You are on page 1of 86

Petroleum Geoscience

The impact of the Messinian Salinity Crisis on the petroleum system of the Eastern
Mediterranean: a critical assessment using 2D-petroleum system modelling
--Manuscript Draft--

Manuscript Number:

Article Type: Thematic set article

Full Title: The impact of the Messinian Salinity Crisis on the petroleum system of the Eastern
Mediterranean: a critical assessment using 2D-petroleum system modelling

Short Title: Eastern Mediterranean Petroleum System

Corresponding Author: Alastair Fraser, PhD


Imperial College London
London, UNITED KINGDOM

Corresponding Author E-Mail: alastair.fraser@imperial.ac.uk

Other Authors: Abdulaziz Al-Balushi

Martin Neumaier

Christopher Aiden-Lee Jackson

Abstract: The offshore Levant Basin demonstrates one of the most phenomenal natural
examples of a working petroleum system associated with a relatively rapid unloading
and loading cycle caused by the the Messinian Salinity Crisis (MSC). In this study, 2D
basin and petroleum systems modelling suggests that the geologically instantaneous
water unloading of c. 2070 m and subsequent rapid salt deposition and refill impacts
the subsurface pore pressure and temperature in the underlying sediments. The
pressure drop is modelled to be instantaneous, whereas the impact on temperature is
more of a transient response. This has important consequences for the shallow sub-
Messinian biogenic petroleum system, which is assumed to have experienced fluid
brecciation associated with massive fluid escape events. Deeper Oligo-Miocene
sediments are far less affected, thus indicating a "preservation window" for biogenic
gas accumulations, which hosts the recent discoveries (Tamar, Leviathan, Aphrodite).
Hydrocarbon accumulations of a "bubble point oil" composition are modelled to have
experienced cap expansion during the drawdown, with the pressure drop being the
primary control. This study suggests that seal-limited traps are expected to have
undergone a catastrophic seal failure whereas the impact of the MSC is modelled to be
less destructive for size-limited and particularly charge-limited traps.

Section/Category: Messinian Salinity Crisis

Manuscript Classifications: Geochemistry; Petroleum geology

Additional Information:

Question Response

Are there any conflicting interests, No


financial or otherwise?

Samples used for data or illustrations in Confirmed


this article have been collected in a
responsible manner

Powered by Editorial Manager® and ProduXion Manager® from Aries Systems Corporation
Article text Click here to download Article text Al_Balushi_et_al_final.docx

1 The impact of the Messinian Salinity Crisis on the petroleum system of the Eastern

2 Mediterranean: a critical assessment using 2D-petroleum system modelling

4 Abdulaziz Nasser Al-Balushi1*, Martin Neumaier2, Alastair J. Fraser1 and Christopher A-L.

5 Jackson1

7 1Basins Research Group (BRG), Department of Earth Science & Engineering, Imperial

8 College London, Prince Consort Road, London, SW7 2BP, England, UK

9 2Aachen Technology Centre, Schlumberger, Ritterstrasse 23, 52072 Aachen, Germany

10

11 *Corresponding author: alastair.fraser@imperial.ac.uk

12

13 1. Abstract

14 The offshore Levant Basin demonstrates one of the most phenomenal natural examples of

15 a working petroleum system associated with a relatively rapid unloading and loading cycle

16 caused by the the Messinian Salinity Crisis (MSC). In this study, 2D basin and petroleum

17 systems modelling suggests that the geologically instantaneous water unloading of c. 2070

18 m and subsequent rapid salt deposition and refill impacts the subsurface pore pressure and

19 temperature in the underlying sediments. The pressure drop is modelled to be

20 instantaneous, whereas the impact on temperature is more of a transient response. This

21 has important consequences for the shallow sub-Messinian biogenic petroleum system,

22 which is assumed to have experienced fluid brecciation associated with massive fluid

23 escape events. Deeper Oligo-Miocene sediments are far less affected, thus indicating a

1
24 “preservation window” for biogenic gas accumulations, which hosts the recent discoveries

