifera indicate the mixing of components from both shoal and lagoon environments in these facies. In addition, the presence of the ostreids is reported to be an indication of the vicinity of rudist biostromes (Scott, 1979).

These characteristics suggest a moderate energy environmental zone separating high-energy shoal and low-energy lagoonal environments. Therefore, they may be assigned to a back-shoal environment. The same environment is termed back-barrier/outer lagoon by Razin et al. (2010) for the Sarvak Formation of Iran. Also it was used for the environmental subdivision of the Mishrif Formation in the UAE by Burchette and Britton (1985).

This facies association is made up of three microfacies types, MFT-17 to MFT-19. The benthic foraminifera in most of these microfacies types (Figure 8e) are diverse, including miliolids, alveolinids, textularids, and Nezzazata. Other constituents are gastropods, ostracods, and scarce rudist debris. In many cases, miliolid foraminifera are the only component (such as in MFT-18). The microfacies in this facies association are characterized by strong micritization and burrowing (Figure 4f), the latter are filled with geopetal fabric of dolomitic micrite. In some cases, such as in MFT-19; peloidal packstone-grainstone (Figure 8f), the benthic foraminifera are rare and of small size and consist of miliolids and textularids. Other associated facies includes rudistid packstone-rudstone, which consists of micritized or recrystallized coarse debris of rudists with a few small miliolids and textularids.

The muddy texture and abundance of miliolids and other benthic foraminifera in MFT-17 and MFT-18 indicate low-energy conditions below wave base in the subtidal zone, which could be assigned to a lagoonal environment. Similar facies associations were reported from the Sarvak Formation of Iran (Razin et al., 2010), and Mishrif Formation of the UAE (Burchette and Britton, 1985). On the other hand, the peloidal packstone-grainstone facies (MFT-19) may be assigned to the intertidal zone due to the abundance of peloids, mud intraclasts, and micritized gastropod shell fragments (Flügel, 1982). The occurrence of rudistid packstone-rudstone facies interbedded within the lagoonal succession may indicate the presence of rudist biostromes within this environment. However, these are considered as inner-shelf and platform biostromes (Ross and Skelton, 1993), different from the dominant shelf-edge biostromes of the Mishrif Formation. They have been recognized in some fields such as Buzurgan, Faqa, and Abu Ghurab with thicknesses ranging between 1–4 meters.

This facies association is represented by MFT-20 (Charophyta wackestone, Figure 8g), and MFT-21 (unfossiliferous lime mudstone, Figure 8h). Other less common components in MFT-20 include small debris of ostracods and shell fragments. In MFT-21, the lime mud can be highly neomorphosed and exhibits a cloudy texture (Figure 8h). These facies correspond to a tidal-flat to brackish environment due to the presence of Charophyta with the high content lime mud (Flügel, 1982). A recent analogue of the lacustrine deltaic setting is the modern marshlands (Ahwar) of southern Iraq (e.g. Aqrawi, 2001). Generally, the association of Charophyta mudstone with MFT-21, and their interbedding with lagoonal succession reflect a tidal-flat environment, rather than a freshwater lake. This facies association is observed in the uppermost part of the Mishrif Formation succession in some oilfields of southern Iraq that border the Kifl facies belt (e.g. West Qurna, Rumaila, and Zubair).

A sequence-stratigraphic framework was built for the Mishrif Formation in southern Mesopotamia based on the correlation of sequence boundaries, maximum flooding surfaces, and the genetic facies tracts that are stacked between the key surfaces. This is combined with the construction of transgressive-regressive cycles through recognition of vertical stacking of facies, which reflects the interplay between relative sea-level changes and sediment input. Sequence identification was also based on wire-line log analysis employing the methodology of Vahrenkamp et al. (1993) by rescaling density (FDC) and sonic (BHC) wire-line logs with narrower ranges. Also, the gamma-ray log is used as an additional marker for shaliness, which increases in protected lagoonal facies compared to the open-marine setting (Vahrenkamp et al., 1993). Sherwani (1998) also used this methodology to differentiate protected lagoonal facies (PL) from the open-marine facies (OM) when studying the Mishrif as a part of a larger megasequence including the underlying Rumaila and Ahmadi formations (Figure 9).

