ABSTRACT

To better constrain the spatial and stratigraphic distribution of the depositional facies, a synthesis of outcrop and subsurface data for the depositional system of the Upper Dalan Member and Kangan Formation in the Zagros to the offshore Fars area was carried out. The areas that were studied in detail are the Kuh-e Surmeh and Kuh-e Dena sections of the Zagros Mountains, Iran, and their equivalent in the offshore Fars subsurface. The observations and interpretations based on these sections were then integrated with the regional subsurface descriptions, interpretations and models, and related to the Upper Khuff system across the region.

The synthesis of the core descriptions and the Zagros outcrop facies data, together with integration of published data resulted in the definition and characterisation of 16 principal facies associations that were used to interpret the depositional environment. Qualitative comparisons of Upper Khuff sections and subsurface cores across the Zagros area, offshore Fars and Middle East Gulf region, showed that this classification of depositional facies is applicable at a larger regional scale and useful in rapid regional comparisons and correlations of the Upper Khuff depositional systems. The large range in documented facies types reflects the great variety in depositional systems and sub-systems that were present across the Khuff platform. The range also shows the temporal evolution of the Khuff environments and palaeoecological conditions from the Permian to the Triassic. The general importance of microbial facies is highlighted and a variety of microbial facies are defined. These microbial events provide reservoir and regional scale isochronous marker horizons that are correlatable over large distances. These microbial facies are associated with periods of poor oxygenation and restriction, but nevertheless can occupy a range of environments from intertidal to mid-to outer-ramp settings.

Several significant stratigraphic surfaces were picked and correlated based on the detailed core descriptions, the bio- and ecostratigraphic analysis, wireline logs, stratigraphic stacking patterns and the regional understanding of other Upper Dalan-Kangan/Upper Khuff sections in the region. The correlations in cored wells for the Upper Dalan cycles are supported by a well-constrained biostratigraphic framework. Four large third-order stacking cycles (Cycle IV to Cycle I) were defined on the basis of cycles bounded by surfaces representing baselevel and accommodation potential minima. The correlations and stratigraphic analysis suggest that the major stratigraphic trends and large-scale stratigraphic architecture are relatively isopachous (“layer-cake”) at the production scales, a function of the almost flat platform geometry. At a larger scale, significant changes in thickness occur: either thickening towards palaeodepocentres or thinning with onlap towards palaeohighs. At this large-scale, progradation of the oolite shoals occurred during the late highstands in the large accommodation areas. However, on the topographic palaeohighs and platform tops, the main stratigraphic locations of the oolite shoal are in the trangressive and maximum accommodation zones of the cycles. Integrating the facies and stratigraphic interpretations, conceptual depositional models have been constructed for the main stratigraphic intervals. From these interpretations and models it is evident that there were significant changes in platform type/geometry, facies organisation and climate from Cycle VI through to Cycle I.

At a large scale the Late Permian depositional setting of the Upper Khuff was organised into a platform profile that gently deepened from the south with a platform-top interior zone, a platform-top edge zone, an intrashelf low, and then rose again in the north with palaeohighs around Kuh-e Surmeh and Kuh-e Dena (structurally-controlled basement highs). There was however a major change in the platform profile in the Early Triassic which had a monoclinal ramp platform geometry which opened to the north to deeper-marine conditions with the absence of effective palaeohigh barriers. These two large-scale palaeogeographic profiles controlled the overall distribution of facies belts across the platform. This change in platform profile was coincident with other events within the lowest part of the Kangan Formation (Triassic Khuff Formation of the Arabian Plate) at the Permian-Triassic Boundary, including: (1) major facies changes on the platform tops with the appearance of thrombolites and associated microbial grainstones; (2) major facies changes in the northern shelf edge areas where there is a change from shallow-water high-energy grainy facies to deeper-water mid-ramp muddy facies; (3) change in pattern of relative stratigraphic thickness; and (4) appearance of high gamma-ray shales in the eastern Zagros subsurface area. These events are all consistent with a major flooding across the Permian-Triassic Boundary causing: (1) drowning of palaeohighs; (2) encroachment of anoxic waters into the intrashelf lows; (3) termination of bioaccumulations at the shelf edges; (4) flooding the platform tops with more grainy facies, and developing microbial facies across the shelf; and (5) the quasi-synchronous end-Permian mass extinction.

Based on the stratigraphic distributions of the biostratigraphically significant fauna and flora, age determinations are interpreted for the main stratigraphic intervals between the Lower Dalan to top Dalan (Lower Khuff to Permian Upper Khuff). Palaeoecologically, five biofacies types have been defined based on the faunal and algal content, the foraminiferal diversity, their sedimentological context and palaeoenvironmental interpretation. This generalised classification is applied to the depositional models developed from the sedimentological analysis and has enabled a validation of the depositional schemes by identifying palaeoenvironmental trends which are not always clear from the sedimentological analysis alone. The analysis of the biofacies distribution has allowed the subdivision of the Upper Dalan Member (Permian Upper Khuff) into six different ‘palaeoecological systems’ that correspond to characteristic faunal assemblages and biofacies sets. The main characteristics of the six palaeoecological systems, and their lateral variability, have been documented. The limits of the defined intervals correspond to important sequence stratigraphic events and markers at various stratigraphic scales. This relationship allowed the integration of ecostratigraphic events to the previously defined sequence stratigraphical framework based on the sedimentological and stratigraphic analysis, and hence confirms and refines the stratigraphic correlations.

A synthesis of stratigraphic, depositional and diagenetic facies, lithological, isotopic, spectral gamma-ray wireline logs and palaeoecological data suggests that there is no major stratigraphic gap between the KS3 and the KS2 stratigraphic intervals, and hence between the Permian and Triassic periods. In the numerous subsurface sections, and outcrop investigations in the Zagros, no evidence for a major unconformity/disconformity or stratigraphic surface is associated with the Permian-Triassic Boundary; furthermore the extinction of Permian fauna occurs within a grainstone body. The faunistic analysis shows that the Permian Fauna Extinction (PFE) event generally occurs within a strongly calcite-cemented and microbially mediated ooid grainstone rich in intraclasts in the lower part of the KS2 sequence. Above the PFE event is a thin Permian azoic interval, followed by the Triassic faunal recovery and associated with the Early Triassic thrombolitic microbial event. In the Zagros area the PFE occurs within pyrite-bearing muds under poorly oxygenated conditions. The outcrop data also show a similar pattern with a thin azoic interval occurring between the last Permian taxa and the first Triassic taxa. In the Zagros outcrops there is a general muddying (deepening-upwards) from the Upper Permian to the Lower Triassic. The analysis suggests there is a low (third) order transgression between upper KS3 stratigraphic interval (Upper Permian) and the KS2 stratigraphic interval (Lower Triassic), and that the ‘Permian-Triassic oceanic event’ is located in the late third-order TST.

INTRODUCTION

Importance of Upper Dalan Member and Kangan Formation and Study Objectives

The hydrocarbon reservoirs of the Upper Dalan Member and Kangan Formation (hereafter “Upper Dalan-Kangan reservoirs” and equivalent to the “Upper Khuff reservoirs” in the Arabian Plate, see Nomenclatural Note below and Figure 1) contain some of the most important gas reserves in the Middle East region, as well as the world. This is exemplified by the North Dome gas field (North field, Qatar and South Pars field, Iran), which is the largest offshore gas field in the world and is estimated to have over 1,100 trillion cubic feet of gas initially-in-place (TCF GIIP) (Upstream Journal, 2004). In particular, the Upper Dalan-Kangan reservoirs of onshore and offshore Iran contain some of the largest gas fields in the Middle East region (Kashfi, 1992, 2000); and yet these sequences, in both outcrop and subsurface, are less well-documented than their equivalents in Abu Dhabi and Saudi Arabia (El-Bishlawy, 1985; Alsharhan and Kendall, 1986; Al-Jallal, 1987, 1994, 1995; Alsharhan, 1993; Alsharhan and Nairn, 1994a, 1997; Al-Aswad, 1997). The landmark study of Szabo and Keradpir (1978) remains the main reference work for the general sedimentology and stratigraphy in the Zagros area, from which subsequent studies base their interpretation (Al-Jallal, 1987, 1994; Alsharhan and Nairn, 1997; Sharland et al., 2001). The current paper is focused on the sedimentological, sequence stratigraphic and biostratigraphic-palaeoecological study of the Permian-Triassic Upper Dalan-Kangan reservoirs in the Iranian offshore Fars area and their equivalent outcrops in the Zagros region, and briefly compares these to what has been documented elsewhere in the region.

Geologically the Khuff carbonate platform system is complex with fine-scale (centimetre-scale) heterogeneities which impact the static and dynamic reservoir characteristics. These heterogeneities are depositional, stratigraphic and diagenetic in nature and therefore reservoir characterisation needs to be an integrated multidisciplinary and multiscale approach. Nevertheless the facies and stratigraphic architecture of the Khuff is of fundamental importance in reservoir characterisation studies for four main reasons.

  • (1) Firstly, the Upper Dalan-Kangan reservoir development, quality and distribution are facies related because the main reservoir storage and best porosities are grainstone-dominated oolitic limestones. Mud-supported facies are generally non-reservoir (limestones), or generally very thin if dolomitised to reservoir quality. Moreover, the detailed depositional characteristics of the facies have an important impact on diagenetic facies and poroperm characteristics of the reservoir facies, these include: (1) mud versus grain-supported; (2) proportion of oolite versus peloids versus bioclasts (i.e. original mineralogy); (3) degree of micritisation; and (4) degree of sorting and packing. These subtle variations in depositional characteristics modulate the porosity and permeability values within poroperm distributions for particular facies groups.

  • (2) Facies distribution is important in terms of the distribution of reservoir-scale barriers and flow baffles, particularly the distribution of laterally continuous anhydrites, but also the very tight microbial facies which can vertically compartmentalise the reservoirs. These will impact the construction of static reservoir models and the strategy of production in field development.

  • (3) Indirectly, the facies type and distribution (particularly depositional poroperm and original mineralogy) affect the diagenetic processes, the distribution of the shallow diagenetic environments, and the subsequent diagenetic pathways. Thus, although diagenesis is the major impact on the final reservoir quality in Khuff reservoirs, the type of final diagenetic product has a strong depositional and stratigraphic component.

  • (4) Facies and depositional environments are important for the distribution of potential intra-Khuff source rocks which may contribute to the charging of the reservoirs. These include both poorly oxygenated organic-rich deeper-water embayments in marine conditions and more internal hypersaline lagoonal organic-rich muds (evaporitic source rocks of Warren, 1999).

To better constrain the spatial and stratigraphic distribution of depositional facies (including potential reservoir facies, intra-reservoir baffles/barriers and potential intra-Khuff source-rock facies) a synthesis of outcrop and subsurface data for the Upper Dalan-Kangan depositional system in the Iranian Zagros Mountains and offshore Fars area (abbreviated Zagros-Fars) has been carried out in order to:

  • (1) characterise the sedimentology and depositional environments and their stratigraphic distribution;

  • (2) develop a Gulf-wide facies classification based on detailed work on the Zagros-Fars area and comparisons with data from other areas;

  • (3) construct conceptual depositional models for the facies distribution for the Zagros-Fars area; and

  • (4) identify the major depositional sequences and construct a sequence stratigraphic framework based on stacking patterns, types of bounding surfaces and biostratigraphy (bio-event markers) which can be used to constrain stratigraphic architecture and reservoir layering.

Nomenclatural Note: In this paper the terms “Khuff” and “Upper Khuff” are used when referring to the stratigraphy of the Arabian Plate and to the general regional depositional system (including the Iranian part of this system). The terms “Dalan” and “Kangan” are used when referring specifically to the Iranian part of the depositional system (note the “Upper Dalan” is also equivalent to the “Upper Carbonates” of Szabo and Keradpir, 1978). Units prefixed with “K” such as “K2” or “K3” refer to reservoir intervals; units prefixed with “KS” such as “KS4b” or “KS1c” refer to correlatable stratigraphic intervals (interval name attached to the surface which marks the top of the interval); and units refered to as “Cycle IV” or “Cycle I” refer to third-order stratigraphic cycles (see figures in stratigraphic section below).

Study Location and Geological Overview

This paper focuses on the well-exposed Upper Dalan Member and Kangan Formation outcrops of the Zagros Mountains, Iran, and their equivalent in the offshore Fars subsurface (Figure 1). The Zagros Mountains outcrops provide excellent analogues of Middle East offshore reservoirs, and allow the seaward part of the Khuff reservoir Neo-Tethys platform system to be constrained. The outcrop sections that are described in detail are located at Kuh-e Surmeh and Kuh-e Dena and provide nearly continuous exposures from the top of the Nar Member (or its equivalent) to the base of the Aghar Shale Member (Dashtak Formation); thus the entire Upper Dalan-Kangan sequence (equivalent to the Upper Khuff) has been studied. The observations and interpretations based on these sections were then integrated with the offshore Fars subsurface descriptions, interpretations and models (see below).

Kuh-e Surmeh Section

The Kuh-e Surmeh structure is a faulted, core-eroded anticline, reversed toward the southeast, and the Permian-Triassic succession is located in the core of this anticline. The studied section, located about 120 km south of Shiraz, was measured along a narrow wadi that cuts through the northern flank of the anticline. This section comprises approximately 400 m in the Upper Dalan Member, Kangan and Dashtak formations. At this location the Permian-Triassic series directly overlie Ordovician shales, and the intervening sedimentary hiatus is generally interpreted as a result of a local palaeohigh located in this area (Szabo and Keradpir, 1978). This palaeohigh was effective and impacted sedimentation patterns until the deposition of the end-Permian Upper Dalan Member; thereafter there is no evidence for an active palaeohigh which affected sedimentation during the Early Triassic Kangan Formation (see below). The observed sedimentary reduction of the Permian series is considered to be related to the palaeohigh structure through erosion and stratigraphic thinning.

The Upper Dalan-Kangan interval is about 300 m thick (compared with approximately 450 m in the South Pars area). Its lower boundary corresponds to a 30–40-m-thick gypsum level, which is considered as equivalent to the anhydritic Nar Member (Szabo and Keradpir, 1978). The Upper Dalan Member consists of an oolitic shoal system which grades upward into deeper-water bioturbated mudstone and crinoidal wackestones. The Permian-Triassic Boundary is characterised by a thrombolite marker bed, as it is the case across the Khuff system in the Gulf region. The lower part of the Triassic Kangan Formation is mud-dominated and consists of laminated mudstone grading upward into thicker bedded bioturbated mudstone. In its upper part, the formation consists of oolitic and bioclastic grain-dominated facies. The Kangan Formation to Aghar Shale Member (Dashtak Formation) transition is characterised by columnar stromatolites followed by the green marls of the Aghar Shale Member.

Kuh-e Dena Section

The Dena Mountain sections, located about 200 km north of Shiraz, crop out as several-km-long cliffs along a NW-SE trend that parallels the main Zagros thrust zone. The first principal outcrop is located about 500 m above a small village accessible from the main road that goes through Pataveh. This section covers the Upper Dalan Member. The second section is located about one kilometer to the southeast of the previous one and comprises the Kangan and Dashtak formations. A third section (Potak Valley section), is located about 5 km to the southeast and covers the Lower Dalan Member. The composite section comprises approximately 400 m of Upper Dalan Member, Kangan Formation and Aghar Shale Member, of which the Upper Dalan-Kangan interval equates to 300 m.

The Upper Dalan Member (approximately 200 m thick) is directly underlain by a series of stacked lateritic palaeosoils and a possible palaeokarstic level. The lower part of the Upper Dalan Member consists of oolitic and bioclastic grainstones which grade upward into more open-marine packstones rich in fusulinids, echinid fragments and tabulate corals, and which dominate the upper part of the Upper Dalan Member. The Permian-Triassic Boundary is marked by a thrombolite bed. The Triassic Kangan Formation comprises outer-shelf mudstone similar to mud-dominated facies observed in the Kuh-e Surmeh area. The base of the overlying Aghar Shales is dominated by tidal-flat type facies.

Offshore Fars Subsurface

The subsurface equivalents of the Upper Dalan-Kangan interval are key hydrocarbon reservoirs in the region and consequently a large amount of subsurface data (logs and core) exists for these formations. These studies are based on the detailed examination of numerous fully-cored Upper Dalan-Kangan intervals. The subsurface Upper Dalan is commonly subdivided into the K4 and K3 reservoirs while the Kangan is subdivided into the K2 and K1 reservoirs. The total thickness of the Upper Dalan-Kangan sequence in the study region is approximately 450 m.

The lower part of the Upper Dalan is dominated by shallow-water restricted facies (lagoons and sabkhas) and small shallow-water oolitic tidal shoals. These are followed by larger bioclastic and oolitic sandwave complexes, which dominate this part of the section. The upper part of the Upper Dalan is mud-dominated, initially by shallow-water lagoonal muds, then by stacked peritidal muds, sabkhas and rooted beds. The lower part of the Kangan is dominated by shallow-water carbonate sands and thrombolitic facies. The upper part of the Kangan is mainly composed of peritidal and evaporitic supratidal flats with small tidal channels, oolitic shoals and local microbial patches. The base of the overlying Aghar Shales is dominated by evaporitic tidal-flat type facies and green and red shales.

Large-Scale Palaeogeographic and Stratigraphic Context

Palaeogeographic Context

The Khuff depositional system developed on the margin of the Neo-Tethys, an ocean separating the Gondwana supercontinent and the Gondwanan terranes (Cimmmerian megablock). The Khuff of the Arabian Plate and its correlative Dalan and Kangan formations of Iran are thus interpreted as reflecting a major tectono-eustatic event related to the onset of rapid thermal subsidence of the early Neo-Tethys passive margin in Arabia and Iran, and the drowning of its rift shoulders. This event has been interpreted as being coeval with initiation of the Neo-Tethys sea-floor spreading in the Middle Permian (Pillevuit, 1993; Sharland et al., 2001), although it has been suggested that the sea-floor spreading may have begun as early as the Early Permian (late Sakmarian) (Angiolini et al., 2003).

By the Late Permian, sea-floor spreading, thermal subsidence and the associated transgression led to the development of a very large epeiric platform shelf which stretched from southern Iran to Saudi Arabia. This regional epeiric platform had very low topographic relief and hence created extremely large facies tracts, many tens to hundreds of kilometers across (Alsharhan and Nairn, 1994b; Stampfli, 2000; Sharland et al., 2001). A consequence of such large facies tracts is that in order to encounter major facies changes, large areas need to be investigated.

The general palaeogeographic context of this system was a marginal marine shelf setting with an inner platform that was very flat, ramp-like, with little topography (Al-Jallal, 1987, 1994; Sharland et al., 2001), but with local depressions (Figure 2). The platform becomes more distal to the north and northeast with the presence of more open-marine influences (stenohaline organisms and ‘mid-ramp’ wackestones). In the inner shelf areas, where most of the reservoir facies sedimentation took place, depositional environments were prone to be restricted with limited circulation due to high carbonate production and low accommodation potential.

In this large, shallow, low-energy platform interior system it was necessary to create accommodation for extensive shoal development and hence reservoir development. The creation of accommodation allows the development of sufficient palaeobathymetry to generate open higher hydrodynamic energy conditions (by both wave and tidal action), and to provide the space for sediment to accumulate – both fundamental prerequisites for the development of thick good-quality grainstone reservoirs in such settings. This essential development of accommodation was initiated by platform flooding and hence transgression. Inversely, sea-level falls can quickly and drastically isolate the platform interiors reducing the internal hydrodynamic energy levels thus stopping the development of reservoirs over vast areas, and creating widespread evaporitic and carbonate mud seals. There are no modern analogues for such epeiric seas in terms of platform morphology and scale, nor extent of facies distributions and their sediment dynamics.

Another consequence of the large-scale and palaeotopography of the Khuff platform was that the shelf was easily isolated during high-frequency sea-level falls, and hence widely favoured faunal and floral provincialism. During periods of highstand or base-level lows faunal and floral populations were isolated by physical barriers such as extensive shoals tracts, from more open-marine palaeobiogeographical systems.

The palaeolatitudes for the Khuff platform were about 20–25°S for the Late Permian and 20–17°S for the Early Triassic (Stampfli, 2000; also see Angiolini et al., 2003). This places the platform in an arid to semi-tropical climatic belt. Moreover this suggests that at a large-scale, the Upper Khuff system evolved from a more arid system during the Late Permian, to a more sub-tropical climate during the Early Triassic as the system drifted northwards. This is consistent with the sedimentological and stratigraphic evidence from the study area which suggests more depositional evaporites and hypersaline lagoons in the Permian and more evidence of freshwater influence in the Triassic (see below). Moreover, during the Permian, the Khuff platform also experienced rising temperatures due to global warming following the Carboniferous-Early Permian glaciation, and the northward latitudinal drift towards lower tropic latitudes (Angiolini et al., 2003). This dynamic climate context is important because it may have played a role in the Permian-Triassic Mass Extinction event (Erwin, 1993).

Chronostratigraphic and Stratigraphic Context

Stratigraphically, this study is focused on the Late Permian-Early Triassic Upper Dalan-Kangan interval which are equivalent to the Upper Khuff in Qatar, United Arab Emirates and Saudi Arabia (Sharland et al., 2001; 2004). These stratigraphic units are subdivided into four reservoir intervals with the K4 and K3 being Upper Dalan (Late Permian) and the K2 and K1 being Kangan (Early Triassic). The global Permian-Triassic Boundary event is recorded near the K3-K2 transition (Figure 3).

The Dalan (lower and upper) and Kangan formations form a large second-order cycle with a major transgression within the Lower Dalan, a highstand located around the base of the Upper Dalan and a sequence boundary at the top of the Kangan (Figure 3) (Sharland et al., 2001).

The large third-order stacking cycles, defined on the basis of cycles bounded by surfaces representing base-level and accommodation potential minima, have been identified and regionally correlated based on surface characteristics, stratigraphic stacking patterns, faunal/floral events, and the large-scale regional context. These large-scale cycles provide the framework for the description and interpretation of the three main study areas of the offshore Fars subsurface, and the outcrop sections of Kuh-e Surmeh and Kuh-e Dena. The cycles defined are as follows (Figure 3):

  • Cycle IV: KS4 stratigraphic interval – lower Upper Dalan Member, and encompassing the K4 reservoir interval. The maximum flooding surface of this cycle would correspond to the P30 MFS of Sharland et al. (2001, 2004).

  • Cycle III: KS3a stratigraphic interval – upper Upper Dalan Member, and encompassing the lower K3 reservoir interval. This cycle includes the P40 MFS of Sharland et al. (2001, 2004).

  • Cycle II: KS3b and KS2 stratigraphic intervals – top Upper Dalan Member and lower Kangan Formation, and encompassing the upper K3 and K2 reservoir intervals. The maximum accommodation zone of this cycle corresponds to the Tr10 MFS of Sharland et al. (2001, 2004).

  • Cycle I: the KS1 stratigraphic interval – middle to upper Kangan and encompassing the lower K1 reservoir interval. The cycle maximum accommodation zone corresponds to the Tr20 MFS of Sharland et al. (2001).

The correlations and stratigraphic analysis suggest that the major stratigraphic trends and large-scale stratigraphic architectures are relatively ‘layer-cake’ at the reservoir geographic scale (10s to many 10s of km) and fourth-order stratigraphic scale, and are a function of the very large-scale nature of the carbonate platform and its flat ramp-like geometry. At a larger geographic scale, significant changes in thickness occur due to thickening towards palaeodepocentres (such as the Nar-Kangan fields intrashelf low during the KS4 and KS3; or the general thickening of the Upper Khuff towards the north and the east), or thinning and onlap towards palaeohighs (such as towards the Zagros Palaeohigh during the KS3 and KS4, and the general thinning towards the south in Saudi Arabia).

FACIES ASSOCIATIONS AND DEPOSITIONAL ENVIRONMENTS

A synthesis of the Gulf subsurface core descriptions and the Zagros outcrop facies data, together with integration of published data has allowed sixteen principal facies associations to be defined, characterised and interpreted (Table 1 and Figure 4). This facies classification and interpretation is based on: (1) macro and microfacies characteristics; (2) stratigraphic positioning and context; (3) types of sedimentary processes involved in facies genesis; and (4) faunal-floral content. Lithology is not explicitly used to define facies type as even very early dolomitisation tends to overprint various facies types depending on their stratigraphic position. It is nevertheless the case that certain facies tend to be dolomitic (internal platform suite of facies), whilst others tend to be limestones (high energy oobioclastic shoals and deeper mid-ramp muds). The facies associations are summarised and interpreted below, and in Table 1 and Figure 4, and Figures 58.