25 (Tamar, Leviathan, Aphrodite). Hydrocarbon accumulations of a “bubble point oil”

26 composition are modelled to have experienced cap expansion during the drawdown, with

27 the pressure drop being the primary control. This study suggests that seal-limited traps are

28 expected to have undergone a catastrophic seal failure whereas the impact of the MSC is

29 modelled to be less destructive for size-limited and particularly charge-limited traps. (200

30 words)

31

32 Key words: Messinian Salinity Crisis, petroleum system modelling, biogenic gas,

33 Tamar, pressure, temperature, phase change.

34

35 2. Introduction

36 Surface processes play a pivotal role in moving loads from one area to another. This

37 redistribution of surface mass causes the Earth’s surface to respond either by subsidence

38 (in the case of loading) or uplift (in the case of unloading), which may subsequently affect

39 subsurface pressure and temperature equilibrium conditions (Allen and Allen 2013). From

40 the perspective of a solid Earth, the nature of the surface load, whether it is water, ice,

41 sediments or rocks, is irrelevant; the only thing that matters is the weight.

42

43 One of the most phenomenal natural examples of relatively rapid unloading and loading of

44 the Earth’s crust occurred ca. 5.96 million years ago in the Mediterranean Sea, during an

45 event that is known as the Messinian Salinity Crisis (MSC). During the MSC, the

46 Mediterranean Sea experienced a geologically instantaneous drop in sea level in excess of

47 ca. 1000 m (Ryan, 1976a; Ryan & Cita, 1978; Bartol & Govers, 2009; Urgeles et al., 2011),

2
48 and the rapid deposition of widespread evaporitic sequences that are up to ca. 2000 m

49 thick in the deeper parts of the basin, typically in areas floored by oceanic crust (Hsü, 1972;

50 Meijer & Krijgsman, 2005). Since the discovery of these evaporitic sequences, numerous

51 studies have attempted to explain the palaeo-geographic setting of the basin and the

52 depositional environment that governed the MSC. Despite these efforts, the relationship

53 between geologically instantaneous water unloading and evaporite loading, and the impact

54 that these had on subsurface pressures, temperatures and the subsequent distribution of

55 hydrocarbons in the eastern Mediterranean, have not been well established. For example,

56 the instantaneous drop in sea level is expected to be associated with a drop in pressure,

57 potentially causing hydrocarbon phase change. The first attempt to address this impact was

58 presented by Fraser et al. (2011) with further supporting evidence from 3D seismic data

59 provided by Bertoni et al. (2013).

60

61 The relationship between the surface processes and subsequent changes in subsurface

62 pressure and temperature conditions is well-established in some basins that have

63 experienced significant periods of tectonically driven uplift (e.g. Hammerfest Basin, offshore

64 northern Norway;Rodrigues Duran et al. (2013)). In the Hammerfest Basin, despite the

65 existence of mature, oil-prone source rocks (Ohm et al., 2008), almost all discoveries are

66 predominantly gas with uneconomical volumes of oil (NPD, 2014). Extensive Cenozoic

67 uplift, ca. 1000-3000 m of erosion and, therefore, removal of a crustal load along the

68 western basin margin (Laberg et al., 2012), as well as the high-latitude Quaternary

69 glaciation (Cavanagh et al., 2006) are thought to have caused trap tilting and exhumation

70 (Doré et al., 2002), leakage and redistribution of hydrocarbons due to phase change, gas

71 expansion, and subsequent flushing of oil from reservoirs (Nyland et al., 1992). This has

72 ultimately led to predominance of gas over oil in the Hammerfest Basin and elsewhere on

3
73 the Barents Sea shelf. The presence of extensive gas clouds and amplitude anomalies, in

74 addition to the documentation of palaeo oil-water contacts, together suggest a dominantly

75 leaky system and show that the structures defining the Snøhvit and Askelad fields, which lie

76 in the Hammerfest Basin, were once filled with significantly larger volumes of hydrocarbons

77 (Linjordet & Grung Olsen, 1978). Even though the origin, order and timing of the unloading

78 and reloading events are somewhat different, this is considered as an appropriate analogue

79 for the effect of the MSC in the eastern Mediterranean.