In this study, three large-scale third-order sequences referred to as Mishrif I, II, III are interpreted within the Mishrif Formation (Figures 10 and 11) employing the sequence-stratigraphic definition of Goldhammer et al. (1991). These sequences are composed of combined transgressive-regressive sub-cycles that can merge laterally into larger regressive cycles, particularly towards the proximal eastern areas of the Mishrif carbonate platform.

Mishrif Sequence I represents the lower part of the Mishrif Formation, which is dominated by shallow open-marine and deep-marine facies, with few thin shoal facies units. Peloidal and bioclastic grainstones cap the shallow open-marine wackestones and packstones, which are underlain by pelagic mudstones and wackestones. They form the first shallowing-upward cycle in the Mishrif succession developed during a sea-level highstand. More shallow depositional conditions occur in Majnoon and Buzurgan fields where thick lagoonal successions were deposited (Figures 10 and 11). Keeping in mind the gradational contact between the Mishrif and Rumaila formations, and the deeper environmental conditions of the Rumaila and Ahmadi formations, Sequence I represents a regressive part of a sequence, which includes the Rumaila and Ahmadi formations (Figure 9), known as Sequence 2 of Sherwani (1998) or the early highstand above MFS K120 of Sharland et al. (2001).

The upper boundary of this sequence is picked at the base of the deep-marine deposits that overlie shallow open-marine sediments. In proximal areas, such as in the Majnoon Field, this boundary is placed between shoal facies and shallow open-marine units (Figure 11). The same boundary is characterized by sharp facies contact between lagoonal succession and overlying shallow open-marine facies in Buzurgan Field (Figure 10). Extensive investigation for this boundary from core and cutting samples show a hardground surface that is characterized by marine phreatic cement founded in shoal facies, but no signs of extensive subaerial exposure, which may suggest a drowning unconformity according to the classification of Schlager (2005). However, the depth of drowning was insufficient to completely prevent continued carbonate deposition.

The faunas of shallow open-marine units of Sequence I include echinoids, green algae, orbitolinids (Neoiraqia sp., Orbitolina sp., Orbitolinidae indet.), serpulids, with rare rudist debris. These units are also interbedded with deep-marine facies indicating distal locations such as Abu Amood Field (e.g. AAm-1 Well in Figure 10). The stratigraphic position of the top of Sequence I in the lower part of the Mishrif Formation and with respect to the underlying Rumaila and Ahmadi formations, in addition to the presence of orbitolinid foraminifera, suggests a Mid-Cenomanian age.

Sequence II starts with a transgressive surface (TS), which is characterized by deep-marine facies, rich in pelagic foraminifera. It also shows high gamma-ray log response indicating to high clay and/or organic content. During relative sea-level rise, up to 25 meters of transgressive deposits accumulated in the Luhais Field (basinward) where greater accommodation space was created; whereas in the Majnoon and Buzurgan fields (shelfward) they were thin and replaced by shallow open-marine facies (Figure 11). The transgressive cycle of Sequence II ends with a maximum flooding surface, which is characterized by a gamma-ray peak that can be followed in most studied wells. This key surface is probably the Mid-Cenomanian MFS K130 as defined by Aqrawi et al. (2010), which is recognizable throughout the Mishrif depositional system in Iraq. This correlation is based on the occurrence of orbitolinid foraminifera in the lower regressive cycle of the Mishrif Formation (below the Mishrif disconformity surface), which indicates a Mid-Cenomanian age according to previous works in Iraq (e.g. Bernaus and Masse, 2006).

The MFS of Sequence II is overlain by a thick regressive cycle. Within this regressive cycle there is a facies change from deep-marine facies at the base, to a shallow open-marine succession at the top (Figure 10). There is also a gradual upward increase in the content of small benthic foraminifera, echinoderms, green algae, and rudist debris, and on logs to a lower gamma-ray response and higher porosity logs.