Anhydritic Facies Associations (Figure 5)

Facies F1 – Massive to laminated anhydrite dominated units

Description: The anhydrite in these facies can have a variety of forms including bedded to massive, finely laminated, nodular, chicken-wire, enterolithic and vertical palmate fabrics. The anhydrite concentrations are generally very high (over 80%) and occur in beds from 10 cm to 3 m thick. There are no sedimentary structures other than some cyclic variation in the amount and type of interstitial sediment and anhydrite type. Their thickness can vary from less than 0.5 m to over 3 m, although thinner beds are present.

Depositional environment and context: These facies developed as supratidal sabkha-type deposits (associated with nodular, chicken-wire and enterolithic anhydritic fabrics and tidal flat sediments) but mainly as shallow coastal salinas (associated with laminated, massive and vertical palmate anhydritic fabrics and hypersaline lagoon deposits). These facies may also result from extensive anhydrite replacement of previously deposited carbonates resulting in beds of massive, laminate or nodular anhydrite, but of later diagenetic origin. These are distinguished by their overprinting nature cross-cutting both sedimentary and stratigraphic fabrics. They also commonly contain ghosts of the original carbonate textures.

Facies F2 – Muddy (mudstone-wackestone) with anhydrite (greater than 20%, nodular (F2N) or laminated (F2L))

Description, depositional environment and context: These are a mixture of dolomudstone/wackestone with laminated (common), massive, enterolithic and nodular anhydrite. The deposits are composite including the host sediments (beige intertidal sediments: intertidal muds, microbial mud flats, mixed flats and intertidal sands) and the eodiagenetic anhydrite overprint. These units are characterised by: (1) numerous internal erosive surfaces (commonly marking deflation surfaces); and (2) a strong erosive surface at the top of the facies unit above which a marine flooding unit, often including lithoclastic debris, is deposited.

When this facies is ‘primary’ it is interpreted as upper inter-tidal to supratidal zone sabkhas (deposited sediments moved into the supratidal zone were early displacive, and to a lesser extent replacive anhydrite is precipitated). In these cases they are found capping shallowing-upward sequences. However also abundant are replacive anhydrites that can mimic depositional carbonate/anhydrite sequences. These anhydrite units are clearly replacive, cross-cut depositional fabrics and not directly related to the stratigraphic stacking. They reflect a later diagenetic overprint, often associated with the local remobilisation of anhydrite.

Facies F3 – Dolobreccia with anhydrite

These are a group of facies and features that reflect subaerial alteration of both dolomitic and evaporitic facies and the resulting meteoric diagenetic overprinting. The altered facies are often intertidal wackestones and mudstones containing evaporitic nodules, microbial lamination, rooted beds and desiccation cracks (facies F5, F6; see below). There are numerous internal exposure and erosion surfaces, and evapo-moldic fabrics may be present. They are generally from 30 cm to 3 m thick and are associated with F5, F6, F1 and F2 facies.

These brecciated dolostones represent evapo-karstic and karstic (epikarstic/surface karst) zones that have undergone evaporite and carbonate dissolution resulting in small vugs, dissolution seams and brecciation collapse. They can occur in both shallow-water limestone/dolostones (intertidal muds and sands) and evaporites sequences, and mark the top of the low-frequency regressive cycles. These zones are important to locate as they invariably define significant sequence boundaries.

Mud-dominated Facies Associations (Figure 6)

Facies F4 – Green-grey shaly mudstone (often dolomudstone) often laminated with liquefaction features

Description: Green shale to shaly mudstone-dolomudstone with millimeter-scale laminae. The thin organic shale laminae are often deformed and may have a wavy to lenticular aspect. They are commonly associated with thin (1–2 cm), black, laminated bands with well-developed laminations (F15; see below). They develop in low-accommodation, tidal-flat settings within small-scale (meter) high-frequency cycles.

Interpretation: These shaly mudstones are interpreted as developed within a marginal marine, shallow, subtidal to tidal-flat environment (intertidal mudflats), and may also reflect restrictive (reducing) waters (commonly in K1 and in Aghar Shale Member). The thin, black, laminated bands (F15) are interpreted as algal mats developed in such environments. The continental influence to such shales is confirmed by the palynofacies analysis, which showed significant amounts of coal material within the shales, suggesting significant terrigenous input. These shales are hence interpreted as continentally influenced marginal marine (to estuarine) deposits. These facies can be rich in organic matter.

Facies F5 – Massive to burrowed and rooted dolomudstone-wackestone

Description: These are beige coloured mudstones and wackestones (dolomudstones and dolowackestones) lacking any significant indications of intertidal influences (such as exposure surfaces, fenestrate, mudcracks, planar and wavy laminations), and hence are slightly more massive and less structure-rich; they are interpreted as being deposited beyond the influence of the intertidal zone in lagoonal settings. They can have a mottled macroscopic fabric suggesting bioturbation (in areas highly bioturbated - F5B) and in some cases can be well-rooted/vertically burrowed (F5R). Where these muds develop in restricted hypersaline conditions, early diagenetic mottling by replacive anhydrite cements develop (F5H).

Interpretation: These facies are interpreted as various types of lagoonal deposits, and based on observational criteria and their sedimentological context either as: (1) marine lagoon, (2) restricted lagoon, or (3) hypersaline lagoon. The depositional environment for marine-lagoon facies are protected low-energy shallow subtidal settings though with enough communication to the normal marine system to ensure a biological impact to the facies. There does not appear to be a strong stratigraphic context to these facies although they tend to be associated with microbial mud flats and intertidal muds. The restricted-lagoon facies reflect slightly more restricted lagoonal settings though not developing excessively high salinities to start precipitating syn-depositional evaporites. These tend to occur during regressive periods and are often overlain by microbial mud flats (F6) and burrowed intertidal muds (F5R). The sedimentary facies and fabrics of a hypersaline lagoon (F5H) indicate less evidence of organic activity (less bioturbation and tends to be of the small monospecific type if present). They have a distinctive early diagenetic overprint where anhydrite replaces the carbonate in a mottled fashion (this is often confused with burrowing); locally dispersed nodular anhydrite can develop where the anhydrite replacement is more advanced. These hypersaline lagoons develop in the most internal restricted parts of the platform interior. These different lagoonal facies are end-member types and complete transitions may be present.

Facies F6 – Laminated mudstone-wackestones

Description: Beige mudstones and wackestones (dolomudstones and dolowackestones) with various types of millimeter- to centimeter-scale laminations (planar, wavy, domal and wrinkled). The microfabrics can be dominated by peloids, and to a lesser extent fine bioclastic and lithoclastic material. They can also contain desiccation cracks, rare but well-developed teepee structures, minor exposure surfaces, fenestrate and occasional burrows. These reflect intertidal microbially-mediated carbonates (originally deposited as microbial mats and mudflats). Variations in the precise intertidal location (from upper to lower intertidal zones) control how well-developed these different features are (generally better developed within the upper intertidal zone, except for the bioturbation). These facies are often associated with evaporitic facies and anhydritic cements.

Interpretation: These are interpreted as microbial mud-flats facies deposited in non-saline low-energy protected intertidal flats, and tend to dominate the upper parts of the high-order regressive-half cycles which are eventually capped by sabkhas or exposure surfaces. Some of these mudstones are interpreted as intertidal muds, lacking evidence for microbial activity (such as domal, wavy peloidal micrites), but having the other features associated with the upper intertidal zone. In some places these thin intertidal mudstones can be intercalated with more grainy beds, forming units of interbedded intertidal sands, microbial mud flats and intertidal mud facies. In these cases flaser to lenticular bedding, and starved ripples are present. These mixed flats are transitional between intertidal sands and intertidal muds, and are invariably located at the tops of regressive fifth-order cycles.

Facies F11 – Very fine mudstone-packstone, often bioturbated

Description: Mudstone to wackestone with occasional thin silty and sandier levels containing bioclasts including echinoid fragments, sponge spicules and forams (F11C). They also form laminated to thinly-bedded dark grey-black mudstones, often mottled with bioturbation and intercalated with thin shaly levels (F11L). They are invariably limestone.

Interpretation: This facies is interpreted as low-energy shallow subtidal deposits associated with a more restricted embayment setting, and with periods of poorly oxygenated waters. The more massive embayment facies represent slightly deeper environments (deep lagoon to mid ramp) with limited circulation. In some areas these facies develop in poorly oxygenated shallow subtidal embayments with significant microbial activity (bedded and domal microbial fabrics; F11L). All these facies can grade into shallow-water facies (lagoonal muds, intertidal mudflat and oolitic/peloidal shoals) suggesting that these facies are rarely representative of deep-water conditions, and are mainly controlled by reduced water circulation. These muds can be rich in organic matter and TOC (total organic carbon) of up to 10% have been recorded within intercalated lagoonal/embayment sequences of the K3 interval.

Another subtype, including beige to grey-beige mudstones to packstones, is highly burrowed with abundant Zoophycos and Cruziana traces (F11B). This facies is interpreted as low-energy, shallow subtidal deposits associated with lagoons and inter-shoal areas (as evidenced by their intercalation with oolitic shoal facies). They can be associated with poorly oxygenated waters and therefore grade into more classical F11 facies. They are also often associated with very shallow-water lagoon and intertidal mudflat facies.

A variant of this facies, which is only seen in the K2 and K1 of Kuh-e Dena and Kuh-e Surmeh outcrops, has a well-bedded nature, can be rich in Claraia bivalves and has occasional slump structures (F14). These are interpreted as mid- to outer-shelf deposits and represent the deepest water, most distal facies of the Dalan-Kangan sequence.

Facies F15 – Thin cm to 5 cm finely laminated black shaly levels

These are thin centimetre-scale finely laminated black shaly to dolomudstone levels. The thin organic-rich shale laminae are often deformed and may have a wavy to lenticular aspect. They are often associated with chemical compaction-dissolution seams. These shale dolomudstones are interpreted as developing within a tidal flat environment with poorly oxygenated-reducing conditions. The thin black laminated bands are interpreted as algal mats developed in such environments.

Grain-dominated Facies Associations (Figure 7)

Facies F7 – Very coarse peloidal and bioclastic grainstone/packstone-conglomerate

Description: These are coarse gravels and polymictic conglomerates in a grainstone to packstone matrix and are generally poorly sorted. They contain angular to rounded clasts of various lithologies (cream mudstone, blackened pebbles and even evaporites clasts), and can be rich in bioclastic material. They can have low-angle planar beds and laminations, and in some areas contain small lags of either bioclastic or lithoclastic material. Minor exposure and erosion surfaces can also be seen. These units are generally thin (less than 50 cm) though sometimes a number of these beds can be stacked on top of each other.

Interpretation: Depending on their depositional context these are interpreted as relatively high-energy upper intertidal settings with significant sediment supply and represent marine pebbly beach-type environments, storm-lag sheets or the bases of channel lags. They are commonly preserved as transgressive lag deposits over erosive or exposure surfaces.

Facies F8 – Medium to coarse bioclastic grainstone-packstone

Description: These are medium- to coarse-grained bioclastic grainstones occasionally and packstones, admixed with significant amounts of oolitic and/or peloidal material. The bioclasts fall into two main groups: (1) bioclastic grainstones dominated by algae, foraminifera, mollusc debris (often also associated with F10 facies; F8I); and (2) bioclastic grainstones dominated by bryozoans, echinoderms and forams (often also associated with F9 facies; F8E). These bioclastic grainstones can have large well-developed foresets with gravely bases, smaller bi-directional cross-bedding and sub-horizontal planar laminations. Gradual transitions with oolitic grainstones, peloidal grainstones and lithoclastic grainstones are common.

Interpretation: The bioclastic grainstones with large foresets with gravely bases containing significant amounts of bryozoan and echinoderms material, and associated with well-developed oolitic shoals, are interpreted as large tidal sandwave complexes (seaward shoal complexes). The smaller bioclastic grainstone-packstone beds containing algal and forams, with small bi-directional cross-beds and subhorizontal planar laminations are interpreted as shallow subtidal (leeward shoals) to intertidal sands (swash sands), and are often associated with peloidal packstones. These types of bioclastic deposits can also develop as canalised storm deposits in the platform interior.

Facies F9 – Oolitic grainstone to packstone

Description: These are shallow subtidal grainstones and occasional packstones forming beds from 10 cm to over 3 m thick. The dominant allochems are oolites (although significant amounts of peloids, bioclasts and lithoclasts can be present). They possess various types of cross-stratification from large decimeter-scale foresets (planar and trough cross-bedding), small-scale bi-directional cross-bedding to low-angle planar laminations. Gradations from pure oolites to mixed oolite-peloidal grainstones (F9/F10) or mixed oolite-bioclastic grainstones (F9/F8) are common.

Interpretation: These sands were deposited in a host of submarine bedform environments ranging from large megaripples (associated with bioclasts and trough cross-bedding) to dunes that represent active mobile sandwave-shoal systems (large composite structures associated with large avalanche foresets), to smaller rippled bedforms that developed in the intertidal zone (with small-scale tidal cross-stratification to low-angle planar laminations). In some cases these oolite shoals represent tidal channel sands that cut into tidal flat muds. The precise nature of the depositional environment needs to be deduced from the bed thickness, type of sedimentary structures and the stratigraphic context.

During certain periods of Khuff deposition oolite facies appear to have had a very large geographic extent, with wide expanses of oolitic grainstones across almost the whole Khuff shelf area (an area of many hundred square kilometers stretching from Saudi Arabia to Iran). This wide geographic extent is in part a result of very low-angle progradation sheets that fill the accommodation from local palaeohighs in all directions on the platform top. These low-angle progradation sheets are difficult to resolve stratigraphically and hence difficult to correlate. Nevertheless, even taking into account the low-angle diachronous nature of these oolite sheets, it is clear that very large geographic areas were covered by oolites during certain periods of Khuff deposition. Considering the hydrodynamic energy of the platform system, it is clear that tidal action alone could not have acted across the whole shelf area where oolitic grainstones are found. This, together with the cross-bedding style suggests that there was significant wave and storm redistribution of oolitic facies across the shelf, in addition to the clear and well-developed tidal shoal – sandwave systems. It was a combination of these hydrodynamic processes that was responsible for the widespread distribution and redistribution of oolites during the maximum accommodation zones of the Upper Khuff. Such processes are consistent with the palaeolatitudinal and geographical setting of the Khuff platform during the Late Permian and Early Triassic which suggest that trade winds and storm tracts could easily have generated these wave and storm processes, in addition to significant local tidal amplification.

Facies F10 – Coarse to fine peloidal grainstone-packstone

Description: These peloidal grainstones-packstones consist of fine- to coarse-grained peloids, highly micritised bioclastic (bivalves, forams, algae) and small reworked micritic lithoclasts. They can have ripple laminations (wave-generated symmetrical forms), low-angle planar cross-stratification, sand laminations, small-scale trough cross-bedding, and reworked lithoclasts. Minor exposure, erosion and scour surfaces are frequent.

Interpretation: Their depositional settings are shallow subtidal to intertidal environments located within an internal platform context of moderate energy, and in some cases could reflect relatively well-sorted beach sands (shoreface), sand sheets, washover sands, channel lags or leeward side of oolitic sand bodies. They are often intercalated with muddy facies as the system gets more protected.

Facies F16 – Oolitic-peloidal grainstone with pebbles and microbial influence

Description: Coarse oolitic-peloidal grainstone with pebbles (often oval to flat) developed under microbial influence and local oncoid development. In areas of high microbial influence, the facies is tightly cemented by low-Mg calcite (microbially mediated) and has large lithoclasts and oncoids/microbially coated pebbles (F16C). In other areas there is less microbial influence and the facies is more porous, less well-cemented and more oolitic in nature (F16P). Reworking of thrombolites, very early marine cementation (forming cement crusts and hardgrounds), microbially-bound sediments and local oncolite development are common. At a microscopic scale, this facies is characterised by coarse calcite isopachous fibrous cements and peloids without well-defined boundaries (commonly associated with microbial peloids).

Interpretation: Depositionally these facies are interpreted as relatively high-energy upper intertidal to shallow subtidal settings and represent pebbly beach-type and sand-sheet environments. They are dominated by early cementation (hardgrounds, reworked lithoclasts), microbially-mediated cements and microbial bounding (grapestone-type textures can develop). They are principally found just after the Permian mass extinction and the first appearance of sea-floor cement crusts marks the precise point of the last Permian fauna. This suggests a possible cause-effect relationship between the syn-sedimentary seafloor calcite cements and the cause of the Permian extinction. For further discussion on this facies and its relation to the Permian-Triassic events see the section ‘Permian-Triassic Transition’ below. The abundant lithoclastic material found around the thrombolitic interval is thought to be due to reworking under progressively storm-wave and fair-weather wave-base conditions of early lithified (microbially-mediated) sediments (including beach rock). The grainy sheets are interpreted as shallow subtidal-intertidal oopebble storm-generated sheets.

This hitherto unrecognised but distinctive facies is of significant regional stratigraphic importance and a valuable reservoir and regional correlation marker unit for the end Permian extinction and the following Permian-Triassic Boundary (see section on the Permian-Triassic Transition below). This facies is present on all the Permian-Triassic platform successions seen in the area (personal observation: Abu Dhabi, Iran, Qatar and Saudi Arabia), even where the more classic thrombolite unit is very poorly developed or absent (such as in Abu Dhabi).

Microbial-dominated Facies Associations (Figure 8)

Facies F12 – Thrombolitic-stromatolitic boundstone

Description: This is a microbial thrombolitic fabric showing a clotted macroscopic texture of micritic masses tightly cemented by calcite. This facies is closely associated with very coarse well-cemented lithoclastic grainstones and black laminated algal mats (see below). The facies is best developed and most widespread just after the end-Permian extinction event and defined in this study as the base of the Triassic Kangan Formation. Within the macroscopic clotted thrombolitic texture there is often a very fine micritic material that fills small cavities. This micritic material has a blue-grey colour and appears to be an unusual marine precipitate as suggested by its distribution and coating nature. In platform top environments the grainstones that follow the thrombolites are often closely associated with the microbialite (microbially cemented or microbially mediated cements; facies F16 see below). Systematically, just before the clotted thrombolites of the Permian-Triassic transition, is a series of more thinly (cm) bedded to laminated microbialites (F12L) which also appear to have the blue-grey marine micrite cement and associated with more classic early marine fibrous and botryoidal cements.

Interpretation: These facies formed in shallow subtidal to intertidal environment, perhaps under peculiar chemical or ecological conditions associated with the end-Permian extinction. The thrombolitic unit developed within a packet of lithoclastic rock (F16) is interpreted as shallow subtidal to intertidal microbial/thrombolitic patches in lower energy more protected conditions, and associated with black algal mats (F15). These microbial units are useful as a broad regional isochronous marker horizons, since they mark the approximate position of the Permian-Triassic Boundary across the region; they are present from Ghawar field Saudi Arabia (Al-Jallal, 1995), United Arab Emirates and Qatar (personal observation), offshore Iran and Zagros outcrops in Iran (present study).

Facies F20 – Microbial bound fabrics

Description: These are dense well-cemented, microbial sedimentary textures that are often thinly bedded to laminated, and associated with microbial conditions (levels with thrombolites and stromatolites) (leolites of Braga et al., 1995). They are very common within the KS1b unit, which is a microbial-rich period within the K1 similar to that of K2. Two broad end-members are identified: (1) microbial cemented dense mudstones to wackestones textures (F21); and (2) microbial cemented dense wackestones to grainstones textures (F22). Associated with these levels are peloidal packstones with blackened grains suggesting reducing conditions. This blackened grained peloidal packstone facies is also closely associated with the KS1b microbial period.

Interpretation: These are essentially microbially bound shallow subtidal to lower intertidal facies (microbially bound sandflats, mudflats and sandy lagoons) related to a period of microbial development and early marine cements in the KS1b. The blackened grained peloidal packstones are suggestive of reducing conditions, poor oxygenation and oxides coating grains.

Summary

The facies associations developed integrate the core and outcrop data where 16 principal facies were identified (Table 1). In general, importance of microbial facies is highlighted and a variety of microbial facies are defined (F12T, F12L, F21, F22). In particular, in addition to the well-known K2 microbial event, there are a suite of microbial facies that have previously been underestimated for the K1. Microbial units are useful as regional isochronous marker horizons, for example the K2 thrombolite is present from Ghawar (Saudi Arabia) to the Zagros (Iran) (personal observation). These microbial facies are often associated with periods of poor oxygenation and restriction, but nevertheless can occupy a range of environments from intertidal to mid- to outer-ramp settings.

Organic-rich Upper Khuff facies include green shales (restricted to K1), and facies related to restricted poorly oxygenated embayments (F11, F11L) and hypersaline lagoons and tidal flats (F5, F5h, F6). These facies can be rich in organic matter and TOC of up to 10% have been recorded in the K3 within a packet of intercalated lagoonal/embayment F11, F5 and F6 facies. These facies are often related to hypersaline systems and can be regarded as ‘evaporitic’ source rocks (Warren, 1999). The main reservoir-prone facies are essentially grainstone and packstone (in terms of reservoir quality and thickness), particularly the oolitic grainstone (F9), oobioclastic grainstone (F9–F8, F8) and the microbially-influenced oolithoclastic grainstone (F16).

These facies have been interpreted in terms of depositional environment including: (1) evaporitic flats or saltern (supratidal/intertidal to subtidal setting); (2) tidal flats (intertidal to supratidal setting, with numerous subenvironments such as beach ridges, intertidal flats, tidal channels, microbial mud and sand flats); (3) subtidal lagoon (subtidal setting, including microbial build-ups); (4) leeward shoals (subtidal to intertidal setting); (5) oolitic to oobioclastic shoal belts (subtidal to intertidal setting); (6) composite sandwave constructions (subtidal setting); and (7) middle shelf (subtidal setting) and outer-ramp settings (see Table 1). This large range in facies types documented in the Upper Khuff sequence reflect the large range in depositional systems and sub-systems present across the Khuff platform (ranging intertidal and supratidal flats, evaporitic salterns to mid-ramp deposits), but also the temporal evolution of the Khuff environments and palaeoceanographic conditions from the Permian to the Triassic with periods of low oxygenation (and possible anoxia) to periods of supersaturated oceans just after the Permian-Triassic extinction event.

Qualitative comparisons of Upper Khuff sections and subsurface cores across the Zagros area, Gulf-Arabian Peninsula (Al-Aswad, 1997; Al-Jallal, 1987, 1994, 1995; Alsharhan,1993; Alsharhan and Kendall, 1986; Alsharhan and Nairn, 1994a, 1994b, 1997; El-Bishlawy, 1985; Schlumberger, 1981), show that this depositional facies classification is applicable at a larger regional scale (including Saudi Arabia, United Arab Emirates, Qatar, Iran, Bahrain, and Kuwait) and useful in rapid regional comparisons and correlations of the Upper Khuff depositional systems. The depositional facies classification should act as a rapid means of studying and comparing the depositional facies of different Upper Khuff sequences, and their corresponding reservoir impact, at both reservoir and regional scale. This facies classification and interpretation form the basis of the stratigraphic analysis developed for the study area.

STUDY SECTIONS - SEDIMENTOLOGY AND STRATIGRAPHY

The detailed sedimentology, evolution of depositional environments and stratigraphic interpretation of the three principal study areas, Kuh-e Surmeh, Kuh-e Dena, and the offshore Fars subsurface is outlined below. The descriptions and interpretations are placed within the large-scale stratigraphic framework outlined previously (see Figures 3 and 4).

Offshore Fars Subsurface

The interpretation for the offshore Fars subsurface area represents a synthesis of subsurface data located in the same palaeogeographic domain, and is summarised below and in Figure 9.

Nar Member (Median Anhydrite Member)

The Nar Member consists of anhydrite with occasional thin dolomicritic intervals. The anhydrite is generally massive to nodular with intervals of bedded-mosaic, nodular-mosaic and distorted nodular-mosaic fabrics. Accessory lithofacies are fine-grained peloidal dolograinstones and dolopackstones, and thin dolomudstone streaks. The top surface (SB NAR) is a strong exposure (with pedogenetic features) and subsequent erosive surface with clasts of eroded evaporites present within the subsequent transgressive lags. Depositionally this unit represents sabkhas, shallow coastal salinas and hypersaline lagoons, with occasional floodings resulting in carbonate deposition. There is also an overall evolution from hypersaline lagoons to coastal salina and finally sabkhas and sabkharised salinas. The Nar Member in this region represents a semi-enclosed saline depression prior to the deposition of the Upper Khuff.

Cycle IV (KS4): Upper Dalan Member (Figure 10)

The KS4 composite depositional sequence has been subdivided into three major intervals (fourth-order depositional sequences): the KS4a at the base, KS4b in the middle and the KS4c at the top. Each of these depositional sequences comprises various depositional units (such as KS4a0, KS4a1, KS4a2, KS4a3, KS4a4, KS4b1…) which equate to parasequence sets and are separated by parasequence set boundaries.

