Clastic Intrusion at the Base of Deep-water Sands: A Trap-forming Mechanism in the Eastern Mediterranean
Published:January 01, 2007
- PDF LinkChapter PDF
Jose Frey-Martínez, Joe Cartwright, Ben Hall, Mads Huuse, 2007. "Clastic Intrusion at the Base of Deep-water Sands: A Trap-forming Mechanism in the Eastern Mediterranean", Sand Injectites: Implications for Hydrocarbon Exploration and Production, Andrew Hurst, Joseph Cartwright
Download citation file:
Three-dimensional seismic data from the continental margin offshore Israel (eastern Mediterranean) show several large-scale mounded structures interpreted to be clastic intrusions. The structures are confined to the Zanclean (early Pliocene) and lower Gelasian (late Pliocene) intervals and restricted to an area of 40 × 20 km (24 × 12 mi) along the Afiq submarine canyon, a former depositional fairway of Oligocene age. Most of the features are circular to oval in plan view, range from 0.5 to 2 km (0.3 to 1.2 mi) in diameter at their base, and are flanked by kilometer-scale depressions interpreted as regions of sediment depletion. In cross section, the mounds are as much as 400 m (1300 ft) in height and have flank dips of as much as 20–25°. The largest structures may reach as much as approximately 0.75 km3 (0.17 mi3) in volume and represent economic hydrocarbon reservoirs.
Well data and direct hydrocarbon indicators show that the mounds are predominantly composed of gas-saturated sandstones along their flanks and crests, whereas their center is heterolithic. Petrophysical interpretation indicates the presence of chaotic and remobilized sediments in the core of the structures. The relationships of the mounds to the overburden exhibit both depositional and deformational geometries (e.g., onlap, forced folding). The proposed model for their formation is hydraulic jacking up of the overburden by forceful vertical and lateral intrusion of clastic sediments during shallow burial. Several episodes of intrusion alternated with the deposition of fine-grained clastic sediment during the Zanclean and early Gelasian to create the complex structures presented in this chapter. The suggested model has implications for the understanding of the trapping mechanism and reservoir properties of the mounded structures and needs to be incorporated in exploration and production strategies.
In recent years, three-dimensional (3-D) seismic data have revealed many examples of clastic sediment remobilization reconfiguring deep-water sands into a complex variety of mounded and intrusive geometries (e.g., Jenssen et al., 1993; Brooke et al., 1995; Dixon et al., 1995; Lonergan and Cartwright, 1999; Lonergan et al., 2000; Davies, 2003; Hurst et al., 2003; Huuse et al., 2003). The dominant processes invoked in this remobilization are the liquefaction, fluidization, and intrusion of the coarse reservoir facies. Understanding the processes and products of large-scale remobilization is critical for the exploration and production of hydrocarbons because remobilization exerts significant control on the geometry (Dixon et al., 1995; Lonergan et al., 2000), properties (Duranti et al., 2002), and intrafield connectivity (Guargena et al., 2002) of reservoir sands.
In this chapter, we describe a series of kilometer-scale, steep-sided, convex-upward, moundlike features from offshore Israel, which we argue are related to clastic sediment intrusion during shallow burial. The structures consist of a complex arrangement of isolated features and compound arrays, forming a sinuous trail above the axis of the Oligocene Afiq submarine canyon. In many cases, the mounded features are sufficiently large to form economic hydrocarbon reservoirs hosting some of the largest gas fields discovered to date in the region. Several of these structures have been targeted by exploration drilling, proving, in many cases, the presence of tens-of-meter-thick sandstones with excellent reservoir quality (30% porosity, as much as 12 d permeability and 81% average net to gross). Such reservoirs, however, represent a challenge to the effective recovery of hydrocarbons because of their geometric variability, lithological heterogeneity, and intricate reservoir compartmentalization.
The aim of this chapter is to describe and unravel the genesis of the mounded features, with special emphasis on their implications for hydrocarbon exploration and production. The chapter commences with an overview of the geological setting and a description of the stratigraphic context of the Afiq submarine canyon. The main arguments of the chapter are subsequently developed through detailed seismic and lithological analysis focused on the interval of the reservoir affected by clastic intrusion. Exceptional seismic imaging provided by the 3-D seismic data and calibration with borehole data allow an accurate analysis of the external and internal geometries of the mound structures. The study is concluded by presenting a model for the origin and evolution of the mounds and their implications for hydrocarbon prospectivity. With substantial accumulations of proven recoverable natural gas remaining to be produced and the potential for new gas discoveries offshore Israel, an in-depth understanding of the mechanisms forming the mounded structures presented in this chapter, and their impact on reservoir quality and architecture, will be critical to an efficient assessment of future exploration and production in the region. Although the examples illustrated here are specific to the study area, the ideas presented may find applicability in deep-water sandstone reservoirs in other parts of the world.
The study area is located in the eastern Mediterranean on the continental margin offshore Israel (Figure 1). This margin occupies an active tectonic and complex stratigraphic setting because of its location at the zone of interaction between the Anatolian, African, and Arabian plates (Figure 1). A complex geological history, dominated by the opening and subsequent closure of the Tethyan Ocean, marks its overall tectonic evolution from the Late Triassic onward (Garfunkel and Derin, 1984; Ben-Avraham, 1989; Robertson et al., 1996). Since then, intense tectonic activity, including plate convergence during the Late Cretaceous (Syrian arc system), rifting during the middle Cenozoic, and subsidence since the Miocene, has resulted in a complex structural setting.
The mounded structures are confined to the lower Yafo Formation (lower Pliocene; Figure 2) and are restricted to an area underlain by the Afiq submarine canyon (Figure 3, Figure 4). The Afiq submarine canyon is one of a series of canyons (e.g., Afiq, el-Arish, and Ashdod) that were incised into the Cretaceous to Eocene shelf-edge strata of the Israeli continental margin during the Oligocene (e.g., Druckman et al., 1995). It extends for more than 120 km (74 mi), trending in a basinward direction northwest from Beer Sheva to Gaza and into the offshore area (Druckman et al., 1995) (Figure 3). Within the study area, the Afiq submarine canyon ranges from 6 to 12 km (3.7 to 7.4 mi) in width and reaches its overall maximum depth of about 500 m (1640 ft) relative to its shoulders (Figure 3, Figure 5).