In most studied wells, the regressive cycles of Sequence II are capped by rudist biostrome facies, whereas in the fields of the eastern part of the studied area, adjacent to the Iran border and in the West Qurna Field, thicker rudist biostrome and shoal units were deposited (Figures 10 and 11). Also, thick lagoonal units were deposited in the Majnoon Field, implying a paleogeographic change across the Najaf Basin with a westward regional progradational trend of the shelf margin within the study area (Figures 10 and 11).

The top boundary of Sequence II is characterized by a disconformity surface, which divides the Mishrif succession into two large-scale regressive accommodation cycles as assumed by Reulet (1982) and Aqrawi et al. (1998). This intra-formational mid-Mishrif disconformity was identified in all studied wells through the occurrence of mud-rich deep-marine or shallow open-marine facies that overlie shelf margin and inner-shelf deposits. In proximal oilfields, the disconformity surface is overlain by retrograding back-shoal or lagoonal successions of Sequence III (Figures 10 and 11). The sharp contact is associated with a relative sea-level fall with evidence of exposure consisting of abundant vuggy pores and geopetal infill. The infill includes carbonate silts, which can be dolomitized, and meteoric and late diagenetic cement. The effect of this dissolution extends to various depths into the upper part of Sequence II; for example, in the Halfaya Field, it may reach 16 meters. This secondary porosity is partially or completely destroyed by later cementation during the late deep burial in some fields, such as in the Rifaiy Field (Figure 3).

The mid-Mishrif disconformity is time-equivalent to similar surfaces recognized in some other oilfields across the Middle East, such as in the Fateh Field of Dubai (Videtich et al., 1988), and in the Fahud Field (Natih Formation) of Oman (Harris and Frost, 1984). Other possible correlative surfaces can be the top of Sequence II of the Sarvak Formation in southwest Iran reported by Razin et al. (2010), and the top of Natih E Member in Oman recognized by Philip et al. (1995). As a result, the Mid-Cenomanian global sea-level fall event of Haq et al. (1987) could be the major cause for the exposure of Sequence II and other equivalent units in the Arabian Plate. Therefore, a Mid-Cenomanian age is suggested for Sequence II. The coexistence of orbitolinids and Praealveolina cretacea in this sequence also suggests a Mid-Cenomanian age as recorded by Bernaus and Masse (2006).

In both correlation sections (Figures 10 and 11), Sequence II shows remarkable thickness variation indicating different accommodation rates created along the depositional profile of eastern and western sides of the Mishrif Basin in southern Mesopotamia. The greater thickness of lagoonal succession in Majnoon Field indicates higher accommodation space in the eastern part of the study area where carbonate production caught up with the rising sea level until the filling of accommodation space with lagoonal deposits (Figure 11).

The deposition of Sequence III began with a transgressive interval that is represented by deep-marine and shallow open-marine deposits overlying the mid-Mishrif disconformity. However, back-shoal and lagoonal deposition continued in some proximal areas such as Buzurgan, Majnoon and West Qurna fields (Figures 10 and 11). The effect of transgression in these fields is marked by the northeasterly retrogradation of both back-shoal and lagoonal facies (Figures 10 and 11). The thickness of transgressive cycles increases towards Abu Amood and Luhais fields (Figures 10 and 11). They are mainly composed of deep-marine facies, and gradually pass upward into maximum flooding surface of Sequence III. This MFS is probably equivalent to the Late Cenomanian MFS K135 as recently defined in southern Iraq by Aqrawi et al. (2010), and may be related to the global transgression that led to outer shelf and ocean basinal environments with anoxic conditions (Schlanger and Jenkyns, 1976; Gale et al., 1993), which is synonymous to the globally recorded Late Cenomanian oceanic anoxic event (OAE2) (i.e. Gale et al., 2008; Kuhnt et al., 2009; El-Sabbagh et al., 2011). However, in order to prove this event more regional information is required.