Depositional Units in KS4a (Cycle IV-a)

The base of the KS4a is characterised by cycles with interbedded coarse peloidal packstones (including rip-up lithoclasts, F10, F7), internal bioclastic grainstones/packstones (F8), and finer rooted/vertically burrowed dolowackestones (F5r) and hypersaline lagoons (F5, F5h). This is followed by cycles that are increasingly peloidal and bioclastic (F10, F8/10), but are still associated with well-developed lagoonal and tidal-flat dolomudstone and dolowackestone units (F6, F5, F5/10). The cycles then become progressively more grain-dominated, including oolitic/peloidal fining-upward grainstones (F9/10), with thinner mudstone lagoons and intertidal mudflat cycle caps. Within these grainstones evidence of tidal cross-stratification and beach laminations is present and the cycles tops have small exposure surfaces; the cycles are about 1–2 m thick. The thickest grainstone cycles with the most well-developed high energy conditions and oolite/peloid/bioclastic grainstone development represents the maximum of accommodation of this unit. The top of this unit is usually marked by well-developed muddy cycles dominated by hypersaline mudstone-wackestones (F5, F5h), evaporite beds (F1) and capped by rooted/vertically burrowed mudstone/wackestone (F5r). Exposure and erosion surfaces are present at the cycle tops and collapse dolobreccias (F3) can also develop.

Depositional environment: The grainy deposits at the base of this depositional unit represent transgressive lags. These are followed by shallow restricted lagoons, evaporitic mud flats and mixed flats, and stacked sabkhas. The first appearance of significant tidally influenced sands represent small shallow subtidal to intertidal protected (leeward) shoals to occasional well-developed tidal oolitic shoals (probably shallow subtidal channel shoal/dune complexes). The top of the parasequence set is capped by restricted hypersaline lagoons, which are occasionally rooted/vertically burrowed, stacked sabkhas and shallow coastal salinas.

Depositional Units in KS4b (Cycle IV-b)

This unit starts with muddy and evaporitic internal facies and then rapidly develops into a series of small fining-upwards peloidal and bioclastic packstone-grainstone (F10, F10/8), with tidal structures in the upper parts. These are then followed by a series of bioclastic and oolitic grainstone-dominated cycles (F8, F9/8, F9) approximately 2–3 m in thickness. These cycles are completely grain-dominated (no mud-cap) and start with a less cemented oolitic grainstone and reflect a deepening-upward motif; they are often capped by a cemented bioclastic grainstone. This deepening upward well-cemented bioclastic cap motif is very characteristic and easily identifiable as this unit on numerous wells (including in Abu Dhabi and Qatar). The cemented cap represents early marine cements and localisation of latter cements around the more bioclastic-rich cycle tops. Tidal cross-sets are common through-out the grainstone sequences. This unit marks the first appearance of thick well-developed tidally influenced transgressive sands in the Upper Khuff succession. The maximum accommodation zone of this unit occurs in the upper part of the thick oolitic, bioclastic-capped grainstone cycles. The top of the sequence is capped by a tidal oolitic grainstone sequence, followed by a significant exposure erosional surface, with the development of oomouldic local vadose silts.

Depositional environment: The base of the parasequence set represents transgressive lags, intertidal sandflats, microbial mudflats and hypersaline lagoons. These progressively evolve into intertidal sandflats, shallow peloidal-oolitic shoals, and intertidal sandy scour/channel beds. The start of the grainstone-dominated cycles are interpreted as shallow subtidal-tidal oolitic and oobioclastic shoals with early cemented bioclastic caps (in the deeper water). The bioclastic caps probably reflect the deepest part of the cycle, and hence are richer in bioclasts and more stabilisation-cementation (the bioclasts also act as local sources for later block cement development). These are inferred to be megaripple-like bedforms.

Depositional Units in KS4c (Cycle IV-c)

KS4c1: This unit starts with a series of thin erosively-based bioclastic, peloidal and oolitic grainstone-packstone units followed by an oolitic grainstone unit which has a cemented cap and exposureerosion surface at its top. This is followed by thin bioclastic transgressive beach lags, and then rapidly by a dark grey, often anoxic, open-marine mudstone-wackestones (F11). These are followed by a series of very fine-grained laminated bioclastic and peloidal micropackstones (F10/8). The top of this depositional cycle set is marked by an exposure surface and then clear transgressive erosive surface (base KS4c2). The maximum accommodation for this parasequence set is placed in the cycle containing the dark grey open-marine mudstone, which can be regarded as the maximum flooding surface (i.e. the interval of maximum depth) of the KS4 interval.

KS4c2: This unit starts with a remarkable development of a thick oolitic and bioclastic grainstone interval. The grainstones are in medium- to coarse-grained (sometimes gravely), with a good to bimodal distribution. Other grain constituents are peloids and mollusc fragments, with the occurrence of bryozoans and echinoid fragments towards the bottom part of the interval where the bioclastic grainstones/gravels (F8, F7) to oolitic/bioclastic grainstones (F9/8) dominate. Asymptotic cross-bedding, occasionally bi-directional, are frequent. Low-angle cross-bedded grainstones occurs throughout the parasequence set (often capping grainstone cycles). The individual depositional cycles are difficult to identify since mud-caps are absent (not deposited or eroded away) or very thin (centimetre to tens of centimetres). Rare inter shoal bioturbated mudstones-wackestones (F5, F11) are present. In the upper part of this parasequence set the oolitic grainstones (F9) dominate and are better developed (tidal cross-stratification well-preserved), bioclastic beds are thinner and less frequent. Cycle mud-caps are virtually absent and the cycles are about 3 m thick. The maximum of accommodation is in the zone of thick aggradational oolitic and bioclastic grainstone packets.

KS4c3 and KS4c4: Progressively the upper part of this unit becomes more mud-dominated with a laminated and bioturbated (Zoophycos) lagoon embayment wackestones (F11) and lagoonal to hypersaline lagoon mudstones (F5, F5h). The cycles vary from 1–3 m in thickness. Anhydrite nodules are present throughout. The cycles are generally completely mud-dominated (mud > 70% of cycle). The topmost 2 m is a brecciated anhydrite dolomudstone unit composed of microbial laminates, with wavy and crinkled laminae, which have been brecciated and partially cemented by anhydrite. The top of this dolobreccia marks the top of this interval (and the KS4) and has been interpreted as a major sequence boundary (SB) and dolobreccia (surface karst) unconformity.

Depositional environment KS4c1–KS4c3: The base of the unit represents a stack of transgressive shallow-marine bioclastic-peloidal-sands, followed by oolitic shoals. The cycle with the dark embayment facies is interpreted as low-energy subtidal open conditions. This may relate to a significant flooding (drowning) of the platform or could also be related to an oceanographic change (change in circulation, less oxygenated waters leading to a suppression in the carbonate platform productivity), or a combination of the two. The very fine-grained laminated bioclastic and peloidal micropackstones that follows the muddy facies are interpreted as shallow-water higher energy ‘open’ shoreface facies. The change from the top of the clean coarse-grained oolitic grainstone shoal to muddy open-marine facies (maximum flooding of the KS4c) is very rapid – these events occur over a few tens of centimetres. It is questionable if this can simply be related to sea-level rise across the ramp, particularly since the more ‘distal muds’ show no sign of their arrival progressively within the KS4b3 (one would expect progressively deeper facies to be increasingly present through the KS4b3 and into the KS4c1 – this is not the case). Neverthelesss the detailed faunal analysis does pick-up a deepening up signal in the KS4b stratigraphic interval (see below), suggesting that a deepening trend is present. The KS4c cycle, could be considered as a platform flooding unconformity – probably associated with another oceanic event (restricted circulation, nitrification – eutrophic event (change in community trophic structure)).

The stack of grainstones that follow represent a series of open-marine transgressive shallow-water subtidal to intertidal sands (transgressive lags, sandflats and beaches). Palaeocurrent and image log data suggest southerly dominated, but tidally-influenced, migration patterns. These quickly evolve into oolitic subtidal sand flats to oobioclastic shoals. The main part of this depositional unit represents coarse transgressive bioclastic sandwave/megaripple complexes at the base, and stacked mobile oolitic shoal belts at the top. All the grainy facies that dominate this parasequence set reflect relatively high energy, open conditions. The thickness of the parasequence set, thickness of individual cycles (up to 3–4 m) and their facies character suggest that this is a high accommodation, aggradational setting. The muddy to sandy mud represent inter-shoal lagoon environments. There is a notable absence of intertidal mud or mixed flat facies in this unit.

The laminated and bioturbated (Zoophycos) wackestones-mudstones in the upper part of the cycle represent subtidal lagoon and embayments with poorer water circulation. The top of the cycle reflects a series of hypersaline lagoons intertidal mud-flats and sabkhas. The massive anhydrite and brecciated dolomudstone at the top of this unit is interpreted as a solution collapse due to partial dissolution of intervening evaporites and carbonates during a significant exposure episode.

Large-scale Depositional Sequences

The sequence stratigraphic interpretation is based on the stacking patterns and depositional environment interpretation. The three major depositional sequences (KS4a, KS4b, KS4c) represent fourth-order transgressive-regressive cycles. The overall transgressive-regressive cycle of the KS4 (from the top Nar Member surface to the major dolobreccia surface (major SB) at the top of the KS4) can be assimilated into a third-order transgressive-regressive stacking pattern. They are summarised in Figures 911.

KS4a Depositional Sequence (Figure 12): The initial transgression in the KS4a depositional sequence is represented with KS4a1 and KS4a2 parasequences sets displaying a similar internal pattern. Evaporitic flats, lagoons and tidal flats are dominant, and small-scale sequences reflect the common evolution of such environments with stacked metric sequences grading from subtidal lagoons to lower intertidal laminates, and upper intertidal to supratidal carbonate sands and anhydrite. The cycles increase in thickness and graininess (and hence porosity development) towards the KS4a4 unit. In the KS4a3 and KS4a4 parasequence sets small oolitic shoals can develop and these represent the maximum of accommodation in this fourth-order KS4a cycle, and the most open-marine conditions. The maximum accommodation zone is placed in KS4a3. The regressive half-cycle of the KS4a cycle is mainly represented by the KS4a5 parasequence set that is dominated by small peloidal-oolitic shoals, internal lagoons and mud tidal flat deposits.

KS4b Depositional Sequence (Figure 13): KS4b is a fourth-order stratigraphic cycle which is dominated by well-developed grainy facies. This cycle is again very asymmetric with KS4b1 and KS4b2 developing within the transgressive half-cycle, and the maximum of accommodation developed within the upper part of KS4b3. The cycles get grainier and more bioclastically-dominated towards the maximum accommodation zone (upper KS4b3). The regressive facies at the top of KS4b3 are represented by small oolitic grainstone shoal facies and intertidal sand flats.

KS4c Depositional Sequence (Figures 14): This depositional sequence is the most important in terms of thickness but also reservoir development. The first fifth-order stratigraphic cycle KS4c1 is highly condensed (about 6 m thick) and contains significantly deeper and/or poorly oxygenated muds. It is unclear whether this significant flooding represents the major accommodation development of the KS4 sequence (platform completely flooded – deeper ramp deposits) or whether this unit reflects a change in the platform/oceanographic organisation (temporary shut-off/slow-down of the carbonate factory), or a combination of the two. The KS4c1 cycle could be considered a platform flooding unconformity – probably associated with another oceanic event.

The KS4c2 sequence is the thickest KS4c parasequence set and has good reservoir characteristics. The individual cycles are the thickest, most grainy and most dominated by bioclasts. The maximum accommodation development of the whole of the KS4 is located within the grainiest, thickest and most bioclastically-dominated cycles of this parasequence set, which is significantly aggradational. The high accommodation of this parasequence set is also reflected by the fact that it is the thickest in the whole KS4 cycle (about 30 m thick). KS4c3 and KS4c4 reflect a decreasing accommodation system and consequently beds and cycles get thinner, mud cycles become better developed and consequently porosity decreases. Also notable is that the KS4c fourth-order stratigraphic cycle is asymmetrical with the regressive half-cycle much more significant (with a maximum accommodation zone in KS4c1 or KS4c2). This compares with the more transgressive-dominant half-cycles of the previous two fourth-order stratigraphic units (KS4a and KS4b). This reflects the dominant flooding of the overall third-order trend.

Cycle III (KS3a): Upper Dalan Member (Figure 15)

The base of this unit is often marked by two well-developed anhydrite beds with associated exposureerosion surfaces. This is followed by peloidal, lithoclastic and bioclastic grainstone-packstones in the lower KS3a. The rest of the unit is a series of mudstone-dominated cycles (dominated by burrowed F11 and F11L facies) intercalated with thin bioclastic and peloidal grainy beds (5–30 cm in thickness). The cycles generally start with an erosively-based grainy bed then fine-upwards into an F11 or F11L mud. Small tidal peloidal shoal-channel can occur, though rare. The cycles are completely mud-dominated and up to 2–3 m thick. The maximum accommodation zone is placed in thick (several meters) dark muddy (F11) unit. The top of this parasequence set is marked by a stack of lagoonal mudstones followed by a well-developed dolobreccia unit and exposure surface.

Depositional environment: The lower KS3a cycles developed in a restricted coastal lagoonal-embayment environment (possibly poorly oxygenated) with occasional transgressive lags representing slightly more high energy conditions at the base of cycles (sub- to intertidal). The tops of cycles represent intertidal mudflat and supratidal sabkha-shallow salina environments. The mud-dominated middle KS3a is more reflective of restricted embayments (with thin grainy stringers), the dark grey colour and high organic content suggestive of poorly oxygenated quiet water environments. Other subtidal environments present in the KS3a are shallow sandy lagoons-embayments and lagoons. No major exposure surface occurs at the tops of the cycles in the middle two-thirds of this unit suggesting that the cycles did not shallow into intertidal conditions. The upper KS3a is dominated by the shallower-water systems including microbial mudflats, mixed flats and well-developed stacked sabkhas and shallow salinas. The top is marked by a brecciated and altered lagoon, hypersaline lagoon and microbial mud flat unit, and results from a significant correlatable exposure surface.

Large-scale depositional cycles (Figures 9 and 16)

The KS3a represents a third-order sequence; the fourth-order cycles are not evident and have yet to be correlated. It is possible that they are either highly condensed or reduced/lost during the trangressive onlap. The base is bounded with the remarkable dolobreccia of the KS4 and the top KS3a is marked by a thick unit of tidal muds and another significant dolobreccia unit. The flooding parasequence sets (KS3a1 and KS3a2) are dominated by a series of vertically stacked evaporitic deposits, minor tidal shoals (only near base), tidal flats and eventually by sandy lagoonal and restricted embayment facies. The maximum accommodation zone of the cycle is within the middle part of KS3a3, here there is a restricted (poorly oxygenated) embayment facies (F11), where individual parasequences are difficult to identify since exposure did not occur at the cycle tops. The regressive limb of this third-order cycle (upper KS3a3 and KS3a4) is dominated by lagoonal muds and muddy tidal and supratidal flats.

Lower Cycle II (KS3b): Upper Dalan Member (Figure 15)

This unit consists of mud-dominated cycles (F5, F5h, F5r, F6, F11) with minor grainy bases (F8/F10 and F7) and the sequence is characterised by numerous erosive surfaces and coarse pebbly lag deposits (F7, F10). The cycles are fining and muddying upwards and between 0.5 and 1.5 m thick; however, the system is highly autocyclic and thus the small-scale allocyclic cycles are not always easy to identify. The tops of the cycles are commonly cemented mud caps that get reworked at the base of the following cycle. Some of the grainstone beds, especially those with significant amounts of lithoclasts, represent canalisation (cannibalisation) of the muddy intertidal flats. The depositional unit is capped by intertidal lagoonal muds, laminated microbial flats and vertically burrowed/rooted mudflats. Within this mud-dominated unit are a series of peloidal sands very rich in foraminifera, and which are correlatable at a field-scale. This unit is highly aggradational and this facies and cycle motif repeats itself with little variation through-out the KS3b sequence.

Towards the top of this unit the system becomes progressively more grainy with the development of series of intercalated peloidal and bioclastic packstone-grainstones (F8, F8/F10, F10) fining up into rooted/vertically burrowed mudstones-wackestones (F5, F5h, F5r) with lithified mudcaps. Significantly there is the reappearance of oolites in some of the grainstone-packstones (F9/F10, F10/F9) (essentially absent from the top of the KS4 – 100 m further down) and the grainier beds are slightly more prevalent, coarser grained and start to show slightly more high energy conditions (cross-lamination, sub-horizontal lamination). The top of the parasequence set is marked again by a well-developed dolobreccia unit (F3) composed of brecciated, altered and cemented F5h and F6. This unit is often capped by a marked epikarst.

Depositional environment: The system is dominated by shallow subtidal to intertidal lagoons (often hypersaline, rooted/vertically burrowed and subsequently exposed and cemented), channel lags, and muddy, sandy and mixed flats which are highly channalised with lots of ripped-up clasts of prelithified sediments. Locally more poorly oxygenated embayment facies (F11) are present. The lithified mudflats are common and reflect significant exposure of the mudflats to allow lithification and then erosion by either the following cycle or a tidal channel. In shallower water muddy intertidal sands, mixed flats and microbial flats prevailed. The anhydrite bed is interpreted as stacked sabkhas and shallow coastal saline ponds.

The bioclastic and peloidal sands represent shallow subtidal to intertidal mixed to sandflats and channels. The grainstone packets are strongly erosional with scour lags, rich in lithoclastic material, represent cannibalistic tidal channels/creeks. The top of the sequence is a significant exposure of intertidal muds and lagoons resulting in brecciation. The increase in grainy beds and significant reappearance of oolites at the top units reflects the progressive opening-up (gradual flooding) of the platform and installation of high-energy hydrodynamic system. This is corroborated by the palaeoecology which suggest progressively more open-marine conditions towards the top of the KS3b and KS2 (see below).

Large-scale depositional cycles (Figures 9 and 17)

This is a fourth-order sequence located in the transgressive half-cycle of a third-order sequence (which also comprises the KS2 fourth-order cycle). This results in the remarkable transgressive aggradational nature of the stacking patterns resulting in a thick stack of relatively ‘internal’ facies. The facies present within the KS3b always remain shallow-water ‘internal’ platform facies including bioclastic and peloidal sands representing shallow subtidal to intertidal mixed to sandflats, cannibalistic tidal grainstone-packstone channels, mudstones representing shallow lagoons, hypersaline lagoons, and rooted/vertically burrowed mudflats-microbial mudflats. Palaeobathymetry was never greater than a few meters throughout the KS3b deposition. The grainier internal facies tend to be concentrated at the bases of the parasequence sets. The maximum accommodation zone of the KS3b is located in the grainy facies of the KS3b4, giving a highly asymmetric, transgressively-dominated, cycle.

Upper Cycle II (KS2): Upper Dalan Member and Kangan Formation (Figure 18)

The KS2 depositional sequence has been subdivided into four depositional units (parasequence sets): KS2a, KS2b, KS2c and KS2d.

KS2a: This unit starts with thin interbedded bioclastic and peloidal packstones-grainstones (F8, F10) then fines and muds upwards to vertically burrowed/rooted wackestone-mudstones (F5, F5r). The following cycles are markedly grainier, cleaner and coarser-grained, and a series of grainstone-dominated cycles begins. These grainstone cycles (F8/10, F10/9, F9) are composed of oolites, peloids and bioclasts, and exhibit well-developed tidal cross-stratification and low-angle laminations. These grainstones mark the return of well-developed tidal oolitic sands. The top of the parasequence set is marked by a muddy parasequence (F5, F6) and a well-developed brecciated mudstone. The cycles are 1–2.5 m thick, and the maximum of accommodation is placed within a 2.5 m thick tidal oolitic sand-dominated cycle.

The depositional environment of the KS2a represents significantly higher energy, more open-marine systems than in KS3b (this is also confirmed by faunal data; see below). The grainy facies represent a stack of transgressive tidally-active shallow subtidal to intertidal oolitic shoals, with sand and mixed flats in the more protected intertidal areas. These grainy cycles represent a significant flooding and opening-up of the platform system over the very internal muddy platform interior facies of the KS3b units. The base of this packet is interpreted as a flooding surface. The more significant mudstone units at the top and base of the parasequence set represent shallow lagoons, embayments, hypersaline lagoons or vertically burrowed mudflats with cemented exposed caps. The top of KS2a is also marked by a well-developed epikarst.

KS2b: The lower transgressive interval of this unit starts with a dolomite parasequence composed of oolitic and bioclastic coarse-grained sands with tidal structures (the bioclastic sand is full of Permian fauna which are evidently not reworked). These basal bioclastic and oolitic grainstones are interpreted as small shallow subtidal to intertidal wave and storm-generated sands, and transgressive bioclastic lags. Progressively within the oolitic grainstone there is an increase in medium-grained ooids and very coarse grainstone intraclasts, cement crusts and oncoids (reflected by a gradual evolution from F9 to F16). Sorting is very poor and shapeless to elongate lithoclasts and oncoids are cemented by a mosaic of calcite crusts (isopachous low Mg calcite fringes). These occur within fining-upward sequences, approximately 1.5 m thick and grain dominated. Within this high-frequency cycle the Permian fauna is progressively lost.

The following cycles are almost completely dominated by grainy facies, generally composed of coarse lithoclastic beds (F16 - oolitic microbially influenced sand sheets with cement crusts and lithoclasts), and range from 1–2 m in thickness. Progressively small (10 cm) thrombolitic layers (F12) are interbedded with the F16 facies. The topmost layers, just before the thrombolites, consist of laminated clotted microbial facies (F12L). This is followed by a 1–3 m thick thrombolitic layer with a typical clotted aspect (F12). After the thrombolites there is a return to a series of coarse intraclastic, oolitic and oncolitic grainstone cycles (1–2 m thick), with varying degrees of microbial bounding. Oolitic cross-sets reappear after the thrombolites and thus the sequence around the thrombolites is quite symmetric. The maximum of accommodation in this depositional unit occurs within the intraclastic grainstones just after the thrombolites. The top of this parasequence set is marked by a flooding surface. In the numerous subsurface section studies there is no indication of a major exposure event between the Upper Dalan (Late Permian) oolitic and bioclastic grainstones of the Kangan (Early Triassic) microbially-influenced sands and microbialites. This contact is conformable.

The coarse lithoclastic grainstones which dominate this cycle (and are characteristic of this stratigraphic interval just after the Permian extinction) are interpreted as a series of intertidal beach ridges, possibly storm generated. The abundant lithoclastic material found around the thrombolitic interval is reworked under progressively storm-wave and fair-weather wave base conditions of microbially-mediated early lithified sediments (such as beach rock and hardgrounds). Very early marine cementation (cement crusts) is a feature of this unit as the importance of the microbial influence, even within the grainstone. The thrombolitic unit developed within the packet of lithoclastic rock is interpreted as shallow subtidal to intertidal microbial/thrombolitic patches in lower energy, more protected conditions (possibly highly nutrient rich). There is an important faunal decrease just below the thrombolitic beds, coupled with the appearance of rare Triassic faunas after, indicating the Permian-Triassic Boundary to be developed around this level (see below). The presence of a thrombolitic boundstone is a regional phenomenon stretching from Ghawar field to the Zagros.

KS2c: This parasequence set is dominated by grainy F16 cycles, 1–2 m thick, and deposited as sandy and intraclastic sheets interspersed by thin 10–30 cm oolitic units, or muddier units (F11L, F5, F5h). The F16 beds are stacked and have an aggradational stacking character. The cycles tend to start with the F16 facies which dominates the cycle, and then are capped by a thin muddier microbially bound cap (10–30 cm thick). The cycles at the top of the parasequence set are more oolite-dominated. The top is marked by a well-developed F16 unit, with well-developed cement crusts, hardgrounds and reworked clasts, which is in turn capped by a microbial mudstone. The maximum of accommodation in this parasequence set is difficult to place due to the aggradational character of the parasequence set – it has been placed within the thickest cycle of F16. The depositional environments are the same as the microbially influenced intraclastic sands of the KS2b.

KS2d: The transgressive facies of this unit are similar to those found in KS2c. In the regressive part of the unit the thickness of the erosively-based grainstone beds decreases upwards, the sands become finer grained, and there is an increasing occurrence of laminated mudstone (F6), anhydrite nodules, and small brecciated levels. The last cycle is completely dominated by these muddy facies, and the top of the parasequence set is marked by an exposure surface and breccia. In terms of depositional environment the lower part of this unit has similar environments to those of the KS2b. Towards the top of this unit intertidal sand flats, mixed flats, microbial mud flats, beach gravels with cement caps, and hypersaline lagoons develop, signifying the return to more shallow internal restricted low-energy conditions. At the top, mud and anhydritic units are interpreted as intertidal mud and microbial flats and sabkha deposits in the shallowest and most exposed part of the cycle.

Large-scale depositional cycles (Figures 9, 17 and 19)

The overall KS2 depositional sequence corresponds to a fourth-order transgressive-regressive couplet displaying an asymmetrical pattern, the transgressive half-cycle being more important than the regressive one (Figure 19).