During most of the early Miocene, submarine erosion or nondeposition prevailed within the Afiq submarine canyon (Druckman et al., 1995). This was followed by the deposition of pelagic marls and mud-rich debris-flow deposits (Bet Guvrim and Ziqim formations) in the middle and late Miocene (Druckman et al., 1995) (Figure 2). During the Messinian salinity crisis (latest Miocene), a relative sea level fall of as much as 800 m (2624 ft) below the rim of the Afiq submarine canyon resulted in the formation of the evaporitic deposits of the Mavqiim Formation (Druckman et al., 1995). Following the Messinian salinity crisis, marine deposition resulted in the accumulation of the Yafo Formation, which includes an important deep-water turbiditic reservoir sandstone: the Yafo Sand Member (Figure 2). The distribution of the Yafo Sand Member in a base-of-slope position is consistent with core-based interpretation of this unit as a submarine fan (Figure 3). These sediments were deposited during a period of increased clastic input in the first stages of refilling of the Mediterranean after the Messinian salinity crisis. From the middle Pliocene onward, sedimentation of large amounts of fine-grained siliciclastics and carbonates of the Yafo Formation built a progradational succession that buried both the Afiq submarine canyon and the preexisting slope (e.g., Druckman et al., 1995; Buchbinder and Zilberman, 1997) (Figure 4). Clay-rich marls passing upward to thinly interbedded sandstones and claystones were deposited over the slope and basin areas, forming a strongly aggradational system with sigmoidal clinoforms linking shelf to slope (Figure 2, Figure 4). Concurrently, several episodes of large-scale slumping and gravitational tectonics occurred (e.g., Almagor, 1980, 1984, 1986; Garfunkel, 1984; Garfunkel and Almagor, 1985, 1987; Frey-Martinez et al., 2005).
Database and Methodology
Three-dimensional seismic and well data represent the main source of information for this study. The Levant and med-Ashdod surveys comprise the 3-D seismic data (Figure 3). The two surveys cover about 2500 km2 (965 mi2), extending from shelf to deep water. They were acquired in 2000 with a crossline interval of 6.25 m (20 ft) and a line spacing of 25 m (82 ft). The final data for the two surveys were defined on a 12.5 × 12.5-m (41 × 41-ft) grid with 6400 bin cells/km2 after processing. The seismic data are approximately zero phase with Society of Exploration Geophysicists normal polarity, meaning that an increase in acoustic impedance is represented by a positive amplitude. The dominant frequency varies with depth, but it is approximately 50 Hz at the level of interest (2500–1750 ms two-way traveltime [TWT]). The data have been time migrated, and the quality is regarded as excellent. When calculating the resolution, an average seismic velocity of 2000 m/s (6600 ft/s) has been assumed within the level of interest. This velocity is derived from the check-shot measurements made in the Gaza Marine-1 well. Vertical (λ/4) and lateral (λ) resolutions are estimated to be about 10 and 40 m (33 and 130 ft), respectively. In this chapter, all stratal thicknesses and depths given in seismic time are in milliseconds TWT.
The well data come from nine hydrocarbon exploration wells located in the study area (Figure 3). The data consist of petrophysical well logs (gamma ray, sonic, velocity, resistivity, and checkshots), along with paleontological and biostratigraphic information. Unpublished commercial stratigraphic reports, mainly based on cuttings analyses, were also available from seven of the exploration wells (Nir-1, Or-1, Or South-1, Noa-1, Yam West-1, Gaza Marine-1, and Gaza Marine-2).
The present study is based on the seismic-stratigraphic and facies analysis of the sedimentary intervals of late Miocene–early Pliocene age. A seismic-stratigraphic framework for the mounded structures in these intervals was established by two different approaches. First, a combination of seismic and lithostratigraphic analysis was employed to provide a detailed stratigraphic context. The interpretation of the seismic data was calibrated with well data by integrating wireline logs and core sedimentological interpretations with biostratigraphic information. Lithostratigraphic data from the Nir-1 well (drilled through a mounded structure) were extensively used for detailed calibration of the lithologies in the critical intervals above, below, and within the mounds. The second approach employed accurate 3-D mapping and seismic attribute analysis to investigate the external and internal seismic geometry, fabric, and lithology of the mounded structures. This approach involved the analyses of timeslice and coherence volumes.
Stratigraphic Context of the Mounded Structures
The seismic-stratigraphic analysis of the mounded structures resulted in the identification of three seismic units in the Oligocene to lower Pliocene successions (units 1–3 in Figure 2, Figure 5). These units are primarily defined based on their seismic facies, thickness variations, and spatial reflection patterns. The boundaries between the seismic units are identified based on their physical discontinuity, i.e., onlap, downlap, toplap, and truncation terminations. Checkshot surveys and calibrated wire-line logs provided direct well ties with the interpreted units, which have been correlated with the lithostratigraphic framework established for the Afiq submarine canyon by Druckman et al. (1995).
Unit 1 (Canyon Fill)
Unit 1 is bounded at its base by a conspicuous erosional unconformity at which it is possible to distinguish clear reflection terminations (base of the Afiq submarine canyon; Figure 5). The upper boundary is coincident with a very irregular and discontinuous seismic reflection of moderate to low amplitude (Figure 5). Unit 1 is only present within the Afiq submarine canyon, where it forms a northeast–southwest-trending depocenter 15 km (9 mi) long by 7 km (4.3 mi) wide, with as much as a 500-m (1640-ft) thickness of sediments. Internally, unit 1 is composed of subhorizontal, medium- to low-amplitude, continuous, and aggradational seismic reflections that onlap the canyon margins (Figure 5).
Petrophysical log data and completion reports show that Unit 1 is principally composed of gray-brown, glauconitic and pyritic, pelagic marls interbedded with gray-brown, soft to firm, deep-marine claystones, limestones, and medium-grained turbiditic sandstones. Unit 1 corresponds to two lithostratigraphic formations as defined in the available wells: the Bet Guvrim Formation (Oli-gocene to middle Miocene) and the Ziqim Formation (middle Miocene) (Figure 2). Lack of accurate dating does not allow clear recognition of the boundary between them and makes correlation of the lithostratigraphic markers with the seismic stratigraphy highly problematic. The completion log from Nir-1 describes the upper parts of the Ziqim Formation as an approximately 400-m (1300-ft) succession of 10–15-m (32–49-ft)-thick intervals of hemipelagic shales interbedded with chert-bearing conglomerates and very coarse- to medium- grained turbiditic sandstones.
Unit 2 (Messinian Deposits)
Unit 2 is bounded at its top by an irregular, high-amplitude reflection that marks a basinwide erosional unconformity (reflector M) caused by subaerial exposure during the Messinian salinity crisis (see Ryan et al., 1971) (Figure 5). Unit 2 forms a thin and uneven veneer that irregularly covers unit 1 and pinches out with onlap against the flanks of the Afiq submarine canyon (Figure 5). Internally, Unit 2 comprises discontinuous to transparent, low-amplitude seismic reflections that are locally interbedded with continuous and moderate-amplitude reflections.