Preservation of Sequence III varies around the Najaf Basin because the top of the Mishrif Formation (i.e. top of Sequence III) is frequently truncated by a sharp contact with the overlying basinal shales of the Khasib Formation (Figure 9). The Khasib-Mishrif unconformity is the most significant sequence boundary that can be clearly identified on the log signatures. It is regional in extent, and has been reported from several locations across the Arabian Plate (Sharland et al., 2001). Exposure resulted from the Mid-Turonian tectonic uplift caused by ophiolite obduction on the northeastern Arabian Plate margin, and a global eustatic fall in sea level (Sharland et al., 2001). Evidence of this event was identified along the upper boundary of Sequence III, which is characterized by extensive dissolution and the presence of conglomerates or karstic breccias in some oilfields such as the Amara Field (Amara-1 Well, Am-1). The effect of exposure and meteoric diagenesis on the top of Sequence III is more pronounced than the mid-Mishrif disconformity. This could be due to a longer period of exposure that lasted several million years during the Turonian (Scott, 1990). Additional evidence for such a subaerial exposure is also recorded in the time-equivalent Sarvak Formation in southwest Iran, which includes brecciation, paleosol development, and bauxite deposits indicating a warm and humid climate (Hajikazemi et al., 2010).

The regressive deposition in Sequence III was also controlled by the regional tectonic activity that resulted from the ophiolite obduction along the margin of the Arabian Plate (Sharland et al., 2001). Therefore, shallowing of the Mishrif platform continued with the deposition of thick lagoonal and shoal deposits that extend southwest into the Najaf Basin areas, as far as the Noor and Abu Amood fields (wells Noor-1 [No-1] and Abu Amood-1 [AAm-1] in Figure 10), indicating a higher rate of progradation compared with the underlying Sequence II. In addition, the wide occurrence of supratidal facies at the top of Mishrif Formation in many oilfields, such as the West Qurna (Well WQ-1 in Figure 11), Zubair and Rumaila, indicates the existence of the shallowest depositional conditions during the end of Sequence III depositional time.

After the exposure and final karstification of the Mishrif Formation in the later Early Turonian, the topography of its upper boundary was covered by the shales deposited during the Khasib Formation transgression.

The stratigraphic position of Sequence III between the mid-Mishrif disconformity and the overlying Khasib Formation (Late Turonian–Coniacian) (Darmoian, 1975) suggests a Late Cenomanian-Early Turonian age for this sequence. Possible position of the Turonian succession occurs in the deep-marine or shallow open-marine units that make a sharp facies contact with the underlying shoal or back-shoal units (Figures 2b and 5). Similar facies relationships in Luhais and Ratawi fields have the same interpretation (Figures 10 and 11). The deposition of thick deep-marine successions in Nasiriyah and Abu Amood fields can be related to the Early Turonian flooding event (Figure 10). The absence of MFS K140 in these successions is related to the tectonic activity that masked the effect of the Early Turonian transgression. Evidence of this assumption is observed through the occurrence of a forced regressive surface that caps the shallow open-marine units as in Nasiriyah, Abu Amood, and Luhais fields (Figures 10 and 11).

A variety of diagenetic features are present within the Mishrif, including recrystallization, cementation, compaction, dissolution and dolomitization; all of which have an impact upon reservoir quality.

Recrystallization in the Mishrif reservoir comprises neomorphism of fine micrite grains and transformation of aragonite to calcite. The effect of recrystallization is clearly observed in shallow open-marine facies where the reservoir quality was reduced. Although the microporosity is usually higher in these facies, it becomes very low when the neomorphism of micritic grains occurs and starts blocking the micropores. Recrystallization has also affected the skeletal grains in different facies. However, grains usually preserve their intrafossil pores within their shell-structures or porosity is present as molds. In rudistid packstone and rudstone facies both coarse rudist debris and micrite are highly recrystallized (Figures 12a and 12b). Extensive recrystallization of micritic facies is commonly linked to prolonged residence of a stagnant fresh-water phreatic lens in the sediment (e.g. Longman, 1980).

Cementation in the Mishrif Formation could be either of early diagenetic origin within a fresh-water phreatic environment or of late origin during deep burial. Both types have affected the quality of the Mishrif reservoir in different facies. In Sequence III at the Mishrif upper boundary where extensive exposure took place, shoal and shallow open-marine reservoirs show fine equant calcite cement linings within various pore types (Figure 12c). The pores are filled with oil indicating that cementation pre-dates the oil migration. This type of cementation may occur in a fresh-water vadose environment (Moore, 1989) and can prevent pore reduction from later physical compaction.