The KS3b4 surface (top KS3b) is interpreted as a significant flooding surface marking the end of a series of intertidal to supratidal mudflat and lagoonal units (KS3b fourth-order sequence). The maximum zone of accommodation of the fourth-order KS2 sequence is placed towards the top of the grainy oolitic and lithoclastic-dominated cycles of the KS2c. The thick F16 - oolitic sequence aggraded to keep-up with relative sea-level rise and increasing accommodation. This places the stacked F16 sand-sheet facies, thrombolites, and intertidal-shallow subtidal sands of KS2a and KS2b into the early part of the fourth-order transgressive half-cycle. This also corresponds to where fauna and flora mark a high diversity and abundance before progressively disappearing just before the thrombolites (biostratigraphically the Permian-Triassic Boundary). The top of this KS2 fourth-order cycle is dominated by dolomitic intertidal mudflats, shallow coastal salinas and sabkhas.

In terms of the third-order cyclicity the KS2 fourth-order cycle is coupled with the KS3b fourth-order cycle (Figure 17). The surface at the top of KS3a4 (significant dolobreccia unit) is seen as the start of a long-term flooding and relative transgression. This results in the thick unit of stacked intertidal deposits of the KS3b, and the flooding continues until the maximum accommodation zone of the KS2c parasequence set (deepest water, thickest cycles, grainiest highest energy open-water facies, and the most and calcareous units). The top of the KS2d parasequence set is seen as the top of this third-order cycle and dominated by intertidal and supratidal facies.

Stratigraphic, facies, lithological and palaeoecological analysis suggests that there is no major stratigraphic gap between the K2 and the K3 reservoirs, and hence no major stratigraphic gap between the Permian and Triassic periods. The transgression onto the platform starts in the top of KS3b where increasing high energy grainy facies occur, suggesting the platform top was becoming more open before the Permian-Triassic Boundary. This is supported by the fact that the dolostone to limestone transition (from dolomite-dominated platform restricted settings to more open limestone facies) also occurs before the Permian-Triassic Boundary. In the numerous subsurface section studies there is no indication of a major exposure event between the Permian Dalan oolitic and bioclastic grainstones and the Triassic Kangan microbially-influenced sands and microbialites; this contact is conformable. Moreover, there is no major unconformity or stratigraphic surface associated with the Permian-Triassic Boundary and the extinction of Permian fauna occurs within a grainstone body. The analysis suggests there is in fact a low-order transgression between upper K3 and the K2, and hence also between the end Permian and the Early Triassic. This Cycle II with a low-order transgression across the Permian-Triassic Boundary would correspond to the Khuff B Member in the Saudi Arabian subsurface (Al-Jallal, 1995) which also forms a cycle from the upper Permian to the upper lower Triassic (also see Strohmenger et al., 2002).

Cycle I (KS1): Kangan Formation (Figure 20)

The KS1 composite depositional sequence has been subdivided into three major fourth-order depositional sequences: KS1a, KS1b and KS1c, with the respective depositional units: KS1a0, KS1a1, KS1a2, KS1a3 for KS1a, and KS1b1, KS1b2 and KS1b3 for KS1b, and KS1c1, KS1c2, KS1c3 for KS1c.

Depositional Units in KS1a (Cycle I-a)

Initially this unit is dominated by alternating laminated dolostones (F6) and anhydrite layers (F2 and F1), interrupted by sharp-based well-sorted fining-upward medium-grained lithoclastic or peloidal grainstones-packstones intervals. Locally there are well-developed tidally cross-bedded grainstones. The dolomitic facies associated with anhydrite displays mud cracks, microbial laminations and fenestral structures, while the anhydrite exhibits nodular and laminated textures. Much of the anhydrite is early replacement in origin (replacing mud-grainstones) suggesting local anhydrite remobilisation. Green shales to laminated dolomudstone-wackestone are also present. The cycles are dominated by mudstones and anhydrites and are thin (1 m thick), and often anhydrite capped. The cycles then become progressively less anhydritic and better developed tidal oolite shoals appear. This evolves to a series of oolitic shoal cycles intercalated with bioclastic and peloidal packstones. The cycles are well-developed, thicker (up to 4 m) and display both fining- and coarsening-upward evolution. The maximum accommodation zone is placed at the top of a well-developed oolitic tidal shoal unit or/and locally deeper lagoon facies. The sequence is capped by anhydrite beds and is marked by a strong erosional surface.

Depositional environment: Upper intertidal to supratidal features are dominant, with evidence of microbial mats, upper intertidal muds, saline flats (including small temporary coastal salinas) and supratidal sabkhas. The peloidal and oolitic dolograinstones interrupting this peritidal sedimentation correspond to small transgressive flooding events and the installation of small tidal shoals and channels. Small meter-scale cycles reflect rapid fluctuations between cross-bedded dolograinstone ooid bars of subtidal origin (with frequent basal storm deposits) to laminated dolomudstones and massive anhydrite from upper intertidal regimes. The ooid bars may be strongly influenced by tidal currents, as suggested by bi-directional cross-sets. Some of the erosively based cycles may represent tidal channels. The system becomes progressively more energetic and open.

Depositional Units in KS1b (Cycle I-b)

This parasequence set starts with a series of oolitic grainstones with well-developed tidal cross-stratification. This is followed by a series of thinly intercalated (decimetric) bioclastic and peloidal grainstones and conglomerate beds arranged in fining-upwards cycles. Also present are a number of muddier levels with a microbial influence (microbial fabrics). These microbial fabrics (F21) are particularly well-developed in the middle of KS1b (microclotted fabrics, blacked coated grains, organic laminations) and are intercalated with thin (centimeter) peloidal layers. These microbial facies are associated with the same microfauna of microgastropods and small thin-shelled bivalves seen within the classic post-Permian thrombolites. The cycles (1–2 m thick) are fining-upwards with slightly erosive bases. Locally large thrombolitic pinnacle build-ups develop at this level. The parasequences are thicker at the base (up to 3 m) and progressively become thinner and more mud prone. The upper part of the unit is dominated by grainier facies (pack and grainstones) and are a series of stacked bioclastic and peloidal beds (with some more oolitic levels) about 1 m thick.

Depositional environment: The basal grainy beds are seen as small transgressive tidal oolite shoals (shallow subtidal to intertidal). The thinner oolitic and peloidal grainstone beds intercalated with muds are seen as mixed flat deposit, intertidal sands, sandy shallow lagoons and washovers. The small oolitic grainstones within the muddy units are seen as small isolated sandy bedforms (shallow subtidal to intertidal oolitic tidal shoals) and intertidal sands that occasionally migrate into more protected areas (mudstones). The microbial fabrics are interpreted as shallow subtidal and intertidal microbial carbonates. The mud-dominated part of the unit represents a packet of shallow subtidal microbial bioturbated mudstones, possibly under the influence of poor oxygenation (dark colour, preservation of organic matter). These poorly oxygenated microbialites also shallow into intertidal settings and create microbial mixed flats (with grainier washovers and stringers). The upper sandier part represents a series of oolitic/bioclastic and peloidal shoals and channels, intertidal sands and sandy lagoons. The cycles are capped by more muddy mixed flats and microbially bound mixed flats.

Depositional Units in KS1c (Cycle I-c)

The base of this parasequence set is dominated by oolitic and peloidal grainstone-packstones with well-developed tidal cross-stratification. Gradually muddy facies (F5, F5h and F6) become increasingly intercalated with thin isolated peloidal, bioclastic and oolitic beds. The upper part of this unit is dominated by medium to fine peloidal packstones, and muddier levels (F5, F6). Towards the top there is the re-appearance of the green shale facies (F4). Interbedded with these facies are occasional coarse erosively-based grainy beds (peloidal and lithoclastic) which then fine upwards into fine sands and then laminated mudstones. Locally more tidal bioclastic/oolitic sand can develop. The Upper KS1c is capped with intercalated thin laminated muds, anhydrite beds and green shales. The top KS1c is marked by a very thick shale bed (1–2 m).

Depositional environment: This unit represents a shallowing of the depositional system with oolitic tidal channels, scours and transgressive lags isolated at the base. The upper part of this unit which is dominated by laminated muds is interpreted as intertidal mudflats and shallow lagoons and sandy lagoons. The muddier units at the top of cycles are microbial mud flat facies. The depositional environment is dominantly intertidal, as suggested from the fenestrate fabrics and mud cracks found within muddy sediments (microbial mud and mixed flats), and the reappearance of anhydritised flats (F2). Coarse-grained, shallow-water carbonates mark minor transgressive events, tidal channels, evolving rapidly into pelletal tidal-flat subfacies with numerous storm deposits (graded beds, sharp contacts). The green shales represent very internal restricted marginal lagoons with continental influence (estuarine). Also present are microbially bound fabrics. The thick shaly unit at the top of the unit often has a very distinctive change in colour from greenish grey to more reddish then back to greenish grey. The reddish tinge to the shales reflects the most continental influence of the sequence (and hence reflects the cycle boundary) and this change in colour is correlatable between the numerous wells.

Large-scale depositional cycles (Figure 21)

The whole KS1 sequence has been subdivided into three fourth-order transgressive-regressive sequences, namely the KS1a, KS1b and KS1c depositional sequences. These fourth-order depositional sequences are incorporated into a third-order depositional sequence (Figures 9 and 21). The facies organisation follows the same pattern as previous large-scale cycles. The maximum accommodation zones of each of the third-order cycles are located in the best developed cycles (KS1a3 for KS1a, KS1b3 for KS1b and KS1c2 for KS1c2). The maximum accommodation zone for the entire KS1 third-order sequence is located in the thick KS1b3. This also corresponds to the thickest zone of limestone development in the KS1, with the highest energy grainy tidal facies and locally most deeper microbially influenced lagoonal facies.

Summary

The large-scale (third-order) stratigraphic architecture is broadly (Figures 9 and 22):

Cycle IV: KS4 (Upper Dalan Member) - this third-order cycle is composed of three fourth-order cycles (KS4a–KS4c) and is essentially transgressive in nature, with a zone of maximum accommodation in the lower half of KS4c. This is also the zone of maximum aggradation, shoal and porosity development. KS4a is dominated by shallow-water restricted facies (lagoons and sabkhas) and then shallow-water oolitic tidal shoals. KS4b equates to small subtidal bioclastic and oolitic shoals organised in deepening-up bioclastic grain-capped cycles with local lagoons. KS4c is composed of larger bioclastic and oolitic transgressive sandwave complexes. The base of KS4c is marked by a ‘flooding/opening’ event that can be correlated across the region. The top is capped by internal platform sediments and a dolobreccia unconformity.

Cycle III: KS3a (Upper Dalan Member) - this is a third-order cycle with a zone of maximum accommodation in KS3a3. It is a system composed of restrictive coastal lagoons and muddy embayments with limited reservoir development. The fourth-order cycles are not evident and have yet to be correlated, it is possible that they are either highly condensed or reduced/lost during the trangressive onlap.

Cycle II: KS3b (Upper Dalan Member) - this is a fourth-order cycle within a third-order transgressive half-cycle. It is consequently aggradational with a thick sequence of stacked peritidal sediments, sabkhas and vertically burrowed/rooted beds. Reservoir development is poor. The KS2 (Upper Dalan Member and Kangan Formation) is a fourth-order cycle within the third-order maximum accommodation zone and regressive half cycle. It is composed of grainy aggradational facies and local mudstones. The cycle has good reservoir development around the maximum accommodation zone. The facies include microbial intertidal sand flats and shallow-water thrombolitic facies, followed by storm generated pebble grainstone beds and shoals, and microbially cemented grainstones. The Permian-Triassic and corresponding Dalan-Kangan boundary is found within the lower KS2.

Cycle I: KS1 (Kangan Formation) - This third-order cycle is composed of three fourth-order cycles (KSla–c). The zone of maximum accommodation is in the KS1b. This is again the zone of maximum aggradation, and where the best developed tidal oolitic shoals and lagoonal microbial facies are located. This third-order cycle appears to have a series of transgressive pulses that generate accommodation that is filled by grainstone facies. The facies are dominated by peritidal and evaporitic supratidal flats with tidal channels and local microbial facies.

Kuh-E Surmeh Outcrop

The Kuh-e Surmeh outcrop section and sedimentary succession are seen in Figure 23. The main sedimentary facies seen in this section are illustrated in Figures 24 and 25.

Cycle IV (KS4): Upper Dalan Member (Figure 26)

Nar Member equivalent and Cycle IV-a (KS4a)

The Upper Dalan series is about 200 m thick, and located immediately above a tight, and locally brecciated, gypsum member, which is equivalent to the anhydrite of the Nar Member. The base of the IV-a cycle is dominated by dolomudstones with rare interbedded peloidal and lithoclastic dolopackstone containing thin shell fragments. The small-scale depositional cycles are typically fining and thickening-upward, and meter-scale in thickness. The slightly grainier cycles are composed of bioclastic and oolitic pack-grainstones in the lower part with well-developed mudstone caps (F5/6) (fining-up mud-capped cycles). These facies are interpreted as shallow lagoons for the more massive mudstones, whilst the fenestral and mud cracked mudstone are interpreted as mud-dominated tidal flats, with abundant reworking of desiccated muds towards the base of the series (upper intertidal to supratidal zone). The thin grainier intervals represent storms washovers and channels, and locally small isolated peloidal tidal shoals and sandflats (shallow intertidal to intertidal).

The middle part of this unit is dominated by peloidal and oolitic dolograinstones. Peloids and lithoclasts are dominant in the lower part of this unit, where they are commonly mixed with ooid material and contain foraminifera (F8/7). They are fine- to medium-grained, generally well- or bimodally sorted. Oolitic pack to grainstones are more dominant in the upper part of the unit, where they develop as well-defined small-scale (0.5–1.5 m) stacked fining-upward depositional cycles. The cycles are erosively based, marked by coarser micritic lithoclastic material, grading into cross-bedded, occasionally bi-directional, sands. The tops of the cycles show more planar beds and laminations; mud-caps are less well-developed or absent. Within this cycle there in a trend of increasing cycle thickness, better developed cross-bedding and planar-bedding within oolitic dolograinstone and less well-developed mud-caps. This unit represents the maximum accommodation zone of the cycle. Progressively towards the top of this unit muddy facies reappear and muds with mud cracks and teepee structures become increasingly frequent, and intercalated with peloidal sands which are common at the base. The lower part of this unit is interpreted as leeward shoals in close association to the lagoon, with mixed oolitic and peloidal components, whereas the upper part may represent the stacking of tidal inlets or channels, as suggested by common bimodal stratification and irregular erosive base.

The top of the IV-a cycle is dominated by peloidal sediments, with admixed lithoclasts towards the base, and more muddy towards the top. The lower part of this unit presents small-scale coarsening-upward cycles with numerous reworked pebbles towards the base and thin oolitic events, followed by peloidal-dominated bioturbated cycles capped by a thick dolomudstone interval about 2 m thick. The upper part contains abundant dolomicrite within poorly individualised small-scale sequences (F5 and F6), suggesting stacked lagoonal and mudflat-dominated cycles. Cracks and teepees are common, especially towards the top of this unit. They are well-developed just beneath the IV-a cycle boundary. These facies represent platform interior mud-dominated systems of tidal flats, lagoons, and occasional grainy events (storm channels and washovers, thin basal transgressive lags of small-scale cycles).

Cycle IV-b (KS4b)

This cycle starts with a lower unit of coarsening-upward oolitic grainstones with pebble-bearing bases, which progressively become more bioclastic and peloidal with pronounced cross-stratifications (F10/8), and then mud-supported towards the top of the unit (F11). The small-scale cycles average 1 m in thickness, and show great variations in grain size, being mostly finely-grained at the base and coarsegrained at the top with large pebbles. Most cycles present cross-stratifications, and erosively based. Major components are peloidal grains, together with variable amounts of shell fragments and strongly micritised ooids. Bioclastic grains include coarse green algae debris at the base, progressively replaced by echinoid fragments, sponge spicules and fusulinids towards the muddy and more bioclastic-rich upper half of these series.

The upper part starts with a bioclastic-dominated interval characterised by medium to fine debris of green algae, bivalve fragments, gastropods, echinoid plates and abundant benthic foraminifera (F8).

There are also coarse bioclastic grainstones with early marine cementation corresponding to facies of KS4b3 bioclastic caps in the offshore Fars subsurface. The small-scale cycles are mainly fining-upward (erosional contact overlain by cross-bedded bioclastic grainstone). This bioclastic-dominated interval grades into a muddy layer dominated by dark bioturbated facies (F11).

As observed in the previous sequence, there is an upward increase in palaeobathymetry and stacking pattern – the IV-b cycle is dominated by peloidal/oolitic facies with abundant cross-bedding towards the base, evolving to more open-marine conditions towards the top with the occurrence of echinoid remains and mid-ramp foraminifera. The platform flooding is manifested by the oolitic/bioclastic facies in the lower part of this unit, evolving to more open-marine conditions towards the top with the occurrence of echinoid remains and muddy bioturbated deposits. Depositional textures are dominated by mudstone at the base, near the maximum flooding, grading into grainstone toward the upper part of this section and finally muddy at the top.

Cycle IV-c (KS4c)

From bottom to top, four sedimentological intervals can be recognised:

  • (1) A tight mudstone-wackestone interval characterised by strong bioturbation (F11), with sponge spicules, rare benthic foraminifera within an organic lime-mudstone matrix. Locally microbial mounds are present. This interval is interpreted as representing the local maximum palaeobathymetry at this point (cf. ‘MFS’) for the IV cycle.

  • (2) A moldic oolitic-dominated interval (F9) interspersed with peloidal, lithoclastic and muddy intervals (F10/7, F11). The small-scale sequences display fining-upward and coarsening-upward trends organised in cyclic units. Numerous erosional surfaces and hardgrounds have been observed throughout this interval. Another surface with abundant sponge remains within a muddy interval (F11) has been observed at 87 m and interpreted as a minor flooding surface.

  • (3) A mixed oolitic, bioclastic and peloidal interval organised into stacked cross-bedded to planar-bedded small-scale cycles (F8/F10). The bioclastic material is dominated by green algae, echinoid fragments, ostracods and relatively common benthic foraminifera.

  • (4) The top part of this cycle grades progressively upward into mud-dominated facies in the form of cryptocrystalline dolomites (F5/F11). This dolomitic interval represents more internal facies (F5) and underlies the cycle boundary of the overall Cycle IV.

Similar to the underlying interval, the thick oolitic unit represents stacked high-energy shoals and sandwaves, generally formed under subtidal conditions that might undergo occasional exposure. These thick shoals are developed under aggrading conditions, in equilibrium with relative sea-level rise, and are progressively capped by lagoonal peloidal sediments that are affected by early dolomitisation processes at the top of the sequence. Hence, as seen in the equivalent cycle in the offshore Fars subsurface, accommodation is still being created for shoal development after the so called “maximum flooding surface” represented by the deepest water facies before the stacked oolitic and bioclastic shoals.

Cycle III (KS3a): Upper Dalan Member (Figure 27)

This limey unit is dominated by an alternation of bioturbated sediments of mudstone to wackestone type (F11) and occasional grain-supported bioclastic sediment (F8); the cycle as a whole is however mud-dominated (F11). The F11 facies is characterised by numerous Zoophycos traces, whereas the bioclastic packstone (F8) is characterised by crinoids debris. It is difficult to distinguish transgressive or regressive pattern in the interval. Nevertheless, the beds are thin (generally around 10 cm), have a wavy aspect, and can be intensely bioturbated. Hardgrounds, burrows and bioturbated surfaces are abundant throughout the interval. These sedimentological observations characterise open-marine conditions associated with a decrease in carbonate productivity. These facies have been interpreted to form in a seaward shoal to middle-shelf environment that occurred during a major transgressive event. This cycle is more clearly defined on the palaeoecological data (see below) and by the lateral correlation with the subsurface.

Lower Cycle II (KS3b): Upper Dalan Member (Figure 28)

This rather homogenous cycle is composed of an alternation of muddy bioturbated sediments (mud-wackestone; F11) and occasional grain-supported bioclastic/crinoidal beds (F8); the cycle is similar to cycle III but more grain-dominated. These facies have been interpreted to form in a middle-shelf environment that occurred during a major transgressive event, perhaps under slightly reduced circulation (and hence oxygenation) conditions. This unit (as with the offshore Fars subsurface equivalent) is highly aggradational.

Upper Cycle II (KS2): Upper Dalan Member and Kangan Formation (Figure 28)

This cycle starts with a series of stacked coarse-grained bioclastic grainstones. The Dalan-Kangan limit is marked by an erosionally-based storm pebble lag (at 165 m), which is followed by a succession characterised by microbial facies developed as dendritic to hemispherical domes with thrombolitic fabrics. Kashfi (1992) documented a well-exposed unconformity surface marked by a dolomitic limestone conglomerate bed with quartz clasts, and with a 10 degree angular unconformity between the Permian and Triassic at Kuh-e Surmeh. This was not seen in the present study and at the section studied there is no indication of a major exposure event in this locality and the transition from the Permian Dalan bioclastic grainstones to the Triassic Kangan microbialites is conformable. The unconformity cited by Kashfi (1992) may represent the unconformity between the base of the Dalan and the Ordovician cited by Szabo and Keradpir (1978), which is quoted by these authors with identical characteristics (dolomitic limestone conglomerate bed with quartz clasts, and with a 10 degree angular unconformity at Kuh-e Surmeh). Moreover the unconformity at the Permian-Triassic Boundary suggested by Szabo and Keradpir (1978) are highly interpretative based on thickness changes, change in physical appearance between the Dalan and the Kangan, and the absence of index fossils (see below). No exposure or erosional surface has been documented.

The thrombolitic domes are laterally continuous (reaching their maximum around 168 m) and form biostromal units interspersed with more detrital sediments (which themselves have a microbial influence). In general the thrombolite bodies develop on the same sedimentological surface suggesting synchronous seeding and growth. Locally more microbial bioherm-like bodies developed consisting of stacked thrombolite bodies, and exhibiting several meters of relief with regard to adjacent beds. The microfabric of thrombolites consists of centimeter-sized patches or clots of microcrystalline limestone separated by spaces infilled by oolitic/oncolitic sediments. Above the thrombolite level, the microbilaites grade to thin-bedded micro-packstone with shell fragments (Claraia sp.). The rest of this cycle is not accessible and described in the Kuh-e Dena section.

Cycle I (KS1): Kangan Formation (Figure 29)

Cycle I-a and Cycle I-b

These two cycles are poorly differentiated in this location. After 32 m of non-accessible section, the Kangan Formation continues as thin-bedded limestones devoid of bioclasts; the sedimentation is mud-dominated and roughly organised in four thickening-up sequences (two in Cycle I-a and two in Cycle I-b). These sequences are made of smaller-scale cycles defined by three main facies from bottom to top:

F13: highly bioturbated, thin-bedded fine-grained packstone-wackestone, with local slump features;

F14: thin-bedded mudstone with abundant centimetre-scale packstone intercalations; and

F15: argillaceous thin-bedded mudstone with occasional millimeter-scale intervals of fine packstone and abundant vermiculate bioturbations.

These rhythmic sediments reflect anoxia associated with rising sea level within a relatively deep-water outer ramp setting.

Cycle I-c

This cycle shows a clear change from the sedimentation seen in Cycle I-a and Cycle I-b, with sedimentation becoming significantly grain-supported. Initially the cycle is dominated by the aggradation of well-sorted oolitic grainstone, which show their first reappearance since the top of the Permian (Cycle IV). The upper part consists exclusively of oolitic and bioclastic grainstone, rich in intraclasts. These oolitic shoals are interpreted as progradation from the south of tidal shoals and channels (as in KS4c). The top of this unit is dominated by proximal storm deposits intercalated in tidal flat deposit, with the development of local columnar stromatolites. These are followed by the green marls of the Aghar Shale Member (Dashtak Formation).

Kuh-E Dena Outcrop

The Kuh-e Dena outcrop section and sedimentary succession are seen in Figure 30. The main sedimentary facies seen in this section are illustrated in Figures 31 and 32. The faunal and sedimentological details of parts of the section are masked by a strong dolomitisation overprint.

Cycle IV (KS4): Upper Dalan Member (Figure 33)

Nar Member equivalent and Cycle IV-a (KS4a)

The equivalent of the Nar Member is composed of a palaeosoil interval of kaolinite and siliciclastic material. The lithofacies range from yellow/red pisolite to grey homogeneous kaolinite, to black friable lignite-like material. Three possible black material-capped cycles were identified and organised in an overall upward-thinning trend. These are interpreted as stacked palaeosoil levels with alteration of siliciclastic material resulting from subaerial exposure in a warm and humid climate. This is overlain by the basal part of the IV-a cycle which comprises small-scale fining-up cycles with coarse sands rich in lithoclasts (F7) at their base and grading to dolomitised mudstone (tidal flat muds) at the top (F11). This interval records an increase in water depth with relatively open-marine conditions above a tidal flat environment as evidenced by fenestrae features. Lithoclastic content indicates storm deposits and channels lags.