80

81 Recent exploration drilling results in the eastern Mediterranean have shown that the

82 offshore Levant Basin is a very prolific gas province. Proven gas reserves totalling ca. 30

83 trillion cubic feet (TCF) have been estimated for the most recent, sub-Messianian salt

84 discoveries in the Tamar, Leviathan and Aphrodite gas discoveries (Needham et al., 2013).

85 However, further exploration activity would benefit from a clearer understanding of the role

86 that the MSC played in shaping the present distribution of oil and gas in the region.

87

88 The eastern Mediterranean Basin, which comprises the Levant Basin, Nile Cone, offshore

89 Sirt, offshore Western Desert and the Herodotus Basin, is located in the south-eastern

90 Mediterranean region, near the complex boundary between the African, Arabian and

91 Eurasian plates. This study focuses on the petroleum system evolution of the Levant Basin

92 during the Messinian, although the results and implications of this modelling are relevant to

93 other basins in the region (e.g. offshore Sirt, Libya and Western Desert, Egypt) which have

94 experienced similar MSC drawdown effects. The Levant Basin is bound by the Nile Cone to

95 the south, the Dead Sea Transform to the east, the Cyprus Arc and the Latakia Ridge to the

96 north, and the Eratosthenes Seamount to the west (Figure 1).

97

4
98 This contribution aims to improve the understanding of the influence of the MSC on the

99 proven biogenic and speculative thermogenic petroleum systems of the eastern

100 Mediterranean. To achieve this we: (i) use 2D basin modelling to reconstruct the

101 subsidence history of the Levant Basin and (ii) assess the impact of the instantaneous and

102 significant drop in sea level and rapid salt deposition on subsurface pressure and

103 temperature (PT) conditions during the Messinian; (iii) use 2D petroleum systems modelling

104 to test the impact of the PT changes on the biogenic source rocks and gas accumulations

105 as well as on hypothetical thermogenic oil accumulations.

5
106 3. Geological Setting

107 The present Levant Basin is a remnant of a larger, Neo-Tethyan oceanic basin that opened

108 between several fragments of the Pangaea supercontinent in the Early Mesozoic (Bein and

109 Gvirtzman 1977; Dewey et al. 1973; Garfunkel and Derin 1984; Robertson and Dixon

110 1984a). Since Senonian times, the basin started to close as a result of the collision

111 between the Afro-Arabian and Eurasian plates which resulted in the subduction of the

112 Tethyan oceanic crust under the Cyprus Arc in the Neogene (Garfunkel 1998; Woodside

113 1977). Convergence between the African and Eurasian plates is still ongoing and is defined

114 by the active subduction boundary of the Cyprus Arc and the Latakia Ridge (Ben-Avraham

115 et al. 2002) (Figure 1).

116

117 The structure of the eastern Mediterranean margin and the adjacent deep-marine basin

118 preserves the signature of the main tectonic events concomitant with the opening and

119 closing of the Neo-Tethyan ocean. These events can be summarized into the following

120 tectono-stratigraphic phases (Figure 2): (i) Middle-to-Late Jurassic continental rifting and

121 break-up; (ii) Early Cretaceous thermal subsidence and formation of a passive continental

122 margin; (iii) Late Cretaceous-to-Recent convergence and contraction; (iv) Early Oligocene

123 Gulf of Suez rifting and Middle Miocene Red Sea break-up and (v) Late Miocene (i.e. MSC)

124 and subsequent Pliocene-Pleistocene flooding. In the following sections we provide a brief

125 summary of the main tectonic phases and depositional events associated with each phase.

126 Such a summary is important because these tectono-stratigraphic events directly impacted

127 the facies distribution in the basin which underpins the permeability structure and thus the

128 fluid flow during hydrocarbon migration.