Unit 2 is correlated to the Mavqiim Formation in several exploration wells in the study area (Figure 2). The Mavqiim Formation in the Afiq submarine canyon is interpreted as the time equivalent of thick deposits of evaporites that were deposited over the former abyssal plain during the Messinian salinity crisis (Druckman et al., 1995). According to Druckman et al. (1995), the Mavqiim Formation consists of thin interbeds of compacted nodular anhydrite interbedded with medium- to dark-gray and moderately firm claystones, limestones, and sandstones (Figure 2). The completion log from Nir-1 describes the Mavqiim Formation as loose and poorly sorted sandstones interbedded with light-medium gray claystones and soft powdery limestones. No evidence from the seismic or well data exists for any significant accumulations of primary evaporites at this stratigraphic level.
Unit 3 (Lower Yafo Formation)
The upper boundary of unit 3 corresponds to a high-amplitude seismic reflection that marks the top of the early Pliocene sedimentary fill in the Afiq submarine canyon (Figure 5). Unit 3 consists of an approximately 500-m (1640-ft)-thick interval that is part of a prograding and aggrading slope wedge that onlaps and down-laps the lateral flanks of the Afiq submarine canyon (Figure 5). All the wells in the study area have penetrated unit 3, which is correlated to the lower Yafo Formation and dated as lower Pliocene (Figure 2).
The basal part of Unit 3 consists of a package of high-amplitude, continuous, subparallel, and high-frequency seismic reflections with an approximately constant thickness of some 70 m (229 ft) (Figure 5). It has been penetrated in six exploration wells, and it is informally termed the Yafo Sand Member (lower Pliocene; Figure 2). Three-dimensional mapping calibrated with borehole data indicates that the Yafo Sand Member is geographically restricted to the offshore extension of the Afiq and el-Arish submarine canyons (Figure 3). It consists of gray, soft to firm marls interbedded with siltstones and sandstones overlain by a succession of as much as three thick units of very fine- to medium-grained, clean and well-sorted, thinly bed-ded turbiditic sandstones. Deposition of the Yafo Sand Member was strongly influenced by the underlying Afiq and el-Arish submarine canyons, which at that time still formed a deep depression along the continental margin and provided a conduit for fine-grained silici-clastics to be deposited onto the basin floor. According to well completion reports, the turbiditic sandstones at the top of the Yafo Sand Member were supplied from the southern Levant hinterland via the Afiq and el-Arish submarine canyons and accumulated in deep-water as base-of-slope submarine fans.
The basal part of unit 3 is overlain by a conformable interval of moderate- to low-amplitude, subparallel continuous and medium-frequency aggradational seismic reflections (Figure 5). This interval reaches as much as about 400 m (1300 ft) in thickness and is herein informally termed the Yafo Mudstone Member of early Pliocene age (Figure 2). The bulk of the Yafo Mudstone Member comprises deep-marine limestones, claystones, and siltstones that locally alternate with sandstones, limestones, and marls. The Yafo Mudstone Member was deposited during a major transgression that inundated the exposed upper Miocene shelf (e.g., Buchbinder and Zilberman, 1997).
General Appearance of the Mounded Structures
The mounded structures are defined in the 3-D seismic data by mapping the two seismic reflections corresponding to their lower and upper boundaries. The main key recognition criteria for the two boundaries are illustrated with reference to Figure 6. The lower boundary is defined as the base of the Yafo Sand Mem-ber as calibrated in well data. This boundary generally forms a relatively concordant, continuous, and flat-lying strong negative amplitude seismic reflection that underlies all the mounds (Figure 6). Three-dimensional mapping of this reflection is relatively straightforward because it is uncommonly deformed.
The upper boundary is herein defined as a surface of stratal discontinuity between the concordant seismic reflections of the underlying Yafo Sand Member and the convergent onlap terminations of the overlying Yafo Mudstone Member (Figure 6). The boundary is correlated with the top of the Yafo Sand Member based on calibration of the seismic data at the Nir-1 exploration well. In 3-D seismic data, it is seen as a convex-upward and moderately continuous surface overlying all the mounds (Figure 6). Mapping of this boundary is locally problematic because of significant variations in its seismic response. On the crest of the mounds, it generally corresponds to an easily traced, continuous, and strong negative seismic amplitude reflection. Toward the flanks, the reflection is more difficult to detect; in some places, it is represented by a weak trough; in others, it is absent (Figure 6). This is probably because of the steeply dipping nature of the flanks of the mounds and the lack of a strong seismic impedance contrast between the structures and the overburden. In this study, the upper boundary has thus been identified by mapping the stratal termination between the steeply dipping reflections dominant within the flanks of the structures and the continuous seismic reflections of the overburden. Following this approach, the boundary has been followed from the top of the Yafo Sand Member up the flanks of the mounds and across their crestal parts (Figure 6). This pick is consistent with the li-thology encountered in the Nir-1 well as further discussed later.
Based on mapping at the top of the Yafo Sand Member and the recognition that mounds are defined by localized structural culminations at this surface, 12 discrete structures have been identified in the med-Ashdod 3-D area. A time-structural map at this stratigraphic level (Figure 7) shows them as elliptical (maximum of about 1.2 ellipticity ratio) to circular culminations in plan view with diameters at their base of 0.5–3 km (0.3–1.8 mi). Some are isolated four-way dip closures; others coalesce to form arrays of as much as three linked bodies. Most of the structures form a southeast–northwest-trending arrangement coinciding with the axis of the underlying Afiq submarine canyon (Figure 3, Figure 7). The more elliptical features occur in the northwesternmost region of the canyon axis arrangement, whereas the near-circular ones appear in its central and southeast-ernmost parts (Figure 7). In detail, the larger elliptical mounds may be composed of several smaller (˜0.5 km [˜0.3 mi] diameter), aligned circular bodies (Figure 7). Their gross rock volume may individually reach as much as about 0.75 km3 (0.18 mi3).
The general architecture of the mounded structures is particularly clearly seen on a representative seismic profile along the axis of the Afiq submarine canyon (Figure 8). On this profile, the mounds are recognized as steep-sided domal features of dominant chaotic seismic facies that are in sharp contrast with the more continuous to transparent reflections of the overlying Yafo Mudstone Member. Their lower boundary varies from a relatively well-defined and undisturbed continuous seismic reflection below the flanks to an upturned seismic reflection below the cores (Figure 8). Whereas some of these upturned geometries may be attributed to overmigration of point diffractions, their spatial coincidence with the mounds, and the fact that they form antiforms and not crosscutting events, suggests that most of them correspond to velocity pull-up features. The upper boundaries of the mounds are consistently expressed as dome-shaped seismic reflections with flank angles of the order of 20–25° (Figure 8). This is a key observation regarding genesis because this range of flank angles is considerably higher than common depositional angles of the margins of deep-water silici-clastic bodies (Nichols, 1995). No evidence of erosion associated with the upper boundaries of the mounds is present, and their approximate heights range from 75 to 400 m (246 to 1300 ft) (Figure 8).