Syntaxial rim cement is partially developed in grain-supported shallow open-marine and shoal facies. It occurs in small amounts and only partially occludes the interparticle pores (Figure 12c). This cement may have formed in a fresh-water phreatic diagenetic environment (Longman, 1980), or in a later deep-burial setting (Tucker and Wright, 1990). However, the preferential occurrence of such cement within shallow-water facies, just beneath a subaerial exposure surface, may favor the first origin.

Drusy mosaic cement is mainly found in rudist biostrome and shallow open-marine facies. It post-dates the dissolution of skeletal grains and may have formed in a meteoric environment (e.g. Scholle and Ulmer-Scholle, 2003). This cement, which is believed to be derived from dissolution of unstable components in the overlying sediment, can partly or completely fill the pores (Figure 12d). The same process can also affect the lagoonal facies beneath sequence boundaries where drusy mosaic cement fills the vugs and burrows within geopetal fabric below the upper sequence boundaries of sequences II and III (Figure 12e). Therefore, the poor reservoir-quality lagoonal facies of the Mishrif Formation become even lower in quality, particularly when located below the sequence boundaries.

The effect of blocky calcite cement on porosity and permeability is similar for many facies types. It fills moldic and vuggy pores, and also fractures. This cement type also shows distinct twinning (Figure 12f) that largely resulted from burial loading or tectonic deformation (Scholle and Ulmer-Scholle, 2003).

It is well known that both mechanical and chemical compaction affect the reservoir quality and are more pronounced in the muddy facies than the grainy facies (Moore, 2001). Mechanical compaction reduced the microporosity in mud-supported facies, and interparticle porosity in grain-supported facies of the Mishrif Formation, due to intense packing of the grains (Figure 12g). Stylolites are numerous in both outer- and inner-shelf facies with abundant micrite (Figure 12h). In shallowing-upward cycles, porosity and permeability decrease gradually downward in the succession where the micrite-rich facies are associated with numerous stylolites. The micrite-rich facies are mostly unaffected by leaching, especially those deposited during early highstand. In addition, the stylolite development post-dates leaching.

Dissolution is the most widespread diagenetic process that enhanced the reservoir quality in all types of the Mishrif facies through increasing the effective porosity. The rudist-bearing facies are particularly affected by this process due to early leaching of the aragonitic components of rudist shells, which mainly occur in the inner-shell layer. Such facies are particularly susceptible to leaching (Kauffman and Johnson, 1988) forming vuggy pores that reach more than 1.0 cm in size. Vuggy pores indicate that most extensive dissolution took place in shoal and biostrome facies (Figure 13a). In these facies, porosity may reach more than 30%, and permeability can exceed 1,000 mD, as dissolution increased the interconnectivity of interparticle pores (Figure 13b).

Dissolution of skeletal grains produced moldic pores in much of the mud-supported facies in both shallow open-marine and lagoonal facies. Green algal (Dasycladaceans) molds are the common pore type recognized in shallow open-marine facies, which have a poor reservoir quality due to the disconnected nature of the molds and vugs (Figure 13c).

Other effects of dissolution are also evident in compacted facies, such as rudistid and bioclastic packstones. In these facies, dissolution crosscuts compacted rudist fragments and other bioclasts forming both interconnected and separated vugs (Figure 13d). These vuggy pores suggest a phase of late burial dissolution (Moore, 2001). It is believed that subsurface dissolution would have resulted from hydrocarbon maturation, or shale dewatering might have provided aggressive fluids for carbonate dissolution (Moore, 1989, 2001). Possible pathways for such fluids can be faults, fractures, bedding planes or porous planes (Mazzullo, 2004).

Fractures can be identified at all scales. They are mostly found in hard, compacted, and stylolitic mud-rich facies such as mudstones and wackestones (Figure 13e). Oriented cores show there is usually a conjugate set, which usually runs both parallel and perpendicular to the axis of the anticlinal structures of most studied oilfields. The maximum width of fractures is 3 mm, and they can be either open or filled with calcite cement or oil residues. The precise role of fracturing and its effect on the reservoir quality in rudistid facies has not been studied and is unclear.