129

6
130 (i) Middle-to-Late Jurassic continental rifting and break-up

131 Break-up of the northern part of Gondwana into several microplates (e.g., Tauride,

132 Cimmeride and the Eratosthenes Seamount) occurred in the Latest Palaeozoic to Early

133 Mesozoic (Garfunkel 1998; Hirsch et al. 1995; Longacre et al. 2007; Robertson and Dixon

134 1984a; Stampfli et al. 1991). A NE-SW trending graben and horst system, which records

135 NW-SE extension, is recognized at deeper structural and stratigraphic levels of the central

136 Levantine Basin as seen in seismic reflection profiles (Ben-Avraham et al. 2006; Gardosh et

137 al. 2010). Jurassic rift-related structures have also been observed in the northern part of the

138 Levant Basin (Al-Balushi et al. 2012).

139

140 (ii) Early Cretaceous thermal subsidence and formation of passive continental

141 margin

142 During the Late Early Cretaceous-to-Eocene, the Levant Margin transitioned into a passive

143 margin and thermal subsidence occurred (Bein and Gvirtzman 1977; Gardosh et al. 2010).

144 A shallow marine shelf formed proximal to the present-day coastline, whereas deeper

145 marine conditions were present in the centre of the basin (Ben-Avraham et al. 2006). The

146 Early-to-mid Cretaceous was characterised by the rapid growth and aggradation of a

147 carbonate platform (Ben-Avraham et al. 2006). In the mid Cretaceous, depositional cycles

148 are characterised by the progradation of the carbonate slopes into the basin, followed by

149 extensive growth of carbonate platforms on the margin (Ben-Avraham et al. 2006).

150

151 (iii) Late Cretaceous-to-Recent convergence and contraction

152 The convergence of the African-Arabian plate and the Eurasian plate commenced during

153 the Late Cretaceous and resulted in the initiation of a northwards-dipping subduction zone

154 in the northern part of the Levant Basin (Ben-Avraham 1989; Robertson and Dixon 1984b).

7
155 As a result of this convergence, a series of shortening structures, such as the Syrian Arc

156 fold belt, which were developed throughout much of the Levant region and northern Egypt,

157 were superimposed on the older (Late Mesozoic) horst and graben system (Ben-Avraham

158 et al. 2006). Shortening continued in the Levant region throughout the Cenozoic and

159 resulted in a second phase of Syrian Arc folding in the Early Miocene (Gardosh and

160 Druckman 2006). The bulk of the productive traps discovered in the Western Desert of

161 Egypt are found in Syrian Arc-related structures (Dolson et al. 2001) and the major recent

162 gas discoveries in the Levant Basin have been found in large Syrian Arc-related inversion

163 structures of Early Miocene age (Hodgson 2012).

164

165 (iv) Early Oligocene Gulf of Suez rifting and Middle Miocene Red Sea break-up

166 The Suez rift, about 70 km wide and 500 km long, forms the north-western branch of the

167 Red Sea. The extension and rifting in the Gulf of Suez began in the Oligocene (Patton et al.

168 1994). However, the main rifting phase occurred during the Miocene (Patton et al. 1994;

169 Steckler et al. 1998) which was followed by the Red Sea break-up in the Middle Miocene

170 (Serravalian age) (Dolson et al. 2001). The tectonic opening of the Red Sea separated the

171 Arabian plate from the African plate and initiated a new strike-slip plate boundary known as

172 the Dead Sea Transform (DST), along the Levant Margin (Mart et al. 2005).

173

174 Uplift associated with the Red Sea rifting played a major role in modifying the river drainage

175 patterns across Africa and acted as a key drainage divide and clastic sediment source for

176 the Proto-Nile River System (Macgregor 2012). The rise of the rift shoulders and other

177 concomitant African swells caused systems, which originally drained to the west, to switch

178 northwards and form the current Nile drainage system (Macgregor 2012). This change in

179 drainage direction increased the sedimentation rates in the Nile Delta to be among the

8
180 highest experienced on the African margins. This is supported by the source-sink volume

181 analysis that shows half of this clastic input is sourced from rift shoulders where the erosion

182 rate is estimated to be in excess of 80 m Ma-1 (Macgregor 2012).