The internal parts of the mounds consist of two main seismic facies. The first corresponds to incoherent and chaotic reflections that are dominant in the cores of the structures (Figure 6, Figure 8). The second is seismically resolvable reflections that can be continuously traced along their flanks and crests (marked X in Figure 6). In plan view, the chaotic cores form near-circular features surrounded by the continuous reflections along the flanks. Such appearance is particularly clear on a flattened horizontal coherence slice across the Mari mound complex (MMC; Figure 9), an array of three mounds in the northwesternmost parts of the med-Ashdod 3-D area. The coherent reflections along the flanks of the mounds correspond to the Yafo Sand Member. This is clearly seen in the Alpha mound (AM), where the Yafo Sand Member can be correlated in 3-D as a package of approximately constant thickness, thus constraining its geometry as a dome-shaped feature that encases the chaotic core of the structure (Figure 8, Figure 10). Numerous direct hydro-carbon indicators (DHIs) are seen within the mounds. They mainly correspond to crosscutting, high-amplitude, acoustically hard seismic events that are confined to their flanks (Figure 6, Figure 8). No evidence of DHIs has been found within the cores of the structures, which strongly suggests poorer reservoir qualities in these parts.
Vertical zones of chaotic seismic reflections are repeatedly observed below the mounds (Figure 8). These are restricted to the areas overlain by the structures and extend in depth down to the Paleocene and Eocene intervals (Figure 8). The chaotic zones could be interpreted as some form of fluid-migration pathway from deeper stratigraphic levels (i.e., Paleocene-Eocene). Gas chimneys, for example, resulting from the vertical movement of gas-rich fluids, are known to have similar seismic expressions (Løseth et al., 2003). In the study area, however, such an origin cannot be invoked given that that there is no evidence of increased gas content below the mounds. Well Nir-1, which penetrated one of such vertical chaotic zones, had low gas readings and no indication of any other type of hydrocarbons. In addition, the fact that the vertical chaotic zones are limited to the areas overlain by the mounds suggests that they are more likely the result of ray bending and scattering caused by high-velocity contrasts and overmigration of point diffractions.
The sedimentary overburden of the mounds is defined here as part of the Yafo Mudstone Member and is restricted upward by the upper limit of visible deformation associated to the structures. On seismic profiles (Figure 8), the overburden appears as an approximately 350-m (1148-ft) interval of continuous to discontinu-ous reflections that terminate sharply against the upper boundary of each of the mounds. The lateral relationships of the overburden to the mounds are highly complex and variable. In some areas, they are interpreted as onlap and can be conveniently assigned a depositional origin; however, evidence of deformational processes (i.e., folds and faults) is also observed. Numerous normal faults, for instance, appear as clusters of concentric and radial features that surround the crests and flanks of the mounds (Figure 11). In addition, thickness variations are seen in those parts of the overburden adjacent to them (Figure 8, Figure 12). The detailed geometry of such variations is critical to assess the mode and timing of formation of the mounded structures and will be discussed in more detail later.
Mapping at the top of the Yafo Sand Member shows the presence of bowl-shaped marginal depressions flanking the mounds. The depressions are elongate and crescentic in plan view and are particularly well defined near the major mounds (marked D in Figure 7). The largest depressions are as much as about 20 km2 (7 mi2) in area (mean length 6 km [3.7 mi], mean width 4 km [2.5 mi]), 150 m (93 mi) in depth, and may individually reach as much as 0.3 km3 (0.07 mi3) in gross volume (calculated above a local undeformed datum). On seismic profiles, they are bowl shaped, with marginal flank dips of between 3 and 5° (D in Figure 12). No evidence of erosion (i.e., reflection truncations) associated with them is present. The thickness of the Yafo Sand Member does not vary across the depressions, which implies that no material has been removed from this interval and points to a mechanism involving the depletion of sediment from deeper stratigraphic levels (i.e., Afiq submarine canyon). The reflection configurations and the thickness variations of the overlying Yafo Mudstone Member are vertically concordant with the geometry of the depressions (Figure 12). This suggests downwarping of the overburden above them and is consistent with a sediment volume loss from the Afiq submarine canyon.
Detailed Description of the Mounded Structures
Nir Mound Complex
In this section, a representative case study of the mounded structures (Nir mound complex; NMC) is analyzed in detail. The NMC comprises two coalesced mounds (NMC-1 and NMC-2) located in the central parts of the med-Ashdod 3-D area and lies directly over the axis of the Afiq submarine canyon (Figure 7). The NMC has a marked elongated planform geometry trending in a northwest–southeast direction. It is 4 km (2.5 mi) long, and its width ranges from 1.5 to 3 km (0.9 to 1.8 mi) (Figure 7). The total structure covers an area of about 4 km2 (1.5 mi2), has a maximum height of as much as 375 m (1230 ft), and involves an overall gross rock volume of about 0.45 km3 (0.1 mi3). The flanks of the two component mounds are almost symmetric, with slopes ranging between 20 and 25° (Figure 13).
Seismic Character of the NMC
On seismic dip sections, the NMC appears as two interlocked, approximately symmetric-upward conical-shaped bodies (NMC-1 and NMC-2; Figure 8). The lower boundary of both mounds is a moderately deformed, continuous, high-amplitude reflection (Figure 8, Figure 13). The deformed appearance of this boundary increases below the cores of the structures, where it forms small-scale antiforms (Figure 12). These geometries are interpreted as seismic velocity pull-up features. The upper boundary of the NMC corresponds to a discontinuous domelike moderate- to low-amplitude seismic reflection, against which the reflections of the overburden terminate by deformed onlap (Figure 8, Figure 12).
The internal seismic reflection character of the NMC is dominated by low-amplitude and chaotic seismic facies (Figure 8, Figure 12). The chaotic pattern is locally broken by short and disconnected blocklike segments that appear as parallel-bedded, high-amplitude reflections (X in Figure 13). Such reflections can only be traced for short distances (<1 km; <0.6 mi) along the flanks and crest of the NMC and dip systematically parallel to its outer boundary (Figure 13). The contact between the chaotic seismic facies and the blocklike reflections is significantly steep (about 15–20°) and irregular (Figure 13). It is difficult to explain such a geometrical relationship by any known depositional process simply because the observed contact is too steep to be of a clastic depositional origin. Instead, we prefer the interpretation that such a contact is the result of some type of forceful intrusion in the NMC. According to this interpretation, the chaotic reflections represent the intruded material, whereas the blocklike reflections are remnants of the primary stratification that has been broken and distorted during the intrusion. From measurements made on serial depth-converted seismic sections, we estimate that as much as approximately 50% of the total gross rock volume in the NMC corresponds to the intruded material (Figure 13). This observation helps to constrain the origin of the mounds because it clearly excludes a depositional origin and forms the basis for inferring a dominant role for intrusion.