Dolomitization is uncommon in the Mishrif Formation, and has a limited effect on the reservoir quality. Two types of dolomite have been identified; dolomite associated within the stylolite zones, and dolomite rhombs scattered within the micrite of shallow open-marine facies. The first type is very common in mudstones and wackestones of different depositional environments, but has no impact on porosity. It is believed to have a burial-related late-diagenetic origin (Tucker and Wright, 1990).

The second type occurs within the mud-rich shallow open-marine facies as a few small clear rhombs scattered within the lime mud of host facies. In these facies, dolomitization may occasionally replace 50–90% of the original facies, and dolomite rhombs become tightly intergrown leaving some remnants of the host facies such as lime mud, skeletal grains, and pores (Figure 13f). Petrographic examination of such a dolomitic facies shows that all dolomite rhombs are of fine to medium crystal sizes and are characterized by cloudy centers (Figure 13f), which may indicate to their late diagenetic origin (e.g. Sibley, 1982).

In the lagoonal facies, late dolomitization develops a heterogeneous, massive, and patchy fabric. It follows the burrowing patterns that extensively cut the lagoonal facies (Figure 13g). Burrows commonly provide conduits for dolomitizing fluids in other areas (i.e. Scholle and Ulmer-Scholle, 2003). Although dolomitization in the Mishrif Formation can locally be high, measured porosity is low in these dolomitic intervals due to the tight intergrowth fabric of the dolomite crystals (or ‘overdolomitization’) (Figures 13g and 13h).

As mentioned above the main diagenetic processes that affected the Mishrif Formation were recrystallization, compaction, cementation, dissolution, fracturing and dolomitization. These processes took place through different stages of deposition and burial (Figure 14), and have significant impact on reservoir quality on the current Mishrif pay zones. The late diagenetic history of the Mishrif Formation resulted in the further modifications (usually destruction) of reservoir quality. Most of the potential porosity and permeability associated with reservoir facies was developed by dissolution during the early meteoric stage. The less common effect of burrowing can enhance porosity and permeability in the muddy facies by creating irregular but connected vugs. Analysis of porosity and permeability data for different depositional facies indicates that the higher porosities are generally associated with higher-energy facies, which include shoal and rudist biostrome (Figure 15). This may indicate the prevalence of the original primary interparticle pores as the main factor controlling the reservoir quality in the Mishrif reservoir of southern Mesopotamia.

In most studied fields, the shoal facies represents the best reservoir, and are characterized by high interparticle porosity and permeability that was further enhanced by dissolution (Figure 15). High-energy currents and wave action at shoals removed most of the lime mud in the rudist shoals leaving the grains to dominate. Generally, this rendered the reservoir quality of the shoal facies much better than the muddier rudist biostrome facies. Exceptions occur in the cemented shoal units that were deposited within the inner-shelf setting (Figure 15), which are characterized by lower porosity and permeability. Therefore, the amount of micrite seems to be an important factor influencing the reservoir quality of rudistid facies (i.e. the more abundant the micrite, the less quality the reservoir facies). Nevertheless, the presence of lime mud in the Cretaceous shallow-marine facies can increase their storage potential through the occurrence of microporosity, which is developed during relative sea-level highstand (Volery et al., 2009). Such reservoirs can be found in the lime mud-rich shallow open-marine wackestones and packstones associated with regressive cycles of the Mishrif Formation. In Sequence II, these reservoirs exhibit blocky porosity log motifs, and cannot be differentiated from the overlying shoal or rudist biostrome facies without core or petrographic examination (Figure 16).

Another depositional texture factor that could have an impact on reservoir quality is the size of rudist debris, which is usually more than 20 mm in both shoal and rudist biostrome facies. Such coarse debris, if well-sorted, may have contained an initial high interparticle porosity.