This is followed by a unit dominated by dolomitised oolitic and peloidal grainstone deposits (F9 and F10) with erosive bases, abundant ripples (and HCS in places) and planar laminations. Locally these are mixed with minor bioclastic material (rich in fusulinids). These represent stacked oolitic shoal to leeward shoal bodies deposited in a high energy environment, with interspersed storm deposits. The upper part of this cycle is dominated by stacked bioclastic and oolitic grainstone deposits (F8 and F9). The bioclastic sands contain fusulinids, gastropods, miliolids and notably rare coral fragments. Oolitic grainstones are clean, well-sorted and exhibit cross-bedding and planar laminations. These facies are interpreted as high energy shoals shallowing to intertidal sandflats. This unit represents the maximum accommodation zone of this cycle.

The top of this cycle is capped by brecciated muddy facies (F5). The cycle in generally grainier and less muddy that Surmeh and particularly the offshore Fars subsurface, and is clearly located in a much more energetic system.

Cycle IV-b (KS4b)

This unit starts with lagoonal muds at the base then becomes grain-dominated with bioclastic, peloidal and oolitic grainstone/packstone. Bioclastic content includes fusulinids, gastropods, sponge spicules, possible miliolids and green algae. The central part of this cycle is dominated by oolitic and bioclastic deposits, with abundant fusulinids, rare crinoids, ripples and flaser bedding structures. The upper half shows the alternation of clean oolitic grainstone and sponge spicule wackestone (F11). These reflect deepening-upward cycles, deepening from shoal to outer shelf conditions (comparable with the oolite – bioclast deepening-up cycles seen in this cycle in offshore Fars subsurface). The top of this cycle becomes muddy again with bioclastic facies (F8: fusulinids, other foraminifera) grading to muddier units toward the top (F11: sponge spicules). Here the cycles are capped by fenestrae and, for the topmost high-frequency cycle, by cavities filled by oolitic material from the overlying unit. This cycle is very asymmetric with the maximum accommodation zone located towards the top of the deepening-upward cycles at the top of the unit just before the reappearance of shallow water muds and sands at the top of the cycle.

Cycle IV-c (KS4c)

The lower part of the unit is dominated by bioclastic facies with abundant coral and echinoderm fragments (facies F11C and F8B). Deposition is interpreted to have occurred in an outer shelf environment with an upward decrease in energy which can be associated to a deepening trend. The coral fragments suggest local reworking of small reefal patches. The upper part of this cycle is dominated by cross-bedded and rippled bioclastic facies with abundant echinoderm fragments (facies F11C) and coral fragments are again locally abundant (F8B). The uppermost part corresponds to slightly shallower conditions with probable reworking of reefal patches. Deposition occurred for most of the unit in an outer shelf environment and reflects the highest energy and most ‘open’ facies associations of the cycle IV, and hence reflects the maximum of back-stepping of the platform shelf system edge (equivalent to the maximum accommodation zone; intra KS4c2). The top of this cycle remains relatively grainy (with stacked erosive surfaces) and internal platform sediments are rare compared to the offshore Fars subsurface (internal platform tidal flats did not prograde out this far).

Cycle III (KS3a): Upper Dalan Member (Figure 34)

This is a homogenous unit dominated by layers of echinoderm-rich facies (F11C) with some alternations of fusulinid and oolitic facies (F8 and F9). Bioturbation is locally well-developed (Zoophycos). Deposition that occurred in the outer shelf is intercalated with occasional grainy input reflecting bioclastic shedding. The strong dolomitisation overprint poses a problem with placing the base of this stratigraphic interval with certainty.

Lower Cycle II (KS3b): Upper Dalan Member (Figure 35)

This unit is similar to the previous unit, being dominated by sheets of echinoderm-rich bioclastic packstone. It also shows abundantly thin intervals corresponding to grain-supported input of reworked material. Interbedded are thin muddy levels (F11). The top of the unit is characterised by more frequent rippled and laminated bioclastic packstone and stacked erosive surfaces. Deposition occurred in the outer-shelf area and recorded abundant storm deposits.

Upper Cycle II (KS2): Upper Dalan Member and Kangan Formation (Figure 35)

The base of this cycle is a continuation of the previous facies – dominated by sheets of echinoderm-rich bioclastic packstones. The Dalan-Kangan boundary is marked by the contact between planar stromatolites (F12) (Kangan, Early Triassic) and the underlying bioturbated crinoid packstone (F11C; Upper Dalan, Late Permian). As with the Kuh-e Surmeh section, there is no indication of a major exposure event in this locality and the transition from the Upper Dalan (Late Permian) bioclastic grainstones to the Kangan (Early Triassic) microbialites is conformable. Following the loss of Permian fauna the cycle is dominated by microbial sediments of the Early Triassic Kangan Formation.

Three types of microbial growth-forms have been observed: planar, columnar and dendritic. The planar stromatolites form isopachous beds or fill the ‘troughs’ that develop between the other non-planar microbial structures. The columnar growth-forms form isopachous beds about 30-cm-thick, made of one growth episode. Dendritic growth-forms form either isopachous beds or hemispherical domes, where several growth episodes can be stacked and form larger structures. The thrombolite microfabric consists of a pink-grey framework of microcrystalline limestone clots or interspaced patches filled by lighter-coloured sediment. Locally successions of interbedded planar stromatolites (with teepee structures), thombolitic growth events and thin oolitic sheets (bound by microbialite) can be seen. The top of the thrombolitic unit shows a gradation from microbially bound oolitic thrombolite through to oolitic grainstone (F9); this in turn is finally capped by bioturbated mudstone (F11). The microbial associations with the oolitic grainstone and teepee structures suggest a shallow water system deepening-up into deeper water bioturbated mudstone. The rest of this cycle is characterised by thin-bedded mud-dominated sediments corresponding to facies F11, F13, F14 and F15 (dominated by F13, F14). These are composed of small-scale deepening-upward high-frequency cycles (F11/13, F14, F15). This cyclic sedimentation pattern reflects changes in oxygenation (dysoxia-anoxia) associated with rising sea-level and possible global anoxic events during the Early Triassic.

Cycle I (KS1): Kangan Formation (Figure 36)

Cycle I-a and Cycle I-b

As with Surmeh the Cycle I-a and Cycle I-b are difficult to differentiate in this location. The units are characterised by thin-bedded mud-dominated sediments corresponding to facies F11, F13, F14 and F15, arranged in small-scale deepening-up depositional cycles (F11/13 - F14 - F15) as in upper Cycle II. This cyclic sedimentation pattern reflects changes in oxygenation (dysoxia-anoxia) associated with rising sea-level and possible global anoxic events during the Early Triassic. At a larger scale the facies F15 is dominant at the base of the section, then F14 becomes progressively more abundant, and finally facies F13 becomes more significant. Slump features have been observed in this stratigtraphic interval suggesting outer shelf deeper water conditions.

Cycle I-c

Again this cycle marks a drastic change to a more grain-dominated system. This unit is characterised by fine peloidal packstones (F10) with abundant sedimentary features such as low to high angle ripples, fluid escape features, and gutter-cast. The base of the unit is characterised by coarser reworked oolitic material (F9). These facies are interpreted as peloidal and oolitic intertidal sand-sheets, and small peloidal shoals. These are followed by the Aghar Shale Member (Dashtak Formation) which is dominated by the alternation of green marls and tidal facies with occasional oolitic washover fans.

DEPOSITIONAL MODEL AND FACIES DISTRIBUTIONS

Conceptual Depositional Model

Integrating the facies and sedimentological interpretations outlined above, with the stratigraphic interpretations and framework developed for the study, conceptual depositional models have been constructed for Cycles IV, III, lower II, upper II, and I (Figures 3741). From these interpretations and models it is clear that there have been significant changes in platform type/geometry, facies organisation and climate from Cycles VI through to I. It is also evident that although most of the facies are seen in all the major stratigraphic cycles, their relative proportions and distributions are not the same (volumetric and facies partitioning). The main exception to this is the microbial suite of facies which tends to be restricted to particular intervals within the KS2 (just after the Permian-Triassic extinction) and KS1 (KS1b microbial event), due to particular palaeoceanographic conditions.

Palaeogeographic Setting and Platform Morphology

At a large-scale the Upper Khuff depositional setting during the KS4 and KS3a was organised into a platform profile that gently deepened from the south (southern North Field) with a platform top interior zone (North field), a platform top edge zone (northern South Pars field), an intrashelf low (Nar and Kangan fields) (personal observation; Szabo and Kheradpir, 1978; Konert et al., 2001; also see Kuh-e Siah-1 well in Sharland et al., 2001; Sadooni and Alsharhan, 2004; and Dalan isopach in Szabo and Kheradpir, 1978). The platform then rose again in the north with palaeohighs around Kuh-e Surmeh and Kuh-e Dena (structurally-controlled basement highs; the so-called ‘Zagros High’ of Szabo and Keradpir, 1978). There is however a major change in the platform profile from this KS4-KS3a profile (Permian), to the KS2b-KS1 profile (Triassic). The KS2b-KS1 profile has a monoclinal ramp geometry which opens to the north to deeper-marine conditions with the absence of effective palaeohigh barriers around Kuh-e-Surmeh and Kuh-e-Dena (as determined by the facies data). These two large-scale palaeogeographic profiles controlled the overall distribution of facies belts across the platform.

This change in platform profile coincided with other events within the KS2b (the Permian-Triassic Boundary), including: (1) major facies changes on the platform tops (offshore Fars subsurface) marked by the appearance of thrombolites and associated microbial grainstones; (2) major facies changes in the northern shelf edge areas (northern Zagros) where there was a change from shallow-water high-energy grainy facies to deeper-water mid-ramp muddy facies; (3) change in relative stratigraphic thickness from a northern Zagros area with a thinner Upper Dalan stratigraphic thickness relative to the offshore Fars subsurface (due to the presence of the Zagros palaeohighs), to slightly thicker Kangan succession of open-shelf muddy facies (due to high accommodation in a deeper open-marine setting); and (4) the appearance of high gamma-ray shales in the eastern Zagros subsurface area, not seen before the KS2b, and only seen in this area (see Dalan-1, Dashtak-1, Varavi-1 cited in Kashfi, 1992).

These events are all consistent with a major flooding across the Permian-Triassic Boundary causing: (1) a drowning of palaeohighs; (2) encroachment of anoxic waters into the shelf area lows; (3) terminating bioaccumulations (such small reefal patches) at the shelf edges; (4) flooding the platform tops with more grainy facies (after the muddy lagoonal Upper Dalan), and developing microbial facies across the shelf; and (5) the quasi-synchronous Permian-Triassic mass extinction. At this stage it is unclear whether the high gamma-ray shales of the eastern Zagros subsurface zone reflect the causal factor of the Permian-Triassic mass extinction (encroachment of anoxic waters – anoxic water turn-over) or the effect of the Permian-Triassic mass extinction (accumulation of decaying organic matter in shelf palaeolows). Moreover, as outlined above, the stratigraphic facies and palaeoecological analysis suggests that there is no major stratigraphic gap nor surface between the latest Permian and the Early Triassic. There is no major unconformity or stratigraphic surface associated with the Permian-Triassic Boundary and the extinction of Permian fauna occurs within a grainstone body (see below). The analysis suggests there is in fact a low-order transgression between upper K3 and the K2, and that the Permian-Triassic mass extinction is associated with major global chemical oceanographic changes which occurred during general global sea-level rise (second-order transgression). This is consistent with recent studies which suggest that the end-Permian extinction event is associated with global transgression and the encroachment of anoxic waters (Wignall et al., 2005; Wignall and Twitchett, 2002).

These changes are also coincident with a major shift of the spectral gamma-ray wireline log data at or around the Permian-Trassic Boundary (K3-K2) (see below). At this level there is a significant drop in Uranium across the boundary in both grainstone facies and muddy facies. This suggests that this ‘Uranium event’ is independent of local facies conditions. There is a significant increase in Thorium and Potassium across the boundary in more muddy facies. This Spectral Gamma Ray ‘Uranium event’ can be easily correlated in the wells for which data is available.

Conceptual Depositional Models

Cycle IV (KS4) conceptual depositional model (Figure 37). KS4 was a zone of significant shoal development where, in the open higher-energy platform edge margin zones, large bioclastic and oolitic sandwave complexes developed. These large transgressive sandwave complexes had dominantly southerly sediment-transport direction (and hence bedform migration) as evidenced by sedimentary structures on core and image log data. Towards the top of the platform interior these graded into well-developed oolitic tidal shoals and ebb-flood tidal deltas. These gave way to small shallow subtidal peloidal and oolitic shoals with local lagoons on the leeward side of the main shoal belt. In the most restricted areas the system was dominated by shallow-water to restricted intertidal facies including lagoons, hypersaline lagoon, intertidal mudflats and sabkhas. Geographically, in the North Dome field area the early KS4 (KS4a) was initially dominated by shallow-water restricted facies such as lagoons and sabkhas, followed by small shallow-water oolitic tidal shoals. To the north within the Nar-Kangan intrashelf low area these passed into hypersaline lagoons, and lagoons with minor sands. These facies shallowed into small oolitic and bioclastic shoals with interspersed lagoons around the palaeohigh in the north Zagros (Kuh-e Surmeh and Kuh-e Dena). The middle and upper KS4 (KS4b and KS4c) were dominated by oolitic and oobioclastic transgressive sandwave complexes and shoals with local lagoons in the North Dome field area. These gave way to embayment facies with oolitic and peloidal shoals in the Nar-Kangan intrashelf low area. To the north, on the open palaeohighs, energetic bioclastic sands with reworked reefal debris, and oolitic shoal complexes developed.

Cycle III (KS3a) conceptual depositional model (Figure 38). The KS3a was a system composed of a series of shallow-water restricted muddy evaporitic coastal lagoons and shallow-water poorly oxygenated embayments in the North Dome field area (which are more anhydritic in the south). These passed into more open, less restricted and deeper bioclastic sands interspersed with embayment muds within the Nar-Kangan intrashelf low area. To the north on the Zagros palaeohighs, bioclastic sheets sands and shoals continued to develop and were interspersed with secondary embayment facies.

Lower Cycle II (KS3b) conceptual depositional model (Figure 39). The KS3b was dominated by a series of stacked peritidal muds, sabkhas and rooted/vertically burrowed lagoons and hypersaline lagoons in the North Dome field area. Towards the Nar-Kangan intrashelf low bioclastic and oolitic sands developed, followed by embayment muds and lagoons. In the palaeohigh around Kuh-e Surmeh and Kuh-e Dena bioclastic sheets sands and shoals continued to prevail.

Upper Cycle II (KS2) conceptual depositional model (Figure 40). In the North Dome field area, KS2 is composed of grainy aggradational facies and local mudstones. The facies include intertidal sands and shallow-water thrombolitic facies followed by storm-generated pebble grainstone beds and shoals, which had varying degrees of microbial influence (microbially bound grainstones, microbially cemented grainstones). In the Nar-Kangan fields area, deeper-water embayment muds dominated, followed by the progradation of oolitic grainstones and lagoons that prograded from the south (see below). In the northern Zagros area around Kuh-e Surmeh and Kuh-e Dena, deeper embayment muds (with occasional slump structures) and lagoons developed, and indicate that this area was no longer a palaeohigh but rather in a more basinal setting. Thrombolites became well-developed across the region just after the Permian-Triassic extinction.

Cycle I (KSI1) conceptual depositional model (Figure 41). On the platform top in the North Dome field area, the KS1a–c system was in a low accommodation state and therefore composed of peritidal and evaporitic supratidal flats, with tidal channels and local microbial facies. This was also a zone of small shoal development. Towards the Nar-Kangan fields region more accommodation and higher-energy conditions allowed the development of embayment muds interspersed with well-developed tidal oolitic shoal complexes and peloidal sand sheets. To the northern Zagros area the system is more basinal with deeper embayment and mid-outer ramp muds with slump features. At the top of KS1 there is late stage progradation of grainy facies from the south (see below).

Local and Regional Facies Variations

In general, at the field-scale (tens to many tens of kilometers) there is a great deal of similarity between wells in terms of facies types, facies distributions and sedimentological systems within individual depositional units. This reflects the very large scale of the carbonate platform and its overall flat ramp-like geometry. There are however subtle but significant variations in facies types, lithologies (calcite versus dolomite versus anhydrite) and poroperm zones from one section to the other. At a larger-scale (100s of km) more systematic trends in facies textures (grain- versus mud-supported), and consequently reservoir potential, are developed. These are related to changes in palaeobathymetry, position of intra-platform lows and hydrodynamic energy gradients, and hence the position of the main shoal belt. These local and regional facies variations are summarised below.

Local variations within the North Dome field area (Qatar and Iran)

At the North Dome field scale the lateral variations for a given stratigraphic layer are very slight, and the regional patterns of facies variations are poorly expressed (personal observation; Alsharhan and Nairn, 1994a, 1994b; Sharland et al., 2001; Schlumberger, 1981; Hamam and Nasrulla, 1989). Nevertheless facies variations between North Dome field wells do occur and represent either high-frequency very local autocyclic events, or subtle regional signals in depositional facies trends.

The high-frequency very local events include: (1) local development of more protected (muddy) periods; (2) local variations in the degree of microbial activity; (3) local presence of autocyclic events (storm beds, chanalisation); (4) better developed shoal facies due to local tidal amplification; and (5) subtle variations in diagenetic pathways. These, by definition, are very local and systematic trends between wells (even a few kilometers apart) and not discernable. Nevertheless facies variations caused by these processes can be very marked.

Notwithstanding this limited systematic lateral facies variation at the North Dome field scale, detailed study reveals that there are subtle but systematic regional signals in depositional facies between the south and north parts of the North Dome. Thus even though the general facies and stratigraphic organisation of north and south North Field are identical, the detailed facies character and their proportion in a given interval shows regional trends. This is particularly evident with the internal platform facies (principally located at the base and top of KS4, KS3b, and base and top of KS1) which are better developed in the southerly areas. Here it is seen that within the internal facies there is: (1) more internal mud, (2) less frequent grainy packets, (3) more early to syn-depositional dolomite, (4) more anhydrite; and (5) more green shale (for KS1). These trends are suggestive of a slightly more internal position on the platform profile for the southerly areas, and are consistent with the large scale regional trend of more distal conditions to the north. A similar pattern is seen for the grainy more energetic facies around the maximum accommodation zones of the KS2 and KS4, where the systems are slightly less bioclastic-rich and the large-scale sandwave complexes less strongly developed in the southern part of the North Dome compared to its northern part. Again this would suggest that even during the highstands, the southern North Dome system is very slightly less energetic compared to the northern part. However it must be stressed that these differences in facies character and facies proportion, in both the internal and energetic facies, are very subtle.

Variations between the North Dome field area and Northern Zagros Outcrops

The Zagros outcrop data provide the most northerly control points for the study, and importantly constrain the Kuh-e Dena section, which is cited as an Upper Khuff palaeohigh with reefal or reefal debris close to the margin of the shelf system (Szabo and Kheradpir, 1978; Al-Jallal, 1987, 1994). This study shows that, compared with the North Dome field depositional system to the south, the main differences for the KS4 and KS3 units (Permian Upper Dalan Member) are: (1) absence of an anhydritic Nar Member (replaced by a laterite/bauxite horizon and or breccias); (2) thinner stratigraphic thicknesses over the palaeohighs; and (3) generally rich in oobioclastic shoals and coarse bioclastic debris (with large coral fragments suggesting reworking of reefal patches). This area is consistently grainier (particularly Kuh-e Dena) throughout the KS4 and KS3, even when lower energy conditions prevail in other areas (such as the Zagros subsurface, North Dome field and the United Arab Emirates).

For the KS2 and KS1 units (Early Triassic Kangan Formation) two main differences between the North Dome field depositional system and the Zagros outcrops to the north are evident. Firstly, the Zagros outcrops had a much muddier system during the KS2 and KS1; they were dominated by deeper embayment muds and mid- to outer-ramp muds. The system only became significantly grainy at the top of the KS1 when the sandy and lagoonal platform facies prograded from the south and reached the northern Zagros area (see below). Hence the high-energy bioclastic facies around Kuh-e Surmeh, and particularly Kuh-e Dena, that were present during the KS3 and KS4, no longer developed due to flooding and drowning of the bioclastic system. Secondly, and concurrent with this facies change, was a change in the relative thickness (the KS2 and KS1 are slightly thicker than further south, compared to significantly thinner during the KS3 and KS4). As previously noted these two differences suggest that there was a major change in the platform profile from the Upper Dalan profile - with a Zagros subsurface intrashelf low bounded to the north by palaeohighs, to a monoclinal ramp profile with deeper marine conditions in the north and the absence of any effective palaeohigh barriers.

Zagros Subsurface Area

At a larger scale the Zagros subsurface wells (Nar and Kangan fields) provide a northern control point in the Zagros subsurface area (personal observation; Szabo and Keradpir, 1978; Kashfi, 1992, 2000). The facies seen in the cores are in general very similar to the facies seen in other parts of the Khuff system in the Gulf. In general, what is remarkable, is the general similarity of the depositional system to that of North Dome field considering the great distance between North Dome field and the Zagros (80 km). Clearly these fields are on the same large-scale shelf system. Nevertheless, compared to the North Dome field area, the Nar-Kangan fields area has some important differences, summarised as follows:

  • KS4: (1) less well-developed top KS4 third-order sequence boundary; (2) less anhydrite-bearing facies; (3) less tidal flat facies at the top of KS4; (4) significantly less dolomite (more limy); and (5) only a minor poroperm break between KS4 and KS3.

  • KS3: (1) good poroperm at base of KS3b; (2) limestone dominated sequence, especially in KS3b; (3) stacked aggradational shallow subtidal poorly oxygenated embayment muddy facies, intercalated with bioclastic grainstones in KS3b; and (4) much less anhydrite.

  • KS2: (1) muddy dominated system at base – dark embayment muds; (2) much less anhydrite; and (3) good poroperm at the top of KS2 as the grainy system progrades from the south (South Pars field) to the north.

  • KS1: (1) much more radioactive at bottom and top; (2) more deeper embayment mud facies; (3) more limestone dominated; (4) very well-developed oolitic shoal complex during later progradational phases; and (5) less anhydritic.

Hence overall, compared to the North Dome field area, the Nar-Kangan fields area has: (1) less abundant anhydrite throughout the KS4–KS1; (2) reduced dolostone to limestone proportion - generally more limey; (3) less marked exposure surfaces; and (4) less well-developed muddy flat deposits. These are consistent with a more ‘open’ (less restricted/evaporitic) external system and a more open ‘basinwards’ position on the platform system – especially at top KS4, KS3 and KS2. Moreover, more poorly oxygenated organic-rich embayment facies are seen, particularly in the KS2, but also in parts of the KS3 and KS1.

These observations, together with the regional variations in formation thickness, suggest that during the KS4 and KS3 the Nar-Kangan fields area was an intrashelf low (‘Nar-Kangan intrashelf low’) and developed slightly more ‘distal’ facies compared to the South Pars field area. This Nar-Kangan intrashelf low was bounded to the south by the North Dome field platform high and to the north by the palaeohighs around Kuh-e Surmeh and Kuh-e Dena (Szabo and Kheradpir, 1978; Partoazar, 1995). During flooding periods (early TST) these were often restricted low-energy embayment muds - compared to more platform top facies on North Dome. During late highstands shallow-water facies prograded from South Pars field area and appeared in this region (see below). However during the KS2 and KS1 the Nar-Kangan fields region was no longer an intrashelf low since the palaeohighs to the north (Kuh-e Surmeh and Kuh-e Dena) were no longer topographic highs. During these times the system was a transitional zone from the platform top facies of the North Dome field to the deeper distal facies of the northern Zagros area. The facies were often embayment muds at the base of the large-scale cycles then shallowed-up into more typical platform top facies as the platform prograded northwards from South Pars.

United Arab Emirates

To the southeast of the North Dome field area, the Upper Khuff of Abu Dhabi subsurface has identical facies types and organisation to those described in this paper (personal observation; El-Bishlawy, 1985; Alsharhan and Kendall, 1986; Alsharhan and Nairn, 1994a, 1997; Alsharhan, 1993). The main differences compared to the North Dome field area are as follows:

  • (1) The large bioclastic and oolitic sandwaves and shoals of the KS4 of the North Dome field are less developed in the Abu Dhabi area, and the system appears to be more protected, but still sandy, and generally lower energy. Moreover they tend to be more muddy and anhydritic in the lower KS4. Likewise the KS4c1 flooding event, though clearly developed, is less marked (see palaeoecological analysis).