No clear evidence of DHIs is found within the NMC, although a gas-water contact has been identified petrophysically (see below). Local anomalously high seismic amplitudes are the only geophysical hydrocarbon indicators identified within the mounded complex. Most of the hydrocarbon indicators are acoustically soft events that represent a decrease in the acoustic impedance. These events are mainly located in the flanks and the crest of the NMC and are associated with the previously described isolated blocklike reflections (X in Figure 13). Accurate 3-D mapping of these soft seismic events is ambiguous because of their limited continuity and intricate geometry. Several of them have, however, been penetrated by exploration wells and calibrated to thin (about 50-m [164-ft]) gas-charged sandstones. A more detailed well-log calibration of these events is undertaken in the following subsection.
Well-log Calibration of the NMC Lithology
The Nir-1 exploration well was drilled through the NMC-1. The well penetrated the crest of the NMC-1 at 1649 m (5410 ft), where it intersected an interval of soft seismic events (Figure 14). The top of this interval is not immediately obvious from the conventional well logs. Instead, it corresponds to weak inflections in the gamma-ray log, indicating low radioactive contrasts with the shale-rich intervals of the overburden. The base of the interval is calibrated to a box-shaped gamma-ray log motif (arrowed in Figure 14). Such gamma-ray response suggests that this interval consists of approximately 140 m (459 ft) of interbedded sandstones and claystones that become more sand-rich toward the base. The completion log describes the sandstones as highly fractured, gas charged, and moderately cemented. Analysis of the azimuth and dip trends in the NMC-1 indicates that at approximately 1649 m (5410 ft), there is a small angular change in dip and a clear variation in azimuth from a northwest direction dominant within the overburden to a south-southwest direction throughout the crest of the structure (Figure 14). This change in the azimuth direction correlates with the top of the NMC-1 and is interpreted as an unconformity.
Beneath the interval of soft seismic events, the Nir-1 well penetrates an approximately 185-m (606-ft)-thick unit of claystones interbedded with loose sands and lightly carbonate-cemented sandstones. This unit correlates on seismic data with the previously described chaotic and low-amplitude facies in the core of the mound (C in Figure 13). The claystones encountered are gray to dark gray, plastic, slightly calcareous, and pyritic. The sandstones are clear to yellow, composed of fine to coarse, subangular to subrounded quartzitic grains. Traces of chert and limestone are present. Dip and azimuth analyses confirm the highly chaotic nature of this part of the NMC-1. The bedding has dips as great as 75°, and the azimuth directions vary from dominant northwest and northeast trends in the approximately 1694–1774-m (5557–5820-ft) interval to a wide spectrum of directions in the central and lowermost parts of the NMC-1 (Figure 14). We interpret these variations in the azimuth direction as the presence of unconformities, faults, and disrupted material. No sedimentary structures (e.g., lamination, ripples, or cross-stratification) are observed. Detailed velocity anal-ysis demonstrates that the core of the NMC-1 has anomalously higher velocities (about 2950 m/s [9678 ft/s]) than the surrounding succession (about 1500–1800 m/s [4921–5905 ft/s]). This supports the above-suggested origin for the velocity pull-up features below the core of the mounds. A gas-water contact has been identified at 1951.6 m (6402.8 ft) (Figure 14), giving a gross gas column of 302.6 m (992.7 ft). The reservoir properties in this interval are relatively poor, with an average net to gross and porosity of 0.56 and 7.5%, respectively. The base of the NMC-1 is penetrated at 2026 m (6646 ft), where the upper Miocene Mavqiim Formation is encountered (Figure 14).
Micropaleontological analyses reveal that alloch-thonous fossil specimens from Cretaceous to upper Miocene stratigraphic levels dominate the microfaunal recovery in the NMC-1. The degree of allochthonous nannofossils in the Nir-1 well is between 44 and 78%, which is consistently higher than those recorded from other wells in the area (about 20%). Microfaunal reworking averages 40–50%, which is considerably in excess of anything recorded in any other well where the content of reworked fossils is limited to a few individual specimens. The taxa composing the reworked assemblages in the Nir-1 well in the NMC-1 contrast markedly with those recorded from other wells in the area. These variations are most obvious in the microfauna: between 1649 and 1770 m (5410 and 5807 ft), where early and middle Eocene foraminifera (Acarinina spp., Morozovella spp., Nummulites spp., and operculi-nids) and Upper Cretaceous planktonic and benthic spe-cies (Bolivinoides draco, Gavelinella spp., Globotruncana ventricosa) are encountered. Reworking of upper Paleo-cene deposits is indicated by the planktonic Morozovella velascoensis that occurs between 1728 and 1980 m (5669 and 6496 ft) (Figure 14).
A similar pattern is repeated in the nannofossil record, which has a major influx of reworked samples from 1671 m (5482 ft) to the base of the NMC-1 (Figure 14). The dominant specimens are primarily of Late Cretaceous age (Arkhangelskiella cymbiformis, Ceratolithoides aculeus, Quadrum trifidum, Reinhardtites levis, and Zeugr-habdiotus compactus) and Eocene taxa (Discoaster kuep-peri, Discoaster barbadiensis, and Helicosphaera seminulun). Oligocene to early Miocene taxa (Sphenolithus belemnos, Cyclicargolithus abisectus, and Reticulofenestra scissura) are recorded in significant numbers.
Timing of Formation of the Mounds
The timing of formation of the mounds is recorded by their geometric relationships to the sedimentary overburden (Yafo Mudstone Member). To define the relationships, three representative seismic reflections were mapped within the overburden across the med-Ashdod 3-D area (horizons A–C in Figure 8). These reflections were dated by correlation with the Nir-1 exploration well. Horizon A is positioned in the lower part of the overburden and is dated as late Zanclean (early Pliocene) (Figure 8). It correlates with a medium-amplitude, subcontinuous seismic reflection that is generally concordant with the underlying top of the Yafo Sand Member in the intermound areas, but shows clear convergent onlap, upturning, and rotation toward the crests of the structures (Figure 8). Horizon B is a moderate amplitude reflection positioned in the middle part of the overburden and dated as early Piacenzian (late Pliocene) (Figure 8). It overlies all the mounds as a continuous surface that is slightly deformed into domal folds above their crests (Figure 8). Horizon C marks the upper boundary of the overburden and is close to the Piacenzian–Gelasian boundary. It is a continuous, moderate-amplitude reflection that is slightly deformed into small-scale folds over the NMC and the MMC (Figure 8).