The porosity versus permeability plots of nearly 500 analyzed core plugs display a scatter relationship (Figure 15). This is mainly due to the diagenetic overprints on the depositional texture, fabric, and pores. Despite diagenetic overprinting, porosity and permeability decrease from high-energy grainy shoal facies to mud-rich deep-marine facies (Figure 17). In the depositional cycles, this change is reflected through the increase of log-based porosity from mud-rich transgressive and early regressive (highstand) facies to grainy late regressive facies (Figure 16). Development of reservoir facies of the Mishrif Formation is related to paragenesis, which is controlled by relative sea level. Three major phases of reservoir facies development are identified and related to changes in relative sea level and their accompanying diagenetic effects. This results in a predictable organization of reservoir units within the regressive cycles showing an upward increase of pore interconnectivity. Sequence II in Noor-1 Well [No-1] is chosen as an example to explain stages within each cycle (Figure 16).

The first phase is represented by relative sea-level rise (transgression) and deposition of low-energy and mud-rich mudstones and wackestones where porosity systems are dominated by micropores (Figure 18a) and limited development of separate vugs derived from burrowing (Figure 18b).

The second phase represents processes of the early highstand or the initial fall in relative sea level. It is characterized by wackestones and mud-dominated packstones with preservation of microporosity and occurrence of moldic pores (Figure 18c). The facies in this stage have poor reservoir quality due to the only rare occurrence of meteoric dissolution. In addition, neomorphism, compaction, and burial cementation are the common diagenetic processes in this and the former phase.

Continuous falling of sea level during the third stage led to flushing of meteoric waters, resulting in extensive dissolution of aragonitic green algae and rudist bioclasts (Figure 18d) that had been deposited in shallow open-marine environments. This phase represents the late regressive (or late sea-level highstand), which is dominated by packstones with development of extensive moldic porosity, but only limited development of vugs (Figure 18e). The effect of meteoric diagenesis in this phase is more pronounced than the second phase due to the decrease in accommodation space, which enhanced meteoric water incursion. The end of this phase records subaerial exposure, karstification, meteoric dissolution, and cementation along a type-2 sequence boundary. The meteoric dissolution of aragonitic and calcitic parts of rudist shells are expressed in the formation of interconnected vugs in rudistid packstones and grainstones facies (Figure 18f), whereas some rudist parts that escaped from dissolution were transformed into neomorphic spar by meteoric solutions.

The above-mentioned phases of development of pore types show that potential reservoir facies are concentrated in rudistid facies along the top of regressive cycles that can be capped by back-shoal, lagoonal facies, or transgressive successions (Figure 19). Therefore, regressive successions within the Mishrif Formation depositional sequences can be specifically targeted as the best potential reservoirs.

Petrographical study of the Mishrif Formation within southern Mesopotamia (Iraq) has identified 21 microfacies types (MFT-1 to MFT-21) through studying thin sections and core samples from 25 wells in 21 oilfields. The wide range of the microfacies represents a variety of depositional environments from deep-marine to shallow open-marine, rudist biostrome, shoal, lagoon, and tidal flats. The microfacies are grouped within seven microfacies associations (MFA-I to MFA-VII) based on their common depositional setting.

Three third-order depositional sequences are recognized through the vertical stacking of the depositional facies into transgressive-regressive cycles. The cycle boundaries can be correlated within and between oilfields in the southern Iraq, using some major geological events (i.e. sequence-stratigraphic surfaces) and unconformities (i.e. sequence boundaries) during the Cenomanian–Early Turonian time. Most are of regional extent and reported throughout the Arabian Plate.

Sequence-stratigraphic analysis shows that regressive cycles consist of prograding rudist biostromes and shoals that partially infilled the eastern margin of the intra-shelf Najaf Basin. The thickest Upper Cenomanian rudistid facies are located along the shallower parts of these structures as in Zubair and Rumaila oilfields (Gaddo, 1971; Al-Khersan, 1975; Sadooni and Aqrawi, 2000; Sadooni, 2005) suggesting that rudists congregated and thrived on growing salt-cored structures, surrounded by their derived coarse grain-supported shallow water facies, whereas muddier sediments were deposited in deeper and lower-energy environments (Aqrawi et al., 1998).