  • (2) Compared to the North Dome field (particularly South Pars field), the KS3 depositional system is lower energy and more restricted in nature. This is particularly the case within the KS3b where the muddy zones appear to be more ‘proximal’ compared to the North Dome field with more shaly material being present. This suggests that during the KS3b deposition this area had a more ‘proximal/continental’ influence compared to the North Dome field area.

  • (3) Compared to the North Dome field area, the KS2 thrombolite in the Abu Dhabi subsurface area is very poorly developed (less than a foot in thickness compared to 3 m in South Pars). The associated microbially early-cemented lithoclastic grainstone-packstones are however present and well-developed. Moreover the pre-K2 thrombolite system is more muddy and anhydritic; the top K3 reservoir opening of the platform in North Dome field (which creates the higher energy conditions and hence grainstone reservoirs) is much less evident. This suggests a more internal position compared to the North Dome field area. This is also consistent with the changes in lithology at the top of K3 from dolostone to limestone (coincident with the opening-up of the platform) which occurred much later in the Abu Dhabi area – just before the thrombolites (this also suggests a more internal system). From the biostratigraphic and palaeoecological data there does not appear to be a major flooding event associated with stratigraphic surface KS2a (i.e. top of K3 reservoir; see below). The fauna remains restricted until the Permian-Triassic extinction. This again suggests that the system was generally more ‘restricted’ and the flooding/transgressive surface associated with surface KS2a is not as marked in this area (though still definitely present).

  • (4) In the KS1, the zone of microbial facies seen in K1 appears to be correlatable to the KS1b microbial zone in the North Dome field area, and provides further evidence of the correlative value of this suite of facies which is related to a K1 microbial event. The K1 appears to be significantly more microbial than in the South Pars field area. More grainy and oolitic facies with well-developed oolitic tidal shoal complexes (which are more reservoir prone) developed in response to very restricted internal conditions in the North Dome field which limits shoal and reservoir development there. This suggests that in the Abu Dhabi area there was slightly more accommodation space, and slightly higher energy conditions. There is also more proximal shaly influence in the K1 reservoir of Abu Dhabi wells compared to the North Dome field.

Summary

Five major facies zones are recognised and characterised: (1) North Dome field area, (2) United Arab Emirates, (3) Zagros subsurface area, (4) northern Zagros outcrop area, and (5) eastern Zagros subsurface area. The characterisation of the main sedimentological and depositional systems of the KS4 to KS1 units is broadly as follows:

North Dome Field Area

KS4a: Dominated by shallow-water restricted facies (lagoons and sabkhas) and then small shallow-water oolitic tidal shoals.

KS4b: Subtidal bioclastic and oolitic shoals with local lagoons.

KS4c: Larger bioclastic and oolitic transgressive sandwave complexes.

KS3a: Shallow-water muddy lagoonal and embayment conditions.

KS3b: Stacked peritidal muds, sabkhas and rooted beds.

KS2a–d: Intertidal sands and shallow-water thrombolitic facies followed by storm-generated pebble grainstone beds and shoals.

KS1a–c: Peritidal and evaporitic supratidal flats with tidal channels and local microbial facies.

United Arab Emirates

KS4a: Dominated by shallow-water restricted facies (lagoons and sabkhas) with minor shoal development.

KS4b: Peloidal sand shoals-sheets, and lagoons; poorly developed bioclastic and oolitic shoals.

KS4c: Poorly developed subtidal bioclastic and oolitic shoals, peloidal sand sheets.

KS3a: Shallow-water muddy lagoonal and embayment conditions with grainy peloidal and bioclastic sands.

KS3b: Stacked shaly peritidal muds, sabkhas and vertically/rooted beds.

KS2a–d: Muddy lagoons followed by intertidal sands and shallow-water thrombolitic facies followed by storm-generated pebble grainstone beds and shoals. Poorly developed thrombolites.

KS1a–c: Shaly peritidal and evaporitic supratidal flats with tidal channels and local microbial facies. Local well-developed oolitic tidal complexes.

Zagros Subsurface Area

KS4a: Hypersaline lagoons, and lagoons with minor sands.

KS4b: Embayment facies.

KS4c: Oolitic and peloidal shoals.

KS3a: Bioclastic sands interspersed with embayment muds.

KS3b: Bioclastic and oolitic sands followed by embayment muds and lagoons.

KS2a–d: Embayment muds followed by oolitic grainstones and lagoons that have prograded from the south. Thrombolites well-developed.

KS1a–c: Embayment muds interspersed by well-developed tidal oolitic shoal complexes and peloidal sand sheets.

Northern Zagros Outcrop Area

KS4a: Small oolitic and bioclastic shoals with interspersed lagoons. KS4b: Open bioclastic and oolitic shoal complexes.

KS4c: Well-developed oolitic and bioclastic shoals with the reworking of reefal debris, and intercalated deeper embayment muds.

KS3a: Bioclastic sheets sands and shoals interspersed with secondary embayment facies. KS3b: Bioclastic sheets sands and shoals.

KS2a–d: Deeper embayment muds and lagoons. Thrombolites well-developed.

KS1a–c: Deeper embayment muds and mid-outer ramp muds with slump features. At the top of KS1 there is late stage progradation of grainy facies from the south.

Eastern Zagros Subsurface Area

KS4a: No core control data available.

KS4b: No core control data available.

KS4c: No core control data available.

KS3a: No core control data available.

KS3b: Lagoonal muds and tidal flats.

KS2a: No core control data available.

KS2b-d: No core data control available - wireline logs suggest the appearance of highly radioactive muds.

KS1a-c: No core control data available - wireline logs suggest highly radioactive muds dominate this level.

From this analysis depositional models have been constructed which account for the local and regional facies variation, and which suggest that there have been significant changes in platform type/geometry, facies organisation and climate from Cycle IV through to Cycle I. The most important of these changes is the large-scale change in platform morphology between the Upper Dalan Member and Kangan Formation where the setting changed from a platform with intrashelf lows and palaeohighs to the north, to a gently sloping homoclinal ramp opening to the north. This change in palaeogeographic profile had a major control on the overall distribution of facies belts across the platform.

BIOSTRATIGRAPHIC AND PALAEOECOLOGICAL ANALYSIS

The aim of this analysis was to: (1) better characterise the palaeoenvironmental setting by detailed palaeoecological analysis; and (2) further constrain the stratigraphic interpretations based on the sedimentological data by biostratigraphic and ecostratigraphic analysis. Due to the major biological crisis at the Permian-Triassic Boundary, the biostratigraphical and palaeoecological analysis mainly concentrates on the Permian Upper Dalan Member (Cycle IV to lower Cycle II). The data used for this study includes the Kuh-e Surmeh outcrop and selected subsurface data. The Kuh-e Dena outcrop was not studied in detail due to strong secondary dolomitisation. The results also integrated Lower Dalan data from the Kuh-e Gakhum outcrop reference section. Table 2 (also see Plates 1 and 2) lists the macrofauna, microfauna and microflora of biostratigraphic and/or palaeoecological interest selected from the outcrop and subsurface Upper Dalan Member.

Biostratigraphy and Dating

This study uses the standard scale, which is based on Chinese and American stages names (Jin et al., 1997), rather than the Tethyan scale, still used by some authors, but not fully correlated with the standard scale (Taraz, 1973; Sharland et al., 2001; Vachard et al., 2002). In this scheme the Upper Dalan corresponds to the Lopingian (Late Permian) Epoch between the post-Capitanian extinction where the giant fusulines disappeared (Sheng, 1992), and the main extinction at the Permian-Triassic Boundary (Jin et al., 2000, 2003). The Lopingian Epoch is divided into the Wuchiapingian Stage (equivalent to the Tethyan Dzhulfian stage), and Changhsingian Stage (here considered as equivalent to the Tethyan Dorashamian stage).

Johnson (1981) proposed a Midian to Dzhulfian p.p. (Capitanian to Wuchiapingian) age for the Kuh-e Surmeh Upper Permian series, and suggested the Permian series was not complete at its top, with a stratigraphic gap of the upper Dzhulfian (note that Dzhulfian could have been considered as complete Lopingian at that time) below the Lower Triassic deposits. In this study, based on the stratigraphic distributions of the biostratigraphically significant fauna and flora listed in Table 2 (figured in Plates 1 and 2), the following age determinations are proposed.

Lower Dalan Member: Permian Period, Guadalupian (Middle) Epoch, Wordian and Capitanian Stages

The study of several samples from subsurface wells has confirmed the presence of Capitanian deposits in the Lower Dalan due to the presence of the following foraminifera: Chusenella ex gr. conicocylindrica (Plate 1.1), in association with Climacammina grandis, Pachyphloia cf. solida, Schubertella sp., Neoendothyra cf. parva, Codonofusiella? sp., Eotuberitina reitlingerae, Dagmarita sp., Globivalvulina bulloides and staffellids such as Nankinella discoidea. In the Kuh-e Gakhum outcrop, previously studied but poorly characterised for the Upper Permian interval (Zaninetti et al., 1978), the Lower Dalan is characterised by the presence of Dunbarula nana (Plate 1.3) that is, at least, Capitanian in age. These results are consistent with the studies of Vachard et al. (2002) and Altiner et al. (2000) which suggest that the Capitanian is characterised in this area by Dunbarula, Chusenella and Shanita. Other samples in the Lower Dalan of the Kuh-e Gakhum outcrop section indicate the presence of Tubiphytes obscurus, Ungdarella sp., Pseudovermiporella sp., the dasycladacean calcareous algae Velebitella sp., numerous sponges, bryozoans, crinoids and corals.

Johnson (1981) documented the Chusenella Zone as occurring at the top of the ‘Lower Carbonates’ (nomenclatural equivalent to Lower Dalan) and an early Capitanian age was proposed for this level. The occurrence of Eopolydiexodina sp. in another section in the Kuh-e Surmeh region indicates that the Wordian is recorded below (Sharland et al., 2001). The importance of Capitanian and Wordian Eopolydiexodina and Shanita as biostratigraphic and palaeobiogeographic markers has been stressed successively by Sengör et al. (1988), Jenny and Stampfli (2000), Vachard and Bouyx (2002), Ueno (2003) and Kobayashi and Ishii (2003).

Consequently, the Lower Dalan Member is considered as Middle Permian (Guadalupian) in age with the presence of Wordian and Capitanian deposits as suggested by biostratigraphic results.

Nar Member: Permian Period, Guadalupian (Middle) Epoch, Capitanian Stage

Both Johnson (1981) and Baghbani (1988) document the Shanita Zone in the Kuh-e Surmeh section, in the upper part of what is called the Nar Member by Johnson (1981). The Shanita zone is considered of Midian (Capitanian) age (Baghbani, 1988; Vachard et al., 2002). The recognition of this genus at the base of the Kuh-e Dena section (Plate 1.2) confirms that the Nar Member, and its lateral equivalents, can be considered as Capitanian in age throughout the study area.

Upper Dalan Member, Lopingian (late) Epoch, Wuchiapingian and Changhsingian Stages

In spite of the relative abundance of fauna, no giant fusulinids (such as Schwagerinids) were observed in the Upper Dalan of Kuh-e Surmeh section and offshore Fars area; this indicates that the environmental conditions in the underlying Nar Member could correspond to the first phase of the end-Permian extinctions (latest Guadalupian), associated with a major regression (Jin et al., 1994). Classical markers such as palaeofusulinids and ‘advanced’ colaniellids (see Vachard et al., 2002) are absent; however this study documents for the first time the presence of numerous Paradagmarita (Biseriamminid) species throughout the study area. The occurrence of Paradagmarita is considered as intra-Lopingian but the precise range has not been unequivocally demonstrated: late Dzhulfian-early Dorashamian (Zaninetti et al., 1981; Altiner, 1981), latest Dzhulfian (Pronina 1988, 1999), latest Dzhulfian-Dorashamian (Altiner and Özgül, 2001), and Dorashamian (Vachard et al., 2002) have been proposed. The first appearance of Paradagmarita species provides a good element of correlation within the KS4b. These species (e.g. Plate 2.1) are sporadically present until the KS4c brecciated top. A ‘bloom’ with new species appearance is observed in the lower KS3a and they are relatively common until the Permian fauna extinction in the lower KS2 wherein the species P. monodi is particularly abundant.

Palaeogeographically this confirms the presence of the Paradagmarita province, roughly corresponding to the Late Permian foraminiferal Southern Biofacies Belt of Altiner et al. (2000) in this area. This province extended from the Taurides (Turkey) to the Arabian Platform (Lys and Marcoux, 1978; Altiner 1981, 1984; Zaninetti et al., 1981; Sengör et al., 1988; Köylüoglu and Altiner, 1989), south Iran (Argyriadis and Lys, 1977; Argyriadis, 1978), Saudi Arabia (Vachard et al., 2005), Oman (Montenat et al., 1977) and in general from Italy to Thailand, and characterises the Neo-Tethys seaway margins (Gaillot and Vachard, 2004).

According to our observations and the stratigraphic ranges proposed in the literature, the KS4 (Cycle IV) is considered as Wuchiapingian due to: (1) the absence of Schwagerinids compared to the relative abundance of fauna; and (2) the appearance of Paradagmarita genus in association with Graecodiscus (Plate 1.10), Rectostipulina, and primitive Colaniella. The KS4-KS3 boundary is considered in this study as the Wuchiapingian-Changhsingian limit and Paradagmarita (?) sp. 1 and P. sp. 4 are generally found just above the limit (‘Paradagmarita bloom’) (Plates 2.9, 2.10 and 2.14). The KS3a sequence is therefore considered as early Changhsingian, up to the appearance of Hemigordiospsis sp. 2 (Plate 1.14) associated with Sphaerulina (Plate 2.7) at the base of the KS3b sequence. The latter event could indicate a late Changhsingian age. The beginning of the KS2 transgressive shoal is associated with abundant Paradagmarita monodi (Plate 2.11), large globivalvulinids such as Urushtenella (?) sp. (Plate 2.18) and Paraglobivalvulinoides sp. (Plate 2.24) and considered as latest Changhsingian in age.

The Permian Fauna Extinction (PFE) event generally occurs within a strongly calcite-cemented ooid grainstone (F16 facies) in the lower part of the KS2 sequence (within the KS2b depositional unit). Above the PFE event an azoic interval, which can vary slightly in thickness from 1 to 3 meters, cannot be biostratigraphically assigned to the Permian nor Triassic periods. This azoic interval was followed by the Triassic faunal recovery and associated with the thrombolitic microbial event (F12). Several important markers such as foraminifers Rectocornuspira kahlori (Plate 2.26), and annelid Spirorbis phlyctaena (Plate 2.25) give an Early Triassic (Induan) age to the upper KS2b sequence following the thrombolitic level. As suggested by the carbon isotope studies (see below), the PFE would occur before the standard Permian-Triassic Boundary and therefore the 1–3 meters azoic interval following the PFE (F16) should be considered as Late Permian in age.

The KS4 and KS3 sequences are thus considered to provide a complete record of the Lopingian without any major gap just below the Permian-Triassic Boundary. The KS4 sequence is considered as Wuchiapingian (Dzhulfian), the KS3 as Changhsingian (Dorashamian) and the base of KS2 as latest Changhsingian. The lack of ‘classic’ Late Permian markers, such advanced Colaniella and Palaeofusulina, may be related to provincialism, different palaeobiogeographic regimes and environmental factors, without invoking a major gap in sedimentation at the end of the Permian Period (Johnson, 1981; Altiner et al., 2000) – the absence of marker fossils does not indicate an absence of stratigraphic time. This is consistent with the sedimentological, stratigraphic and lithological data which do not show any evidence for a major gap in sedimentation, nor exposure at the Permian-Triassic Boundary (see above).

Palaeoecology

Palaeoecologically, five biofacies types and nineteen subtypes have been defined based on thin section analysis, taking into account the faunal and algal content (mainly foraminifera and calcareous algae), the foraminiferal diversity (taxa richness), and the microfacies texture. This generalised classification was applied to the depositional models developed from the sedimentological analysis and has enabled a validation of the depositional schemes by identifying palaeoenvironmental trends (particularly proximal to distal tendencies, oxygenation levels and salinity gradients) which are not always obvious from the sedimentological analysis alone. These bio-data were integrated within depositional models with the development of detailed depositional-palaeoenvironmental schemes for the KS4 and KS3 sequences.

Marine faunal and algal distributions are related to many environmental factors including salinity, palaeobathymetry (by proxy), temperature, water energy, oxygenation, and illumination. Observations on Khuff subsurface and outcrops, within the context of this study, suggest that the fauna and flora distribution is mainly controlled by salinity (gradient of diversity from open marine sands and shoals to restricted lagoonal conditions) and oxygenation on the platform (water circulation and marine influence with protected poorly oxygenated zones showing low-diverse assemblages; Figure 42). Because in shallow-marine systems faunal migration events are strongly related to marine influences and flooding events, there is a direct link between sequence stratigraphy, biostratigraphy and palaeoecology (see Brett, 1995). Consequently the boundaries of the ‘palaeoecological systems’ (PS) defined in this work correspond to important sequence stratigraphic events and markers (such as maximum flooding events, transgressive surfaces and sequence boundaries) at various stratigraphic scales (sixth- to second-order). This relation has allowed the integration of ecostratigraphic events to the previously defined sequence stratigraphical framework based on the sedimentological and stratigraphic analysis (Figure 43). This integrated sequence stratigraphic-ecostratigraphic approach helps confirm and refine the stratigraphic correlations (Figure 44).

According to the pattern of recurring faunal/floral assemblages, their sedimentological context and palaeoenvironmental interpretation, a number biofacies (BF) types and subtypes have been created in order to support distal-proximal and palaeobathymetric interpretations. The principal Khuff Biofacies definitions and their palaeoenvironmental interpretations are summarised in Table 3 (also see Figures 42 and 43). The biofacies types (BF0 to BF5) were classified to take into account two main parameters: (1) the texture (fine- versus coarse-grained) to detect the hydrodynamic signals (shallowing-up or deepening-up and/or increasing – decreasing hydrodynamic protection); and (2) the general foraminiferal diversity (low, moderate, high) that is closely linked with salinity and oxygenation (and thus by proxy circulation, open-marine influence, and anoxia). In addition nineteen subtypes have also been defined based on their faunal content (calcareous algae and foraminifers) and are organised in hierarchical order from a more proximal protected system (attributed number = 1) to relatively more distal open-marine system (attributed number = 19).

The analysis of the biofacies distribution has allowed the subdivision of the Upper Khuff K4 and K3 reservoirs into 6 different palaeoecological systems that correspond to characteristic faunal assemblages and biofacies sets. The main characteristics of the six systems (PS1 – PS6) based on an offshore Fars subsurface reference well, and their lateral variability, are summarised as follows.

Palaeoecological System 1 (PS1)

In the offshore Fars subsurface reference well, PS1 mainly occurs in units KS4a, KS4b1, and KS4b2, and includes: (1) intertidal sands, leeward oo-peloidal shoals with dominant Biofacies BF2 subtypes; (2) mixed flats, salinas and hypersaline lagoons with dominant Biofacies BF1 subtypes, and associated with Biofacies BF0. Typical associated flora and fauna comprise the calcareous algae Mizzia and forams such as Staffellids, Globivalvulinids, small Miliolids and Earlandia.

These patterns of diversity and biofacies stratigraphic distribution are quite uniform from the south (Abu Dhabi; personal observation; El-Bishlawy, 1985) subsurface to the north (Kuh-e Surmeh outcrop). There is no major differences with the Iranian subsurface reference well in term of biofacies suggesting a very stable system in a very internal setting, well-protected from an open-marine influence (shoal/barrier present towards the north or northeast), and lacking significant marine flooding.

Palaeoecological System 2 (PS2)

In the offshore Fars subsurface reference well, PS2, mainly occurs in the KS4b3, KS4c1 to lower KS4c2, and comprises: (1) the rapid transgression of oolitic shoals (Biofacies BF2 subtypes) and bioclastic megaripples (organised in opening-up cycles (see above) with Biofacies BF3 and BF5) and lead to the appearance of important biomarkers such as Rectostipulina, Tetrataxis (Plate 2.4), Insolentitheca ? (Plate 1.21), and a typical mixing of calcareous algae such as Mizzia (Plate 1.4), Permocalculus (Plate 1.5), and Gymnocodium (Plate 1.6); and (2) the maximum flooding interval of the KS4 sequence with Lagenid-rich Biofacies BF4 to BF5 subtypes (inner to mid-ramp setting) followed by a tidal complex but still mid-ramp influenced.

Both events are assumed to be correlatable at a regional scale and correspond to a major transgressive event, but with regional differences in its characterisation. In the Zagros wells and outcrops, this transgressive event is characterised by the absence of oolitic shoal but with the presence of a thick well-cemented high diverse bioclastic grainstone with BF5 biofacies. This is followed by lower diversity BF1 biofacies subtypes that replace the Lagenid-rich deposits during the maximum flooding because of deeper environments and associated poorer circulation. Moreover, muddy deposits still prevail in this locality during lower KS4c2 while laterally in the offshore Fars subsurface the environments are grainier indicating shallower water. The transgressive event is less marked southwards and BF5 facies can be absent. Only a thinner Lagenid-rich muddy interval corresponds to the maximum flooding interval and diversity is noticeably lower (but without anoxia). These results strongly suggest a more ‘distal’ marine-influenced position on the platform for the Zagros wells to more proximal conditions southwards. These interpretations again corroborate the sedimentological analysis and depositional model. Compared to PS1, this system gets increasingly diverse and reflects the establishment of new foraminiferal communities during a general drowning event that probably extended widely on the Arabian Platform during the Wuchiapingian.

Palaeoecological System 3 (PS3)

PS3 is mainly in units KS4c2 to KS4c4, and shows: (1) depositional units with an aggradational then regressive trend with a large tidal sandwave complex and oolitic shoals with dominant BF2 biofacies subtypes (with biomarkers Charliella (Plate 2.8) and Rectomillerella (?) (Plate 1.19)), and rare BF4 biofacies representing inter-shoal environments; and (2) an abrupt change within the KS4c4 sequence with a return of hypersaline lagoons and marine muddy embayment with low foram diversity to azoic facies, and dominated by BF1 and BF0 biofacies subtypes.

As Palaeoecological System 1, PS3 is encountered from Abu Dhabi to the Zagros wells suggesting the stability of the system during this stratigraphic interval. Nevertheless, there are some differences within the KS4c4 that can be slightly richer in forams in the Zagros wells, suggesting again a more open-marine influence and ‘distal’ position on the platform compared to the offshore Fars reference well and the Abu Dhabi data.

PS3 is mainly characterised by a decreasing foraminiferal diversity, suggesting that the whole system get progressively isolated by shoaling northwards and/or reducing marine circulation on the platform.

Palaeoecological System 4 (PS 4)

This palaeoecological system is mainly encountered in KS3a1 to KS3a3 and is composed of depositional units showing rapid palaeoenvironmental variations with BF1, BF2, BF3 biofacies subtypes. KS3a2 sequence may include the interval of maximum marine influence of the KS3a sequence and includes an important correlatable event with abundant brachiopods and large lagenids (Cryptoseptida (?), Plate 2.13). In the Zagros wells, the foraminiferal diversity is noticeably higher suggesting better oxygenated environments, less affected by restricted circulation and hypersaline influence. This is also compatible with a more distal position for these wells.

The rapid variation in biofacies subtypes and faunal diversity of PS4 suggests high-frequency fluctuations in circulation effectiveness and/or rapid sea-level fluctuations in a general subtidal to intertidal setting, which can be locally poorly oxygenated (dysoxic – anoxic) and hypersaline (with subaqueous gypsum precipitation). However, in shallower and/or better oxygenated environments, assemblages show similarities with those of PS2 with abundant new Paradagmarita species and reflect an important flooding event at the base of KS3.

Palaeoecological System 5 (PS 5)

PS5 dominated the KS3a4 to KS3b4 interval and comprises very characteristic and repetitive high-frequency shallowing-up depositional units (1–2 meters thick) with BF2 biofacies at the base of the cycles to BF0 biofacies subtypes near the top. The cycles developed within the channelised tidal flats and lagoons that characterised these intervals during the late KS3a and KS3b, and the biofacies cycles show an aggradational stacking pattern which was also seen in the stratigraphic stacking analysis (see above). Basal transgressive sands are generally composed of dominant calcareous algae Mizzia (BF2d, intertidal sands), followed by rich-foraminiferal sands with Hemigordiopsis sp. 2 (FAD within KS3b1 interval), large Staffellids (BF2c) and occasionally, in more proximal environments, abundant well-preserved Paradagmarita (?) sp. 1. Small forams (such as Earlandia and Pseudomidiella) progressively replaced the large forams, and the sands finally become azoic and muddier (BF2a to BF1a biofacies evolution). Similarly, the KS3b of Zagros subsurface shows alternation of bioclastic grainstones (mainly tempestites) with lagoonal muds (reduced circulation). These environments are quite similar to the PS4 of the offshore Fars reference well and correspond to the migration of these environments towards the Zagros area during the KS3b which was a low accommodation (but aggrading) system.