The most significant of the previous observations is that horizon A exhibits convergent onlap across all the mounds (Figure 8). This specific form of onlap demonstrates that the structures were present on the seabed during deposition of this unit. Provided that horizon A is dated as late Zanclean, we can thus establish the approximate timing for the initiation of the mounds as middle Zanclean. The package between horizons A and B varies laterally in thickness (Figure 8). Near the AM, it forms an interval of constant thickness that terminates in parallel onlap onto the flanks of the structure (Figure 8). Conversely, toward the NMC and the MMC, it thins appreciably by convergent onlap (Figure 8). The identification of parallel onlap onto the AM indicates the fill of an established topography by postdefor-mational sedimentation. In contrast, the convergent onlap onto the NMC and the MMC implies growth of a series of established topographies on the paleosea-bed affected by synsedimentary folding. The different geometries demonstrate that the growth of the AM concluded before the deposition of the A–B interval, whereas the growth of the NMC and the MMC mounds continued. The geometric relationships of the B–C interval to the mounds are equivalent to those of the A–B interval, but with diminution of the thickness variations. No evidence for forced folding at horizon C or in the overlying strata exists. From these observations, we interpret that the development of the AM was completed during the late Zanclean (about 4 Ma), whereas the NMC and the MMC remained active up to the early Gelasian (late Pliocene) (about 2.5 Ma).
Canyon Margin Ridge Structures
Although the main focus of this chapter is on a series of mounded structures along the axis of the Afiq submarine canyon, there are several other features observed along the northern canyon margin that are of fundamental significance for the conclusions of this study. These features show many of the key character-istics previously identified for the mounds (e.g., convex-upward geometry, internal chaotic facies), but they lack a well-defined lower boundary and have clear differences in terms of planform geometry, volumes, and internal character. In this section, we provide a detailed description of their 3-D appearance.
The general planform appearance of this second group of structures is best described with reference to the structural time contour map of the top Yafo Sand Member (Figure 7). On this map, they appear as two ridge-shaped structural culminations (marked R) oriented in a south-southeast–north-northwest direction. The imaged parts of the two ridges cover areas of as much as about 40 km2 (15 mi2) (mean length 15 km [9 mi], mean width 4 km [2.5 mi]) and have a maximum structural relief of approximately 150 m (492 ft). Both ridges are restricted to the areas underlain by the northern flank of the Afiq submarine canyon. On strike seismic profiles (Figure 12), they appear as steep-sided domal features (marginal flank dips of as much as 15°) that are invariably positioned above the highest point of the northern flank of the Afiq submarine canyon. Their internal seismic character is dominated by chaotic seismic reflections that are most likely derived from the canyon fill (Figure 12, Figure 15). Velocity pushdown effects are seen below them (Figure 15).
The lower Yafo Formation is the sedimentary overburden of the ridges (Figure 12, Figure 15). Correlation of intraoverburden seismic reflections is possible above the structures, and this shows that there are significant thickness variations (i.e., growth intervals) adjacent to them. Such variations demonstrate that the ridges initiated after deposition of the Yafo Sand Member, and that there was an important phase of growth during the deposition of the younger Yafo Mudstone Member (horizons A and B). Growth of the ridges was likely to be complete soon after the deposition of horizon B. From these observations, we constrain the approximate timings for initiation and completion of the structures as middle and late Zanclean, respectively. This is a highly significant conclusion because it demonstrates that the ridges and the mounded structures initiated synchronously along the Afiq submarine canyon during the middle Zanclean.
Comparing the preceding observations against the criteria of Morley (2003) for the classification of mud intrusions and their seismic signature, we interpret the ridge structures as mud diapirs. The source rock for the intruded material is uncertain because of a lack of well calibration penetrating the structures; from the seismic data, we propose that it corresponds to the shallower levels of the infill of the Afiq submarine canyon (i.e., Mavqiim and Bet Guvrim formations). An input of material from deeper stratigraphic levels (i.e., Paleocene, Eocene) is rejected based on the continuity of the seismic reflections at these intervals (Figure 12). This conclusion is further supported by the absence of fluid pathways (e.g., gas chimneys) below the structures. The ridged planform morphology of the diapirs provides a suggestion of some form of alignment of sediment-injection feeder routes. Based on the spatial coincidence, we propose that the northern flank of the Afiq Canyon acted as a preferential zone of weakness for remobilized mud to move vertically and build the diapirs.
Origin of the Mounded Structures
The unusual geometry and high flanking angles of the mounded structures suggest that they are not primary depositional features but the product of post-depositional deformational processes (sensu Brooke et al., 1995). Several genetically distinct post-depositional geological features are mounded in cross section, circular to subcircular in planform, and similar in scale to the observed structures. These, among others, comprise igneous, salt and mud diapirs, mud volcanoes, sand mounds, and slumps. In the case of the mounded structures presented here, the number of possible interpretations is constrained by seismic and well data. For instance, the internal parts of the mounds are composed of well-mixed sands and clays. No evidence from the well data of salt or igneous lithologies is present, which rules out an igneous or salt diapir origin (e.g., Alsop et al., 2000; Davies et al., 2002). A sand mound origin (Jenssen et al., 1993; Brooke et al., 1995; Dixon et al., 1995; Huuse et al., 2004) is precluded on the basis that large amounts of clay have been recorded from the structures. Slump deposits may appear as chaotic mounded geometries on seismic profiles; they are, however, excluded as a possibility because there is no evidence of a basal shear plane, a headscarp, and a toe region, which are essential elements of slump deposits (Varnes, 1978; Frey-Martinez et al., 2005).
The overall geometries of the mounds could suggest mud volcanism as an origin; this is precluded based on several important observations previously presented. For instance, no evidence of erosion associated with the upper boundaries of the mounds is present, which strongly suggests that they were not exposed at the seabed. In addition, it is evident that the Yafo Sand Member is domed upward above the structures. This observation, together with the identification of a steep (about 15–20°) and irregular contact between the chaotic core and the Yafo Sand Member (Figure 10, Figure 13), suggests a forceful intrusive origin. Additionally, the overburden (Yafo Mudstone Formation) is deformed into a series of folds and faults against the flanks and crests of the structures (Figure 11). These styles of deformation are morphologically consistent with the characteristics of upwarped domes formed by flexure and folding caused by the forceful intrusion of material into shallow-level sedimentary strata (Price and Cosgrove, 1990). Alternatively, it could be argued that such deformational styles may be partly the result of differential compaction of the overburden after burial of the mounds being formed on the paleoseabed as mud volcanoes. However, the high angles of the folds against the structures (20–25°), and the recognition of concentric and radial faults systems, are clear indications of brittle deformation of the overburden by hydraulic jack-up. It is thus clear that the mounds are the result of some form of intrusive processes, and only this possibility is given further consideration.