A combination of diagenetic processes and depositional environments control reservoir quality. Due to the relative sea-level falls, rudists located on paleohighs were most affected by meteoric dissolution that produced potential reservoir facies. Therefore, the reconstruction of detailed paleogeography during the deposition of various system tracts, particularly highstands, is critical for the successful exploration of the Mishrif reservoir in Iraq (Aqrawi and Horbury, 2008). Dissolution and fracturing enhance the reservoir characteristics, whereas neomorphism, cementation, and compaction have destructive effects on the Mishrif reservoir quality.

The best reservoir units occur in rudist-rich facies including both shoal and rudist biostrome, where aragonitic components have been dissolved under shallow vadose and meteoric diagenetic environments.

Sequence-stratigraphic layering of the Mishrif reservoir facies shows that regressive systems tracts host the best reservoir units along exposure and karstified upper boundaries of sequences II and III, which are capped and locally sealed by transgressive or lagoonal units.

The first author would like to acknowledge Statoil ASA for full sponsorship of his PhD study at the University of Bergen, Norway. The authors thank the anonymous reviewer for the suggestions that have improved the manuscript, GeoArabia’s Assistant Editor Kathy Breining for proof-reading it and Assistant Editor Heather Paul-Pattison for designing the paper for press.

Thamer A. Mahdi is a Carbonate Sedimentologist, and currently a PhD student at the Department of Earth Science, University of Bergen (UiB), Norway. He holds a BSc in Geology and MSc in Petroleum Geology from the University of Baghdad, Iraq. His present activity and research include the sequence stratigraphic interpretation and reservoir geology of the carbonate formations in southern Iraq.

[email protected]

Adnan A.M. Aqrawi is a Leading Consultant in Research, Development and Innovation (RDI), Innovation-Agents and Networks at Statoil, based in Stavanger, Norway. He obtained his BSc in Geology and MSc in Petroleum Geology from University of Baghdad, Iraq, and his PhD in Sedimentology and DIC in Sedimentary Geology from Imperial College-London, UK. Adnan has around 30 years of international experience as a Petroleum Geoscientist from the Middle East, North Africa, southeast Asia, Caspian Sea, Red Sea, North Sea and Gulf of Mexico while working for several research centres and oil companies in the Middle East, Malaysia, USA and Norway. Adnan joined Statoil in 2001 and has been involved mostly in international activities of Exploration and Business Development in the MENA Region and Gulf of Mexico (USA). He is an active member of AAPG, EAGE, SPE and NPF. He has a long list of publications in various international journals and conferences, and co-authored the book ÒThe Petroleum Geology of IraqÓ in 2010 with J. Goff, A. Horbury and F. Sadooni, which was sponsored by Statoil and BP, and published by the Scientific Press, UK.

[email protected]

Andrew D. Horbury is a Carbonate Geologist with Cambridge Carbonates, which he co-founded in 1992 after spending six years working the Middle East region with BP Exploration. He gained a BSc from Bristol University (UK) and a PhD in Carbonate Sedimentology from Manchester University (UK). Andrew is a co-author of the books ÒArabian Plate Sequence StratigraphyÓ and ÒPetroleum Geology of IraqÓ. His present interests include the sequence-stratigraphic evolution, sedimentology and reservoir geology of northern Arabia and the eastern Mediterranean, Mesozoic play systems and reservoirs in Mexico.

[email protected]

Govand H. Sherwani is a Petroleum Geologist and currently serves as Director General of Scholarships and Cultural Relations, Ministry of Higher Education, Kurdistan Regional Government (KRG), Iraq since 2009. He was awarded a BSc in Geology (1980), an MSc in Geology (1983), and a PhD in Petroleum Geology (1998) from Baghdad University, Iraq. Govand has held lecturing and many managerial posts at Salahaddin University, Erbil, including as the Head of Basic Science Department and the Dean of College of Science. He also held the post of General Coordinator at the Ministry of Higher Education at KRG, Iraq. Govand has acted as Consultant to several UN agencies and foreign oil companies based in Iraq. He has published more than 30 scientific articles and reports in local and international periodicals. His main areas of interest include sequence stratigraphy, basin analysis, and strategies of petroleum exploration.

[email protected]