Palaeoecological System 6 (PS 6)

The palaeoecological system PS6 characterises the KS2a and KS2b intervals and includes four successive environmental units, from base-up:

  • (1) After a level of transgressive sands, an important and first appearance BF3c biofacies event with abundant lagenid Robuloides (usually associated mid-ramp conditions), and the return to normal well-circulated marine waters as suggested by fauna such as bryozoan, Paradagmarita and Eostaffella (?). This event has been found on neighbouring wells and an equivalent can be found on the Kuh-e Surmeh outcrop. This event however appears to be absent southwards, once again suggesting a more proximal position. Nevertheless this event appears to have at least semiregional significance (offshore Fars to Zagros).

  • (2) The top KS2a breccia is characterised in microfacies by exposed-mudflat dolomudclasts, with well-preserved Paradagmarita (?) sp. 1, reworked within a more oxygenated mud (BF2a brecciated). This event is less marked in the more open ‘distal’ positions (Zagros subsurface and outcrops), but is nevertheless considered to correspond to an important, possibly semi-regional, flooding surface.

  • (3) The Permian Fauna Extinction (PFE) occurs within an oo-peloidal grainstone in the subsurface wells. In the offshore Fars subsurface reference well, the diversity drops from a transgressive bioclastic grainstone (BF4b) to the final oo-peloidal grainstone (BF2d) where the algae and forams seem to be subjected to intense calcification and syn-sedimentary calcite cementation. The same trend is observed southwards but with a lower diversity (BF2) which corresponds to more internal environments (lagoonal mud and mixed flats).

  • (4) The following units correspond to a coarse cemented lithoclastic grainstone with clasts possibly reworked from hardgrounds, and where mainly gastropods and bivalves seem to be able to survive (F16 facies). The first occurrence of Triassic fauna (Spirorbis phlyctaena and Rectocornuspira kahlori) is associated with the appearance of the thrombolitic facies (which are less developed southwards).

STRATIGRAPHIC ARCHITECTURE AND CORRELATIONS

The significant stratigraphic surfaces have been picked and correlated based on the detailed core descriptions, the bio- and ecostratigraphic analysis, wireline logs and the regional understanding of other Dalan-Kangan/Upper Khuff sections in the region (Figure 45, also see Figure 44) (Szabo and Keradpir, 1978; El-Bishlawy, 1985; Alsharhan and Kendall, 1986; Al-Jallal, 1987, 1994, 1995; Kashfi, 1992; Alsharhan, 1993; Alsharhan and Nairn, 1994a, 1997; Al-Aswad, 1997; Sharland et al., 2001). The correlations in cored wells for the Upper Dalan cycles are supported by a well-constrained biostratigraphic framework (see above). At the production-scale (tens to many tens of km) a series of Upper Khuff ‘event’ and ‘marker’ beds have been recognised, defined and correlatable over the North Dome field area – their lateral extent is a function of the very large scale of the carbonate platform and its flat ramp-like geometry. In particular, microbial events provide reservoir and regional scale marker units that are correlatable over large distances. These include the major thrombolite event within the KS2b, the approximate position of the Permian-Triassic Boundary, and of at least regional significance, similar units and occur in other parts of the Early Triassic Induan stratigraphy (such as the correlatable microbial event within the KS1b).

The major stratigraphic surfaces are clearly visible on the wireline logs, however difficulties in using just wireline logs for correlating some of the intermediate surfaces are evident, particularly when major lateral facies changes are present. Another significant problem encountered when using the wireline logs for correlations is that they do not recognise the different types of dolostone (especially the difference between the ‘early’ and ‘later’ overprinting dolostones) - the early dolostones have a sequence stratigraphic significance and can be used sequentially whereas the later dolostones are not directly linked to the stratigraphic architecture. The later dolostone overprints (often coarse sucrosic dolostones) are often responsible for anomalous thick dolostone sequences.

Based on the integration of the regional correlations, the sedimentological model and the biostratigraphic-palaeoecological data, a large-scale stratigraphic architecture was established. The correlations and stratigraphic analysis suggest that the major stratigraphic trends and large-scale stratigraphic architecture are relatively ‘layer-cake’ at the production scale (at the scale of individual fields), a function of the almost flat platform geometry. At a larger scale significant changes in thickness occur, either thickening towards palaeodepocentres (such as the Nar-Kangan fields intrashelf low during the KS4 and KS3, or the general thickening of the Upper Khuff towards the northwest), or thinning and onlap towards palaeohighs (such as towards the Zagros High during the KS3 and KS4, and the general thinning towards the south). Based on this analysis four large third-order stacking cycles have been identified and regionally correlated (Figure 3).

Stratigraphic Architecture

The large-scale (third-order) stratigraphic architecture is summarised as follows:

Cycle IV (Figures 46 and 47)

The base of this unit is marked by a major sequence boundary at the top of the Nar anhydrite or its equivalent (the Nar anhydrite thins northwards of its depocentre (near the Nar-Kangan fields region) and is absent in the northern Zagros area which were palaeohighs during the KS4; the Nar Member anhydrite is correlated with a thick gypsum level in the Kuh-e Surmeh section which grades into a palaeosoil unit in the region of the Kuh-e Dena section.) The KS4 is a third-order cycle composed of three fourth-order cycles (KS4a–c), and is essentially transgressive in nature with a zone of maximum accommodation in the lower half of KS4c. This is also the zone of maximum aggradation, shoal and porosity development, particularly over the North Dome.

At a large scale it can be seen that the system has four major phases:

  • (1) Flooding with a backstepping architecture onlapping palaeohighs, and with the maximum development of muddy embayment facies (and its most landward extent) within the transgressive half-cycle.

  • (2) Construction of a well-developed aggrading grainy platform system (KS4c – zone of maximum accommodation). At the base of KS4c is a marked and correlatable ‘flooding/opening’ event that can be correlated across the region and is well-characterised by the faunal studies (see above). This opening event represents the peak of flooding, but accommodation is still being developed even towards the top of KS4c.

  • (3) Progradation northwards at the top of the KS4 cycle; the progradation of shoals over embayment muds in the Nar-Kangan fields intrashelf low area is particularly notable.

  • (4) Platform-high exposure during base level lows at the top of the KS4. This is marked by a well-developed collapse breccia exposure unit on the platform top part of the succession, which becomes much less marked or absent further towards the north where the palaeolows were situated.

Each of the three fourth-order cycles also shows a dominant flooding half-cycle (retrogradational-backstepping) and a comparatively minor progradational thickness.

Cycle III (Figures 46 and 47)

This is a third-order cycle with a zone of maximum accommodation in KS3a3 (middle to upper KS3). It is a system composed of restricted evaporative coastal lagoons and shallow-water muddy embayments with limited reservoir development, except towards the north where the system gets increasingly grainy. The maximum of accommodation is well-picked out by the maximum development of grainstones in the platform top edge sequence. The top of this cycle is marked by the reappearance of platform interior mud and evaporitic flat facies with associated exposure surfaces. The fourth-order cycles are not evident and have yet to be correlated, it is possible that they are either highly condensed or reduced/lost during the trangressive onlap.

Cycle II (Figures 48 and 49)

This can be clearly split into two units the lower Cycle II (equivalent to the KS3b stratigraphic interval, and essentially the upper part of the Upper Dalan Member) and the upper Cycle II (equivalent to the KS2 stratigraphic interval, and essentially the lower part of the Kangan Formation).

Lower Cycle II: This is a fourth-order cycle within a third-order transgressive half-cycle. It is consequently highly aggradational with a thick sequence of stacked peritidal sediments, sabkhas and vertically burrowed/rooted beds on the platform top where there is very limited accommodation (hence reservoir development is poor). In the Nar-Kangan fields intrashelf low area a packet of aggradational embayment mud facies developed. Towards the north in the palaeohighs of Kuh-e Surmeh and Kuh-e Dena, the units thin and are dominated by bioclastic sheet sands coming off the palaeohighs areas with more open high energy conditions (reworking of local reefal patches). At the top of the KS3b more internal lagoonal facies can be seen prograding over the embayment muds in the Nar-Kangan fields intrashelf low area. The increase in accommodation in this unit may also be related to increased subsidence rates in the North Dome field area.

Upper Cycle II: This is a fourth-order cycle within a third-order late transgressive half-cycle, maximum accommodation zone and regressive half-cycle. It is composed of grainy aggradational facies and local mudstones on the platform tops where the cycle has good reservoir development around the maximum accommodation zone. The facies include intertidal sands and shallow-water thrombolitic facies (at the base of the Triassic), followed by storm-generated pebble grainstone beds and shoals, and microbially cemented grainstones.

Towards the north in the deeper areas of the palaeolows the system is dominated by deeper embayment muds and mid- to outer-ramp muds with a landward stepping architecture of outer embayment muddy facies. This deeper area includes the former Nar-Kangan fields intrashelf low, but also the northern Zagros region which is no longer acting as a palaeohigh and is now a deeper mid – outer ramp position (Triassic change in shelf geometry - see above). Within this unit, just after the Permian-Triassic limit (within the KS2b – lower part of KS2) the regionally correlatable thrombolite event occurs. In the Eastern Zagros zone there is the appearance of high gamma-ray muddy shales for the first time in the Upper Khuff.

At the top of the KS2 higher energy platform top facies prograde from the south to the north over the embayment facies with grainstones and lagoonal muds prograding over embayment muds in the Nar-Kangan fields area. This gives rise to good reservoirs at the top of the KS2 in these more distal areas. As noted above the stratigraphic, facies, lithological and palaeoecological analysis suggests that there is no major stratigraphic gap between the lower and upper Cycle II (and therefore between the end of Permian and the Early Triassic). The analysis suggests there is, in fact, a low-order transgression between upper K3 and K2.

Cycle I (Figures 48 and 49)

This third-order cycle is composed of three fourth-order cycles (KS1a–c), with a zone of maximum accommodation in the KS1b. This is again the zone of maximum aggradation and maximum backstepping of the poorly oxygenated embayment-influenced facies onto the platforms (local thrombolites and other microbial facies within the KS1b). This third-order cycle appears to have a series of transgressive pulses that generate accommodation that is filled by grainstone facies. On the platform top, the low accommodation setting limits the development of shoals which are preferentially developed in the higher accommodation settings off the palaeohighs. Also noteable is that the shoal complexes prograde northwards over the embayment facies in the upper part of the KS1 cycle.

What is also significant is that within Cycle I (KS1b) there are microbial facies and local thrombolites which are regionally broadly correlatable. This represents a second significant microbial oceanographic event and has a similar associated fauna than that associated with the Permian-Triassic thrombolites (microgastropods, thin-shelled bivalves/ostracods). Similar microbial facies are present in other wells at the same stratigraphic layer and are often associated with blackened grained peloidal packstones (F20). This suggests a second correlatable microbial event in the K1. As with the K2 thrombolites they are located in the late third-order TST. Similar correlatable microbial events, after the main PTB thrombolites, occur within the shallow-water carbonate platforms of the Nanpanjiang Basin, South China, where it is suggested that detrimental oceanic events within the Early Triassic Induan result in a number of discrete calcimicrobial framestone horizons (Lehrmann et al., 2003).

The top of the KS1 is dominated by shaly peritidal and evaporitic supratidal flats with tidal channels and local microbial facies in the low accommodation zones (North Dome field area) and more grainy oolitic shoal and peloidal sand sheet facies in the high accommodation settings to the north (Nar-Kangan fields area to northern Zagros area). At the top of the KS1 in the eastern Zagros zone the high gamma-ray muddy shales (more or less developed since the KS2b) retreat and are no longer picked-up by the well control. The top of the Kangan Formation is followed by the flooding related to the regional deposition Sudair and Aghar shales.

THE PERMIAN-TRIASSIC TRANSITION

The precise nature of the K3–K2 reservoir transition, and thus the Permian-Triassic Boundary, has always been problematic with many interpretations being put forward. Among the different proposed scenarios, some models suggest a major sequence boundary at the top of the K3 (Top Permian) and a significant stratigraphic time gap between the Permian and the Triassic (Szabo and Kheradpir, 1978; Johnson, 1981; Kashfi, 1992; Erwin, 1993; Altiner et al., 2000; Heydari et al., 2001; Sharland et al., 2001). Other models suggest no major sequence boundary at the top of the Permian and a continuous transgression with no major sedimentation gap between the Permian and Triassic (Wignall and Twitchett, 2002; Heydari and Hassanzadeh, 2003; Heydari et al., 2003; Weidlich and Bernecker, 2003; Lehrmann et. al., 2003; Wignall et al., 2005). These different models for the Dalan-Kangan transition have a major impact on the correlation strategy of the K3 and K2 reservoir intervals, at both the reservoir and Gulf scale.

The study data provides an ideal opportunity to examine the details of the Permian-Triassic Boundary and the K3–K2 reservoir transition on a large Gulf-scale database (subsurface and outcrop). This study integrates at high-resolution the sedimentological, sequence stratigraphic, biostratigraphic, wireline log and 13C/12C isotopic analysis across the Permian-Triassic Boundary.

Sedimentological, Stratigraphic and Palaeoecological Synthesis

Stratigraphic, facies, lithological and palaeoecological analysis suggests that there is no major stratigraphic gap between the K2 and the K3, and hence no major stratigraphic gap between the Permian and Triassic periods. This is based on the following observations and interpretations:

  • (1) The transgression onto the platform started in the top of KS3b where increasing high energy grainy facies occur - suggesting the platform top was becoming more open before the Permian-Triassic Boundary. This is confirmed by the biofacies analysis which suggests a significant transgression and opening of the platform system prior to the end Permian.

  • (2) Paralleling this pre-end Permian appearance of higher energy grainy facies, is the dolostone to limestone transition which shows that the transition from dolomite-dominated platform restricted settings to more open limestone facies also occurs before the Permian-Triassic Boundary (see above).

  • (3) In the numerous subsurface sections, and outcrop investigations in the Zagros, no evidence is seen for a major exposure event between the Upper Dalan (Late Permian) oolitic and bioclastic grainstones and the Kangan (Early Triassic) microbially-influenced sands and microbialites – this contact is conformable.

  • (4) There is no major unconformity/disconformity or stratigraphic surface associated with the Permian-Triassic Boundary and the extinction of Permian fauna occurs within a grainstone body.

  • (5) The analysis suggests there is in fact a low-order transgression which started at the base of the KS3b (base Cycle II) and continued through to the KS2 (upper Cycle II), and hence a transgression across the Permian-Triassic Boundary.

  • (6) Also significant is that the Szabo and Kheradpir (1978) and Kashfi (1992) publications, from which the commonly quoted Permian-Triassic unconformity at the Kuh-e Surmeh section originates, provide no physical evidence of an erosional or exposure surface at outcrop, and the unconformity is mainly based on interpretational criteria (change in thickness, change in colour, absence of index fossils) (see above).

As is demonstrated by the faunal analysis, the Permian Fauna Extinction (PFE) event generally occurs within a strongly calcite-cemented ooid grainstone (F16 facies) in the lower part of the KS2 sequence (within the KS2b depositional unit). Above the PFE event a Permian azoic interval, which can vary slightly in thickness from 1–3 meters, this is followed by the Triassic faunal recovery and associated with the Early Triassic thrombolitic microbial event. Several important markers such as Rectocornuspira kahlori (foraminifera, Miliolid) and Spirorbis phlyctaena (annelid) give an Early Triassic (Induan) age to the upper KS2b sequence following the thrombolites. This is corroborated by carbon isotope studies (see below). In KS3b, the Permian fauna and flora are clearly not reworked without any major exposure surface between the PFE and the basal Triassic fauna suggesting that, if one considers a gap exists, the latter may be situated at the top of KS2a sequence (KS2a surface), and not at the Permian-Triassic Boundary (i.e. directly above the PFE).

In the Zagros area (Kangan, Kuh-e Surmeh), the microbially-influenced strongly-cemented grainstone (F16 facies) is absent and the PFE occurs within muds under strongly anoxic influence (as suggested by early pyrite). This is similar to the pyrite framboids seen in the Permian-Triassic succession of the Khunamuh Formation, Kashmir, where it is linked to oxygen-poor deposition during the Late Permian faunal crisis (Wignall et al., 2005). The thrombolites appear less than 1 meter above the last Permian fauna. Hence the outcrop data also show a similar pattern with a thin azoic interval occurring between the last Permian taxa and the first Triassic taxa. Again no major surface is seen in the Zagros outcrops linked to a major SB nor PTB. Indeed there is a general muddying (deepening-upwards) from the lower Cycle II (KS3a) to the upper Cycle II (KS2a). The analysis, together with the facies, stratigraphic and biostratigraphic data presented earlier, suggests there is in fact a low (third-) order transgression between upper K3 and K2. These represent the ‘Permian-Triassic oceanic event’ and are located in the late third-order TST. This Cycle II with a low-order transgression across Permian-Triassic corresponds to the Khuff-B member in the Saudi Arabian subsurface (Al-Jallal, 1995), which also forms a cycle from the Late Permian to the late Early Triassic (also see Strohmenger et al., 2002).

This is consistent with recent studies which suggest that the end Permian extinction event is associated with a global transgression and the encroachment of anoxic waters (Wignall and Twitchett, 2002; Lehrmann, et al., 2003; Wignall et al., 2005). Moreover Weidlich et al. (2003) suggest that not only is there evidence of general transgression over the Permian-Triassic, but that the extinction was due to oxygen decline (by long-term atmospheric oxygen drop) in shallow-water carbonate reefal systems. In this model there may be no direct evidence of anoxic events at the extinction level (as is the case for the offshore Fars subsurface sections). Indeed the sections in the current study, and their interpretation, are very analogous to the Permian-Triassic Boundary sections from shallow-water carbonate platforms of the Nanpanjiang Basin, South China (Lehrmann et al., 2003). In these sections the end Permian extinction event occurs a meter or so before the Permian-Triassic Boundary, and is coincident with the appearance of calcimicrobial framestone. Also significant is that the stratigraphic section is associated with long-term transgression over the Permian-Triassic Boundary with no evidence of sedimentary hiatus nor exposure (Lehrmann et al., 2003). These observations and interpretations are completely inline with those of the current study in the Zagros-Fars area. This interpretation is also supported in the current study by the biostratigraphic analysis which suggests the Upper Dalan Member represents a complete record of the Lopingian (Late Permian) without a major gap at the Permian-Triassic Boundary (see ‘Biostratigraphy and Dating’ section above).

Other studies on the Permian to Triassic succession in other parts of the world, and in different platform settings suggest that there is an overall global transgression across this boundary and have associated the PFE, at least in part, to transgression of anoxic waters (Wignall and Twitchett, 2002; Wignall et al., 2005). Moreover the global Permian extinction event has also been attributed to massive gas hydrate (CH4) degassing into the ocean and promoting extensive syn-sedimentary sea-floor cementation as a result of the highly carbonate-supersaturated seawater (Heydari and Hassanzadeh, 2003), with no evidence of sea-level fall (Heydari et al., 2003). These “smoking-gun” syn-sedimentary sea-floor cements, which according to Heydari and Hassanzadeh (2003) mark the precise PTB, appear to be the lateral equivalent to the F16 facies which are characterised by coarse calcite isopachous fibrous cements and reworked intraclasts (see above). As with the sea-floor cements (Heydari and Hassanzadeh, 2003), the first appearance of the F16 facies marks the precise PTB event in this study.

If this is the case and there is no major exposure surface (sequence boundary and associated stratigraphic gap) at the Permian-Triassic Boundary in the study area, what does the surface often cited in Middle East literature as the Permian-Triassic third-order sequence boundary actually represent? (Strohmenger et al., 2002; Sharland et al., 2001). One possibility is that it could represent the Late Permian KS2a surface, which is well-expressed and interpreted as a fourth-order cycle boundary (within an overall third-order transgression). However this surface is still within the Late Permian and is followed by a transgressive bioclastic grainstone with a rich and well-developed Permian open-marine fauna representing flooding (taphonomic analysis of the fauna show that they are not reworked). The stratigraphic stacking, facies analysis, lithological analysis and palaeoecology all show that this unit occurs within an overall flooding sequence (progressive flooding of the platform interior settings), and hence unlikely to represent an intra-Upper Permian third-order sequence boundary, and is a surface of lower order. In the deeper settings, such as the Zagros outcrops, Late Permian muddying-upwards (deepening) continues across the Permian-Triassic Boundary and is consistent with a general transgression. It is however possible that over palaeohighs (such as in the southern Arabian Plate) non-deposition at the KS2a surface encloses the Permian-Triassic Boundary. The non-deposition may correspond to a hiatus preceeding onlap by the transgressive sequence; however the data is not currently available to test this stratigraphic model.

Isotopic Data and Spectral Gamma Ray Analysis

Carbon (δ13C) and Oxygen (δ18O) Isotopic data

Carbon and oxygen isotopic analysis of the Permian-Triassic Boundary on a offshore Fars section was carried out with the objective of establishing the stratigraphic location of the well-documented δ13C and δ18O negative minima event associated with the Permian-Triassic Boundary (see Yin and Tong, 1998). Analysis was carried out on bulk rock samples avoiding zones with too much anhydrite (lack of large bioclastic debris did not allow selective sampling), and treated with an acetic acid attack. Samples were taken within the Permian sequence (which contained a rich microfauna), through the anoxic interval and finally into the thrombolites at the base of the Early Triassic.

The 13C/12C results (Figure 50) showed Permian values of 3 to 4 drop gently, from the KS3b surface, towards the top of the Permian interval, and then more rapidly up to the negative minima associated with the Permian-Triassic Boundary. The minima is reached within the thrombolite level with a value of –0.9. The δ18O minima coincides with the appearance of the F16 facies and hence also the last of the Permian taxa in this locality. The signature of the δ13C curve is identical to other published 13C/12C curves for the same stratigraphic interval (see Jin et al., 2000 and Yin and Tong, 1998). This negative shift is thought to correspond to a dramatic decline in organic productivity resulting from the mass extinction or slow oxidation of the carbon reserve over a few million years. Once the 13C/12C data is calibrated on: (1) extinction of most benthic taxa; (2) the facies distribution; and (3) the 13C/12C minima event found in the thrombolitic bed, it is seen that there is an interval which is still Permian (upper Dorashamian), but devoid of any Permian benthic microfossils (due to the end Permian mass extinction event). This unit is dominated by the F16 facies. In fact, the first “Triassic newcomers” would appear in this interval, and in some cases benthic Permian microfauna are quoted as occurring within this interval (Yin and Tong, 1998).

Sulphur isotopic data (δ34S)

The sulphur isotope analysis carried out on the anhydrites from an offshore Fars well show there is a well-developed environmental signal (Figure 51). The details of the sulphur isotope–depth curve shows that not only were Permian values seen in the K4 and K3, and Triassic values in the K2–K1, but the detailed trend within these reservoirs reflects the early environmental isotopic signal. The interpolated inflection point would occur within the area defined as the Permian-Triassic Boundary from other lines of evidence (see above). This would suggest that if remobilised anhydrite was sampled, the late stage remobilisation was extremely local and little mixing of anhydrites of different stratigraphic levels occurred.

Spectral Gamma Ray analysis

A number of Upper Khuff logs which have Spectral Gamma Ray wireline log data show major shifts at or around the Permian-Triassic Boundary (K3–K2). At this level there is a significant drop in Uranium across the Permian-Triassic Boundary in both grainstone facies and muddy facies. This suggests that this ‘Uranium event’ is independent of local facies conditions. In more muddy facies there is a significant increase in Thorium and Potassium across at the Permian-Triassic Boundary in muddy facies. The ‘Uranium event’ can be easy correlated on the wells for which data is available. This suggests that this event is a good marker event at the Gulf-scale and that it represents a chronostratigraphic time-line from standard logging tool. High-resolution calibration with palaeontological data suggests that this event is coincident with the faunal turnover at the Permian-Triassic Boundary. High resolution core calibration of the Spectral Gamma Ray event shows that it does not relate to a major stratigraphic surface but correlates with the end of the Permian fauna and the start of the early cemented microbially mediated grainstones with lithoclasts (F16 facies), just before the thrombolites (Figure 52). This suggests that the cause of the Spectral Gamma Ray shift was a chemical oceanographic change associated with the Permian-Triassic mass extinction and influences from changes in ocean biomass.

The precise relationship between the appearance of the F16 facies (and hence also the last of the Permian taxa in this locality) with: (1) the end Permian and Early Triassic δ18O and δ13C isotopic signatures (the δ18O isotopic minima); (2) coincidence of the appearance of the F16 facies with the last Permian fauna; (3) appearance of the F16 facies and spectral gamma ray signature; (4) their correlation of the syn-sedimentary sea-floor cements of Heydari and Hassanzadeh (2003) which mark the precise PTB; and (5) their association with the Early Triassic thrombolites, suggests that this facies is a key sedimentological units for the recognition of the end Permian extinction and the following Permian-Triassic Boundary. This facies is present on all the Permian-Triassic platform successions seen in the area (personal observation: Saudi Arabia, Abu Dhabi, Qatar and Iran), even where the more classic thrombolite units are very poorly developed or absent (such as in Abu Dhabi). This illustrates the regional and reservoir correlative value of identifying this facies.