Another fundamental observation that helps constrain the origin of the mounded structures is their lithological composition. Well data have revealed the presence of large amounts of allochthonous clastic sedi-ments in the NMC. Nannofossils and microfauna recorded from the Nir-1 well indicate that these sediments are polygenetic and originate from various stratigraphic levels (Cretaceous to upper Miocene). This requires an external input of material in the mounds and forms the basis for inferring two possible origins for the sed-iments: horizontal or vertical reworking. The evidence for a preexisting depositional fairway (Afiq submarine canyon) could suggest that the allochthonous material in the mounds was derived from the reworking of the canyon fill; this is excluded as a possibility because of the presence of a much higher ratio of al-lochthonous fossils in the mounds compared with the same stratigraphic levels nearby. Furthermore, in such a depositional environment, sedimentary structures (e.g., lamination, ripples, or cross-stratification) would be expected, and these have not been observed in the available well data. Instead, dip and azimuth analysis from the Nir-1 well have revealed that the material in the structures is highly chaotic and remixed, which strongly suggests an origin caused by some type of postdepositional remobilization process. It is proposed that the allochthonous material was derived vertically from deeper stratigraphic sections and injected into shallower levels. Such input of material would require a vertical feeder system such as pipelike structures or fractures (Hanken et al., 1996; Netoff, 2002; Draganits et al., 2003) for fluids and rock being mobilized upward from deeper stratigraphic levels and intruded into the mounds. No evidence of such structures has been observed; however, this may be explained by the difficulty of making any primary observation of vertical feeders in the highly disturbed zones directly beneath the mounds. The lack of any mapped feeder does not therefore preclude the presence of such features directly beneath each major mounded structure.
Three-dimensional mapping has shown the presence of elongated and crescent-shaped marginal topo-graphic depressions surrounding most of the mounds (Figure 7, Figure 12). Although the depressions could be interpreted as some form of erosional features, the absence of any evidence of erosional truncation and their bowl-like geometry does not support such an origin. Instead, we prefer the interpretation that they are the result of local subsidence around the area of emission (feeder system) because of the accommodation formed on the paleo-sea floor caused by the loss of material at depth. Similar depressions were documented around other feeder systems (Nichols et al., 1994; Netoff, 2002; Davies, 2003). Importantly, the thickness of the Yafo Sand Member remains constant across the depressions, indicating that no material has been removed from this stratigraphic interval. We thus propose that remobilization affected the intervals beneath the Yafo Sand Member, which subsided as coherent layers caused by sediment evacuation deeper in the section. Such a mechanism requires the volume of allochthonous material in the mounds to be approximately balanced by the loss of an equivalent volume from the depressions, excluding the loss of any pore fluid from the system. From measurements on isopach maps and depth-converted seismic sections, we estimate the volume of sediment intruded within the mounds to be approximately equivalent to the material evacuated from the surrounding depressions. This evidence suggests that the material injected within the structures was remobilized from the Afiq submarine canyon and reinforces an injected origin for the mounds.
Based on the previous observations and discussion, the mounded structures in the study area are inter-preted to have formed by clastic forceful intrusion into shallow-level sedimentary host rocks (Yafo Sand Member). From the microfauna records, the sources for the intruded material were most probably the Bet Guvrim and Ziqim formations in the Afiq submarine canyon. Comparable structures involving forceful intrusion in association with diapirism have been generated in theoretical modeling (Schultz-Ela, 2003) and described from outcrop and seismic-based studies (Jenyon, 1986, 1988; Kempler et al., 1996; Alsop et al., 2000; Davison et al., 2000).
Genetic Model and Evolution
The arguments advanced in the previous section point to a genesis for the mounds in which mud re-mobilization in the Afiq submarine canyon became a key factor. Clastic injection occurs when fluidized sedi-ment is depleted from a source area and flows through a series of feeder systems (e.g., fractures, blowout pipes) to be entrained into a host rock. Sediment becomes mobilized because it is in a condition of insufficient strength to resist the forces driving it to move (Maltman and Bolton, 2003). A sufficient pressure differential must be then maintained for clastic intrusions to propagate and develop (Morley, 2003).
We propose that overpressures were triggered within the Afiq submarine canyon by earthquake activity. A direct link between earthquakes and mobilization of clastic sediments has been widely described in the literature (Bankwitz et al., 2003; Yassir, 2003). Earthquakes would significantly increase the fluid pressure and weaken the seals in the canyon infill, thus facilitating seal rupture and clastic injection. Because of its position at the zone of interaction between the Anatolian, African, and Arabian plates, the continental margin of Israel is a seismically active region. Local earthquakes along structures underlying the Afiq Canyon, such as the Syrian arc system, could have caused local overpressures and triggered sediment mobilization. A high probability exists that the region was seismically active during the early Pliocene. Paleoseismic activity is to have been expected because tectonic studies give evidence for active geodynamic processes in the Pliocene (Garfunkel and Almagor, 1985; Garfunkel, 1998; Robertson, 1998; Vidal et al., 2000; Huguen et al., 2001).
Fluids were trapped in the sand-rich intervals (i.e., turbiditic sands) interbedded with poorly permeable claystones of the canyon infill. The clay-rich intervals built a series of permeability barriers that prevented drainage and consolidation of the sand-rich units. This induced the formation of overpressured, compart-mentalized sandstone reservoirs that created a metastable pore-fluid regime in the Afiq submarine canyon (Figure 16A). Eventually, fluid overpressure approached lithostatic pressure, and the seals failed by hydraulic fracturing (Mandl and Harkness, 1987). This created a differential pressure gradient in the canyon that trig-gered rapid vertical movement of fluids. Fluids migrating through the canyon infill fragmented and mobilized sediments along their path. As a result, a low-density mixture of fluids and fine-grained sediments (fluidized sediment) were mobilized and moved upward (Figure 16A). In most of the cases, the fluids moved discordant to bedding in the canyon fill (Figure 16A); however, they probably also used bedding-concordant, high-permeability intervals (e.g., turbiditic sandstones) as subhorizontal migration pathways. This created a lateral dissipation of the overpressures along the canyon fill and would explain the presence of fluidized sediments along its northern flank (Figure 16A).
The fluidized sediments migrating upward exerted a continuous stress to the base of the Yafo Sand Member along the Afiq submarine canyon (Figure 16A). Eventually, this stratigraphic level failed under tension and fractured, leading to catastrophic intrusion of fluidized sediment. During emplacement, the fluidized sediment moved upward to hydraulically deform the overall parallel-bedded strata of the Yafo Sand Member (Figure 16A). In the northern flank of the canyon, intrusion occurred though a system of aligned feeders that built the ridge structures; several aligned feeders are inferred to have formed the mounds. After the first major episode of mud injection, growth of the canyon-margin ridge structures and of those mounds positioned in the eastern parts of the canyon axis (e.g., AM) ceased. Concurrently, relief at the paleoseabed was almost eliminated by deep-water sedimentation (horizon A; Figure 16B). Development of the mounds continued in the western axis of the canyon (e.g., NMC and MMC) for a more protracted period (Figure 16C). Because intrusion of fluidized sediment persisted, the Yafo Sand Member was intensively fragmented into a series of blocklike segments with primary bedding preserved surrounding a core of forcefully intruded clay-rich sediment (Figure 16D). Finally, as several mounds formed in close proximity, they merged and intersected to form complex arrays (e.g., NMC and MMC) that were flanked by elongate and bowl-shaped depressions.
A final and important point to be addressed regarding the genesis of the mounded structures is the burial depth at which they formed. It has been argued that the first, major episode of mud injection along the Afiq submarine canyon occurred during the middle Zanclean. At this time, the top of the Yafo Sand Member was exposed at the seabed. Considering that the injected mud reached the base of the Yafo Sand Member and that the average thickness of this interval is about 70 m (229 ft) (see previous section), it is clear that mud injection occurred up to very shallow depths (about 50–70 m [164–229 ft]). Such a shallow occurrence raises the question of why the injected material did not extrude to the seabed. It is possible that a minor part of the injected material reached the seabed, but this cannot be resolved in the seismic data. It seems unlikely that muds would intrude at the base of the Yafo Sand Member if there was already a connection to seabed that would allow the mud to flow freely from the overpressured canyon fill, as long as overpressure persisted. Mud extrusion at the seabed is in conflict with the previously presented observations that proved many of the mounds to occur through several discrete episodes of mud intrusion throughout a long period. Another possibility is that mobile muds may have been restricted in their vertical ascent by a sealing unit in the Yafo Sand Member. The Yafo Sand Member consists mainly of a succession of marls, siltstones, and sandstones that are overlain by thick units of turbiditic sandstones. Limestones may form seals, especially where they are mixed with shale-rich lithologies, but we do not consider the thin carbonate stringers present in the Yafo Sand Member to be sufficiently competent to prevent extrusion to the seabed in an overpressured system. An alternative that appears more plausible is that when muds intruded vertically, they encountered the highly bedded lower Yafo Sand Member and the permeable massive sandstones of the upper Yafo Sand Member. It is possible that the sandstones allowed overpressure to dissipate, thus diminishing the intrusion drive in a manner analogous to an internal blowout. In this context, bedding anisotropy and lateral spreading of the intruded mud mass in the Yafo Sand Member may have interacted to form the mounded structures. This assertion is supported by the idealized mechanical theory of clastic intrusion (Jolly and Lonergan, 2002). According to these authors, the intrusion of dikes occurs at greater burial depth than sills where the effects of bedding anisotropy exceed the effects of the vertical stress. In a system where mobile muds intrude vertically and encounter a highly bedded succession close to seabed (such as the Yafo Sand Member), this theory would thus predict the formation of laccolithic bodies that jack up the overburden in preference to dike intrusion to the seabed.
Implications for Hydrocarbon Exploration and Production
Complex deformational features of the type discussed here are important to understand in the context of hydrocarbon exploration and production. The most important implication is that they form structural traps for hydrocarbons; mud remobilization and intrusion can cause deformed reservoir intervals into four-way dip closures that act as excellent traps for oil and gas accumulations. Another implication is that intrusion of mud-rich sediments into a reservoir interval may produce a network of permeability barriers (i.e., dikes and silts) that compartmentalize reservoir lithologies and, thus, reduce intrareservoir connectivity. In addition, mud intrusion may modify the stratified structure of the host rocks into complex, disrupted units that contain only remnants of the original stratification.
This can significantly affect the geometry, thickness, and reservoir properties of the intruded intervals.
The impact of mobile shales is not just confined to trapping geometries, permeability barriers, and alter-ation of the reservoir geometries. It may also affect hydrocarbon migration and remigration; mud diapirs are frequently associated with large accumulations of hydrocarbons (Morley et al., 1998) and maybe associated with areas of fluid flow (Løseth et al., 2003). The feeder systems associated with the structures may create effective migration pathways for hydrocarbons through otherwise low-permeability intervals. Structures of the type presented here can thus be used to enhance the evaluation of the paleofluid system of hydrocarbon-generating basins.
Seismic and well data have revealed the presence of 12 large-scale mounded structures with diameters of as much as 4 km (2.5 mi) in the lower Pliocene, offshore Israel. The structures are restricted to an area 40 × 20 km (25 × 12 mi) aligned along the Afiq submarine canyon, a former depositional fairway of Oligocene age. They may form isolated four-way dip closures or arrays of coalesced features.
The mounded structures are interpreted as the result of clastic forceful intrusion into shallow-level sedimentary host rocks (Yafo Sand Member) and subsequent hydraulic jack-up of the overburden (Yafo Mudstone Member). Their localized occurrence is attributed to the spatial constriction of the clastic successions that housed the overpressure in the Afiq submarine canyon. Overpressure was likely to be triggered by seismic activity.
The mounded structures were active between the middle Zanclean (about 4 Ma) and the early Gelasian (about 2.5 Ma). They formed during several episodes of activity, which alternated with deep-water sedimentation. Initiation of the mounds was coeval with the formation of large-scale ridgelike mud diapirs along the northern flank of the Afiq submarine canyon. The mounds first became inactive at the landward parts of the Afiq submarine canyon, while those in the seaward areas remained active for a longer period.
The mounded structures presented here are among the most complex examples of clastic intrusion features described to date. They are important for hydrocarbon prospectivity as they may create eco-nomically significant trapping geometries. Conversely, mud-rich clastic intrusions may form complex intrareservoir networks of permeability barriers that degrade reservoir quality.
BG-Group is thanked for providing the 3-D seismic and well data, and for permission to publish this chap-ter. We acknowledge the support on GeoQuest IESX applications provided by Schlumberger Systems Infor-mation. A. Fraser and S. Adiletta provided very valuable comments and suggestions in their reviews, which greatly improved the original manuscript. C. Bertoni, S. Beavington-Penney, and S. Maddox are thanked for their editorial comments and I. Campbell for his help with gaining permission for the publication of the data. The ideas and interpretations presented herein are those of individuals and, thus, do not necessarily reflect those of BG-Group or its partners.
Figures & Tables
Sand Injectites: Implications for Hydrocarbon Exploration and Production
Sand injectites are described in scientific literature as an increasingly common occurrence in hydrocarbon reservoirs, in particular in deep-water clastic systems, where they are known to influence reserves distribution and recovery. Seismically-detectable injected sand bodies constitute targets for exploration and development wells and, subseismic sand bodies provide excellent intra-reservoir flow units that create field-wide vertical communication through depositionally extensive, low-permeability units. As sand injectites form permeable conduits in otherwise low-permeability units they facilitate the expulsion of basinal fluids; hence they act both as a seal risk and mitigate timing and rate of hydrocarbon migration. Injected sand bodies form intrusive traps, which are distinct from structural or stratigraphic traps. Included in this publication are 10 chapters on subsurface examination of sand injectites, 1 chapter on theoretical considerations, and 13 outcrop analogs in reservoirs across the world. Captured in this volume is at least a taste of the global and stratigraphic distribution of sand injectites, and an attempt to introduce readers to sand injectites and their significance in the context of hydrocarbon exploration and production. The book is not intended as a complete review of the field-based literature, but emphasizes high quality case studies from the surface and subsurface. The geographic scope of the book is large, and illustrates the diversity of geological settings in which these fascinating and economically significant features are found.