SUMMARY AND CONCLUSIONS

The principal conclusions are as follows:

Facies Associations and Depositional Environments

The synthesis core descriptions and the Zagros outcrop facies data, together with integration of published data has allowed sixteen principal facies associations to be defined, characterised and interpreted. Qualitative comparisons of Upper Khuff sections and subsurface cores across the Zagros-Gulf-Arabian Peninsula, show that this depositional facies classification is applicable at a larger regional scale (including Bahrain, Iran, Kuwait, Qatar, Saudi Arabia and the United Arab Emirates) and useful in rapid regional comparisons and correlations of the Upper Khuff depositional systems. The depositional facies classification should act as a rapid means of studying and comparing the depositional facies of different Upper Khuff sequences, and their corresponding reservoir impact, at both reservoir and regional scale.

The facies in this framework have been interpreted in terms of depositional environment including: (1) evaporitic flats or saltern (supratidal/intertidal to subtidal setting), (2) tidal flats (intertidal to supratidal setting, with numerous subenvironments such as beach ridges, intertidal flats, tidal channels, microbial mud and sand flats), (3) subtidal lagoon (subtidal setting, including microbial build-ups), (4) leeward shoals (subtidal to intertidal setting), (5) oolitic to oobioclastic shoal belts (subtidal to intertidal setting), (6) composite sandwave constructions (subtidal setting), (7) middle shelf (subtidal setting) and outer-ramp settings. This large range in facies types documented in the Upper Khuff sequence reflect the large range in depositional systems and sub-systems present across the Khuff platform (ranging intertidal and supratidal flats, evaporitic salterns to mid-ramp deposits), but also the temporal evolution of the Khuff environments and palaeoceanographic conditions from the Permian to the Triassic with periods of low oxygenation (and possible anoxia) to periods of supersaturated oceans just after the Permian-Triassic extinction event.

The general importance of microbial facies is highlighted and a variety of microbial facies are defined. In particular, in addition to the well-known KS2 stratigraphic interval microbial event, there are a suite of microbial facies that have previously been underestimated for the KS1 interval. Microbial units are useful as regional isochronous marker horizons, for example the KS2 thrombolite is present from Ghawar field (Saudi Arabia) to the Zagros (Iran). These microbial facies are often associated with periods of poor oxygenation and restriction, but nevertheless can occupy a range of environments from intertidal to mid- to outer-ramp settings.

Organic-rich Upper Khuff facies include green shales (restricted to KS1 interval), and facies related to restricted poorly oxygenated embayments and hypersaline lagoons and tidal flats. These facies can be rich in organic matter and TOC of up to 10% have been recorded in the KS3 interval. These facies are often related to hypersaline systems and can be regarded as evaporitic source rocks. The main reservoir prone facies are essentially the grainstones and packstones (in terms of reservoir quality and thickness), particularly the oolitic grainstones, oobioclastic grainstones, and the microbially influenced oolithoclastic grainstones.

Stratigraphy architecture and correlations

The significant stratigraphic surfaces have been picked and correlated based on the detailed core descriptions, the bio- and ecostratigraphic analysis, wireline logs and the regional understanding of other Dalan-Kangan/Upper Khuff sections in the region. The correlation in cored wells for the Upper Dalan cycles is supported by a well-constrained biostratigraphic framework. At the production-scale (tens to many tens of km) a series of Upper Khuff ‘events’ and ‘marker’ beds have been recognised, defined and correlated over the North Dome field area – their lateral extent is a function of the very large scale of the carbonate platform and its flat ramp-like geometry. In particular, microbial events provide reservoir and regional scale marker units that are correlatable over large distances. This includes the major thrombolite event within the KS2b interval, the approximate position of the Permian-Triassic Boundary, and of at least regional significance, similar Early Triassic Induan intervals.

Four large third-order stacking cycles have been defined on the basis of cycles bounded by surfaces representing base-level and accommodation potential minima, have been identified and regionally correlated based on surface characteristics, stratigraphic stacking patterns, faunal/floral events, and the large-scale regional context.

Based on the integration of the regional correlations, the sedimentological model and the biostratigraphic-palaeoecological data, a large-scale stratigraphic architecture has been established. The correlations and stratigraphic analysis suggest that the major stratigraphic trends and large-scale stratigraphic architecture are relatively “layer-cake” at the production scales (at the scale of individual fields), a function of the almost flat platform geometry. At a larger scale, significant changes in thickness occur, either thickening towards palaeodepocentres (such as the Nar-Kangan fields intrashelf low during the KS4 and KS3, or the general thickening of the Upper Khuff towards the northwest), or thinning and onlap towards palaeohighs (such as towards the Zagros High during KS3 and KS4, and the general thinning towards the south). At this large-scale, progradation of the oolite shoals (and hence reservoir facies) is seen during the late highstands in the high accommodation areas. On the topographic highs and platform tops, however, the main stratigraphic locations of the oolite shoals are in the trangressive and maximum accommodation zones of the cycle.

The cycles defined are as follows:

  • Cycle IV: KS4 stratigraphic interval – lower Upper Dalan Member, and encompassing the K4 reservoir interval. The maximum flooding surface of this cycle would correspond to the P30 MFS of Sharland et al. (2001, 2004).

  • Cycle III: KS3a stratigraphic interval – upper Upper Dalan Member, and encompassing the lower K3 reservoir interval. This cycle includes the P40 MFS of Sharland et al. (2001, 2004).

  • Cycle II: KS3b and KS2 stratigraphic intervals – top Upper Dalan Member and lower Kangan Formation, and encompassing the upper K3 and K2 reservoir intervals. The maximum accommodation zone of this cycle corresponds to the Tr10 MFS of Sharland et al. (2001, 2004).

  • Cycle I: the KS1 stratigraphic interval – middle to upper Kangan and encompassing the lower K1 reservoir interval. The cycle maximum accommodation zone corresponds to the Tr20 MFS of Sharland et al. (2001).

Conceptual depositional model and platform morphology

Integrating the facies and sedimentological interpretations, with the stratigraphic interpretations, conceptual depositional models have been constructed for the intervals: Cycle IV, Cycle III, lower Cycle II, upper Cycle II, and Cycle I. From these interpretations and models it is clear that there have been significant changes in platform type/geometry, facies organisation and climate from Cycle IV through to Cycle I. It is also seen that although most of the facies occur in all the major stratigraphic cycles their relative proportions and distributions are not the same (volumetric and facies partitioning).

The general palaeogeographic context of this system was a marginal-marine shelf setting with an inner platform that was very flat, ramp-like, with little topography, but with local depressions. The platform becomes more distal to the north and northeast with the presence of more open-marine influences. In the inner shelf areas, where most of the reservoir facies sedimentation took place, depositional environments were prone to be restricted with limited circulation due to high carbonate production and low accommodation potential.

In this large shallow low-energy platform interior system it was necessary to create accommodation for extensive shoal development and hence reservoir development. The creation of accommodation allows the development of sufficient palaeobathymetry to generate open higher hydrodynamic energy conditions (by both wave and tidal action), and to provide the space for sediment to accumulate – both fundamental prerequisites for the development of thick good-quality grainstone reservoirs in such settings. This essential development of accommodation was initiated by platform flooding and hence transgression. Inversely, sea-level falls can quickly and drastically isolate the platform interiors reducing the internal hydrodynamic energy levels thus stopping the development of reservoirs over vast areas, and creating widespread evaporitic and carbonate mud seals. There are no modern analogues for such epeiric seas in terms of platform morphology and scale, nor extent of facies distributions and their sediment dynamics.

At a large-scale the Upper Khuff depositional setting during the KS4 and KS3a was organised into a platform profile that gently deepened from the south (southern North Field) with a platform top interior zone (North Field), a platform top edge zone (northern South Pars), an intrashelf low (Nar-Kangan fields), and then rose again in the north with palaeohighs around Kuh-e-Surmeh and Kuh-e-Dena (structurally-controlled basement highs). There is however a major change in the platform profile from this KS4–KS3a profile (Permian), to the KS2b–KS1 profile (Triassic). The KS2b–KS1 profile had a monoclinal ramp geometry that opened to the north to deeper marine conditions with the absence of effective palaeohigh barriers around Kuh-e-Surmeh nor Kuh-e-Dena (as determined by the facies data). These two large-scale palaeogeographic profiles control the overall distribution of facies belts across the platform.

This change in platform profile is coincident with other events within the KS2b (the Permian-Triassic Boundary), including: (1) major facies changes on the platform tops (offshore Fars subsurface) where there is the appearance of thrombolites and associated microbial grainstones; (2) major facies changes in the northern shelf edge areas (northern Zagros) where there is a change from shallow-water high-energy grainy facies to deeper-water mid-ramp muddy facies; (3) change in relative stratigraphic thickness from a northern Zagros area with a thinner Upper Dalan stratigraphic thickness relative to the offshore Fars subsurface (due to the presence of the Zagros palaeohighs), to slightly thicker Kangan succession of open shelf muddy facies (due to high accommodation in a deeper open-marine setting); and (4) the appearance of high gamma-ray shales in the eastern Zagros subsurface area, not seen before the KS2b, and only seen in this area.

These events are all consistent with a major flooding across the Permian-Triassic Boundary causing: (1) a drowning of palaeohighs; (2) encroachment of anoxic waters into the shelf area lows; (3) terminating bioaccumulations (such as small reefal patches) at the shelf edges; (4) flooding the platform tops with more grainy facies (after the muddy lagoonal Upper Dalan), and developing microbial facies across the shelf; and (5) the quasi-synchronous Permian-Triassic mass extinction. This is consistent with recent studies which suggest that the end Permian extinction event is associated with global transgression and the encroachment of anoxic waters.

In general, at the field-scale (tens to many tens of kilometers) there is a great deal of similarity between wells in terms of facies types, facies distributions and sedimentological systems within individual depositional units. This reflects the very large scale of the carbonate platform and its overall flat ramp-like geometry. There are however subtle but significant variations in facies types, lithologies (calcite versus dolomite versus anhydrite) and poroperm zones from one section to the other. At a larger-scale (hundreds of kilometres) more systematic trends in facies textures (grain- versus mud-supported), and consequently reservoir potential, are developed. These are related to changes in palaeobathymetry, position of intra-platform lows and hydrodynamic energy gradients, and hence the position of the main shoal belt.

Biostratigraphic and Palaeoecological Analysis

Based on the stratigraphic distributions of the biostratigraphically significant fauna and flora the following age determinations are proposed:

Lower Dalan - Midian

Nar Member - Midian

Upper Dalan - KS4 interval – Wuchiapingian (early Lopingian)

Upper Dalan - KS4/KS3 boundary – Wuchiapingian/Changhsingian boundary

Upper Dalan - KS3a interval – early Changhsingian

Upper Dalan - base of the KS3b interval – late Changhsingian

Upper Dalan – base of KS2 – latest Changhsingian.

Palaeoecologically, five biofacies (BF) types have been defined based on thin section analysis taking into account the faunal and algal content (mainly foraminifers and calcareous algae), the foraminiferal diversity (generic and species richness), their sedimentological context and palaeoenvironmental interpretation. This generalised classification is applied to the depositional models developed from the sedimentological analysis and has enabled a validation of the depositional schemes by identifying palaeoenvironmental trends (particularly proximal to distal tendencies, oxygenation levels and salinity gradients) which are not always obvious from the sedimentological analysis alone. These bio-data have been integrated within depositional models with the development of detailed depositional-palaeoenvironmental schemes for the KS4 and KS3 sequences which integrate palaeoecological and sedimentological information.

The analysis of the biofacies distribution has allowed the subdivision of the Upper Khuff K4 and K3 reservoirs into 6 different ‘palaeoecological systems’ (PS) that corresponds to characteristic faunal assemblages and biofacies sets. The main characteristics of the six palaeoecological systems (PS1–PS6) based on an offshore Fars subsurface reference well, and their lateral variability, have been documented. The boundaries of palaeoecological systems defined in this work correspond to important sequence stratigraphic events and markers (such as maximum flooding events, transgressive surfaces and sequence boundaries) at various stratigraphic scales (sixth- to second-order). This relation has allowed the integration of ecostratigraphic events to the previously defined sequence stratigraphical framework based on the sedimentological and stratigraphic analysis and hence confirm and refine the stratigraphic correlations.

The palaeoecological and biostratigraphic studies are shown to be a powerful tool in: (1) both high-resolution reservoir and regional-scale correlation, being able to provide correlation schemes where little sedimentological or stratigraphic data is available (such as studies based on thin-sections only), but also able to test and validate correlation schemes derived from sedimentological or stratigraphic data where doubts exist; and (2) palaeoenvironmental and depositional characterisation, being able to validate and propose depositional schemes and identify proximal to distal tendencies which are not always obvious from the sedimentological analysis.

The Permian-Triassic Transition

Stratigraphic, facies, lithological and palaeoecological analysis suggests that there is no major stratigraphic gap between the KS2 and the KS3 stratigraphic intervals, and hence no major stratigraphic gap between the end Permian and the Early Triassic. This is based on the following observations and interpretations:

  • (1) The transgression onto the platform starts at the top of KS3b where increasing higher energy grainy facies occur – suggesting the platform top was becoming more open before the Permian-Triassic Boundary. This is confirmed by the biofacies analysis which suggests a major transgression and opening of the platform system prior to the end Permian.

  • (2) Paralleling the pre-Permian-Triassic appearance of higher energetic grainy facies is the dolostone to limestone transition which shows that the transition from dolomite-dominated platform restricted settings to more open limestone facies also occurs before the Permian-Triassic Boundary.

  • (3) In the numerous subsurface sections, and outcrop investigations in the Zagros, no evidence is seen for a major exposure event between the Upper Dalan (Late Permian) oolitic and bioclastic grainstones and the Kangan (Early Triassic) microbially-influenced sands and microbialites – this contact is conformable.

  • (4) There is no major unconformity/disconformity or stratigraphic surface associated with the Permian-Triassic Boundary and the extinction of Permian fauna occurs within a grainstone body.

  • (5) The analysis suggests there is in fact a low-order transgression which started at the base of the KS3b (base Cycle II) and continued through to the KS2 (upper Cycle II), and hence a transgression across the Permian-Triassic Boundary.

The faunal analysis shows that the Permian Fauna Extinction (PFE) event generally occurs within a strongly calcite cemented and microbially mediated ooid grainstone rich in intraclasts (F16 facies) in the lower part of the KS2 sequence. Above the PFE event is a Permian azoic interval, which can vary slightly in thickness from 1 to 3 meters. This is followed by the Triassic faunal recovery and associated with the Early Triassic thrombolitic microbial event.

In the Zagros area the PFE occurs within pyrite-bearing muds under strongly anoxic influence. The outcrop data also show a similar pattern with a thin azoic interval occurring between the last Permian taxa and the first Triassic taxa. No major surface is seen in the Zagros outcrops linked to major SB nor PTB. Indeed there is a general muddying (deepening-upwards) from the lower Cycle II (KS3a) to the upper Cycle II (KS2a).

From the biostratigraphic analysis the Upper Dalan Member is considered to provide a complete record of the Lopingian (Late Permian) without any major gap just below the Permian-Triassic Boundary. The KS4 sequence is considered as Wuchiapingian (Dzhulfian), the KS3 as Changhsingian (Dorashamian) and the base of KS2 as latest Changhsingian.

The facies, stratigraphic and biostratigraphic analysis suggests there is a low (third-) order transgression between upper K3 and the K2. These represent the ‘Permian-Triassic oceanic event’ and are located in the late third-order TST. This is consistent with recent studies which suggest that the end Permian extinction event is associated with global transgression and the encroachment of anoxic waters.

The global Permian extinction event has also been attributed to massive gas hydrate (CH4) degassing into the ocean and promoting extensive syn-sedimentary sea-floor cementation as a result of the highly carbonate supersaturated seawater (Heydari and Hassanzadeh, 2003). These syn-sedimentary sea-floor cements, which according to Heydari and Hassanzadeh (2003) mark the precise PTB appear to be the lateral equivalent to the F16 facies which are characterised by coarse calcite isopachous fibrous cements and reworked intraclasts. As with the Heydari and Hassanzadeh (2003) cements, the first appearance of the F16 facies marks the precise PTB event in this study.

The 13C/12C data showed Permian values of 3 to 4 drop gently up to the top of the Permian interval, and then more rapidly up to the negative minima associated with the Permian-Triassic Boundary. The minima is reached within the thrombolite level with a value of –0.9. Once the 13C/12C data is calibrated on: (1) extinction of most benthic taxa; (2) the facies distribution; and (3) the 13C/12C minima event found in the thrombolitic bed, it is seen that there is an interval which is still Permian (latest Changhsingian), but devoid of any Permian benthic microfossils (due to the end Permian mass extinction event). The δ18O minima coincides with the appearance of the F16 facies and hence also the last of the Permian taxa in this locality.

A number of Upper Khuff logs which have Spectral Gamma Ray wireline log data show major shifts at or around the Permian-Triassic Boundary (K3–K2). At this level there is a significant drop in Uranium across the boundary in both grainstone facies and muddy facies, and hence independent of local facies conditions. This Spectral Gamma Ray ‘Uranium event’ can be easily correlated on the wells for which data is available. High-resolution core calibration of the Spectral Gamma Ray event shows that it does not relate to a major stratigraphic surface but correlates with the end of the Permian fauna and the start of the early cemented microbially-mediated grainstones with lithoclasts (F16 facies), just before the thrombolites.

The precise relationship between the appearance of the F16 facies with: (1) the δ18O isotopic minima; (2) coincidence of the appearance of the F16 facies with the last Permian fauna; (3) appearance of the F16 facies and spectral gamma ray signature; (4) the correlation with the syn-sedimentary sea-floor cements which mark the precise PTB; and (5) its association with the Early Triassic thrombolites, suggests that this facies is a key sedimentological units for the recognition of the end Permian extinction and the following Permian-Triassic Boundary. This facies is present on all the Permian-Triassic platform successions seen in the area (Saudi Arabia, Abu Dhabi, Qatar and Iran), even when the more classic thrombolite unit is very poorly developed or absent (such as in Abu Dhabi). This illustrates the regional and reservoir correlative value of identifying this facies.

Summary

The depositional and stratigraphic interpretations based on the integration of the sedimentological, stratigraphic, wireline log, palaeontological and geochemical data has provided a consistent sedimentological and sequence stratigraphic framework for other studies on the Upper Khuff in the Arabian Plate area. The study also provides a coherent geological framework for Upper Khuff regional exploration and prospect evaluation, but also for reservoir-scale diagenetic and reservoir quality studies.

The results also demonstrate that the precise nature of the Upper Khuff system can only be understood by an integrated approach. The resulting models developed reconcile the sedimentological, stratigraphic, palaeoecological, geochemical and petrophysical data and provide models for the stratigraphic architecture and facies distribution. Moreover the study illustrates that over-relying on a single discipline (such as sequence stratigraphy or biostratigraphy) no matter how “convincing” the data may appear, can lead to miscorrelation and misinterpretations of reservoir units.

ACKNOWLEDGEMENTS

The authors would like to thank RIPI-NIOC, POGC and TOTAL EP for permission to publish this study. The authors would also like to express thanks to the following people for their valuable contribution to various aspects of this study: P. Masse (biostratigraphy and palaeoecology), F. Walgenwitz (geochemical analysis), M. Sudrie and P. Thiry-Bastien (core and thin-section descriptions), D. Vachard (determination of foraminifera), A. Meyer, A. Roumagnac and C. Laporte-Gala (correlations), S. Rouyer (descriptions of Zagros thin-sections), C. Javaux, M. Lescanne, C. Fraisse and F. Vieban (project and management support), and M. Namati, M. Moradpour, A. Ghaemi, A. Khalili and N. Efteckhari for much discussion on various aspects of the Dalan and Kangan system, and assistance during fieldwork. Constructive reviews by M. Al-Husseini, D. Vaslet, Y.-M. Le Nindre and L. Angiolini were also much appreciated, as is the excellent editorial work of the GeoArabia team.

ABOUT THE AUTHORS

Enzo Insalaco holds a PhD in Carbonate Sedimentology from the University of Birmingham. He is presently Senior Sedimentologist in the Carbonate Sedimentology Group for the Total group in Pau, and Project Leader for Total’s Khuff (Dalan/Kangan) Synthesis, a regional to reservoir-scale multidisciplinary project integrating regional seismic interpretations, structural analysis, sedimentology, diagenesis and poroperm analysis. Enzo has worked on the sedimentology, stratigraphy, reservoir characterization and modelling of a number of Middle East petroleum formations. From 1998 to 2001 he was based at Total’s Geoscience Research Centre in London and Project Leader on the 3-D Sedimentary Modelling Project, aimed at better constraining 3-D reservoir geomodels, at high resolution, through better integration of sedimentological concepts. From 1996 to1999 Enzo was a Lecturer in Sedimentology at the University of Birmingham, England. He has published numerous scientific articles and edited a book on carbonate platforms.

Enzo.Insalaco@total.com

Aurélien Virgone joined the Total Group in 1998. He received his PhD degree in Carbonate Sedimentology in 1997 from Marseille University (France). Aurélien has worked for five years on the sedimentology and reservoir characterisation of the Khuff (Kangan/Dalan) Formation and the Thamama Group. He has acquired a good knowledge of the outcrops in the Zagros Mountains. After two years as a Wellsite Geologist at the rig site (deep offshore Angola) and one year as a Log Analyst, he is presently Senior Geologist in the Carbonate Sedimentology team at Total CST.

aurelien.virgone@total.com

Bruno Courme graduated from the Ecole Polytechnique and from the Paris School of Mines. He received a Masters degree in Geological Sciences from the University of Texas at Austin in 1999. Bruno joined Total in 2000 and worked as a Carbonate Sedimentologist for three years, mostly on CEI and Middle East subjects related to Paleozoic environments. He then moved on to geological operations, first on the rigsite in the North Sea, then office-based for Total subsidiary in Nigeria, until 2005. He recently joined the Deep Offshore Exploration team of Total in Nigeria.

bruno.courme@total.com

Jérémie Gaillot received his MSc in Biodiversity of present-day and fossil ecosystems in 2002 from the University of Science and Technology, Lille, France. After his studies, he participated as a Junior Geologist in a synthesis project on the biostratigraphy of the Upper Dalan Formation of Iran. In June 2003, Jérémie began his PhD on the Biostratigraphy and Palaeoecology of the Khuff Formation of the Arabian Platform at the University of Lille 1, funded by Total. The main focus of his PhD is the sequential foraminiferal biostratigraphy of epeiric platforms and the mechanisms of mass extinction at the Permian-Triassic Boundary.

Jeremie.Gaillot@total.com

Mohammad Reza Kamali is Director of the Center for Exploration and Production Studies and Research at the Research Institute of Petroleum Industry, Iran. He received BSc and MSc degrees in Geology from the University of Mysore in 1982 and 1984, respectively. After joining the National Iranian Oil Company in 1985 as Senior Geologist, Mohammad continued his PhD in Petroleum Geology at Adelaide University, Australia and graduated in 1996. His research interests are petroleum geochemistry and reservoir geology. Mohammad is a member of the AAPG, SPE, EAGE, Geological Society of Iran and Petroleum Engineering Society of Iran.

Ali Moallemi is a PhD candidate in the Department of Geology at Shahid Beheshti University, Tehran, Iran. He obtained an MSc in Sedimentology from the Azad University of Tehran in 1994. Ali is presently a Senior Research Scientist at the Research Institute of Petroleum Industry, working on carbonate reservoir characterization research. His major interests include sequence stratigraphy, depositional systems, and diagenesis of carbonates.

moallemisa@ripi.ir

Masoud Lotfpour obtained a PhD from the Department of Geology at Shahid Beheshti University, Tehran, Iran in April 2005. His thesis was on the sequence stratigraphy and biostratigraphy of Permo-Triassic successions in the Zagros area. He also holds an MSc (1997) in Sequence Stratigraphy from the Tarbiat Moallem University of Tehran. Masoud is a Reservoir Geologist at the Research Institute of Petroleum Industry, working on carbonate reservoir characterization research. His major interests include sequence stratigraphy, sedimentary environment and biostratigraphy.

lotfpourm@yahoo.com

Saeed Monibi has an MSc (1992) in the biostratigraphy of Permian successions in the Alborz area from the Department of Geology at Azad University, Tehran, Iran. He is a Micropalaeontologist at the Research Institute of Petroleum Industry, working on carbonate reservoir characterization research. His major interests include biostratigraphy and sedimentary environment.

monibis@ripi.ir