183

184 2D seismic isopach mapping by Macgregor (2012) shows that deep-water Oligocene-

185 Miocene sediments exist as far north as northern Levant and their distribution is controlled

186 by the pre-existing structures. A long run-off turbidite system extending hundreds of

187 kilometres basin-ward from the southern shelf edge, such as the present-day Niger or

188 Mississippi systems, seems to provide a good analogue for predicting deep-water Nile

189 reservoirs.

190

191 (v) Late Miocene MSC and Pliocene flooding

192 At the end of the Miocene (c. 5.96 Ma), the Straits of Gibraltar closed as a consequence of

193 the continued convergence of the African plate with the Eurasian plate. The subsequent

194 reduction in the oceanographic connection of the Mediterranean Sea with the Atlantic

195 Ocean led to a drastic palaeo-environmental change known as the MSC (Hsü et al. 1977;

196 Meijer and Krijgsman 2005). A shortage of saline water recharge and high evaporation

197 rates driven by the arid Mediterranean climate resulted in a dramatic drop in sea level,

198 increasing the seawater salinity and promoting deposition of thick (locally over 2 km)

199 evaporite sequences (Hsü 1972; Hsü et al. 1977; Ryan 2009; Ryan and Cita 1978). The

200 magnitude of the Messinian sea-level drop has been a matter of debate since the event

201 was first recognised by(Hsü 1972). It has been estimated as 1300 m (Ben-Gai et al. 2005;

202 Urgeles et al. 2011), 1500-1900 m (Ryan 1976; Steckler et al. 2003), 1730-2230 m (Bartol

203 and Govers 2009), and up to 3500 m (Ryan 1978). A new estimate of c. 2070 m, which has

204 been used in this modelling study, has been suggested based on evidence for low-stand

9
205 Messinian shorelines offshore from the Western Desert, Egypt (Al-Balushi et al. 2013).

206 During this significant and geologically rapid sea level fall, the Levant Margin underwent

207 extensive erosion, resulting in the formation of deeply incised submarine canyons

208 (Gvirtzman and Buchbinder 1978).

209

210 Pliocene (Zanclean) flooding marked the end of the MSC (c. 5.33 Ma) and re-establishment

211 of normal marine conditions in the Mediterranean basins (Hilgen et al. 2007; Hilgen and

212 Langereis 1993). The Pliocene-to-Recent succession unconformably overlies the Messinian

213 Salt and consists of mainly Nile-derived sediments that reach a thickness of up to 4 km over

214 the Nile Cone area (Gardosh and Druckman 2006; Segev et al. 2006). Along the Levant

215 Margin, however, the Pliocene-to-Recent succession is thinner (c. 0.5-1 km) and

216 characterised by mud-dominated sequences composed of claystone and siltstone (Gardosh

217 and Druckman 2006).

10
218 4. Dataset and Methodology

219 This study utilises a 2D regional seismic reflection dataset, composed of multiple surveys

220 acquired between 2000 and 2005. Data are pre-stack time-migrated and cover ca. 185,000

221 km2 of the eastern Mediterranean Basin, extending from northern Lebanon across the Nile

222 Delta and westward to the offshore Western Desert and Herodotus Basin (Figure 1). The

223 average spacing between lines is ca. 10 km. The dominant frequency of the seismic data in

224 the sallower section (< 5 sec two-way time, TWT) ranges from ca. 7.5 to 25 Hz, giving a

225 vertical resolution of ca. 25-80 m based on an average velocity of 2600 ms -1. In the deeper

226 section, however, the dominant frequency ranges from ca. 10 to 15 Hz, giving a vertical

227 resolution of ca. 80-125 m based on a higher average velocity of 5000 ms -1. The data have

228 a record length that ranges from ca. 8 sec TWT in the Herodotus Basin, deep-water Nile

229 and Eratosthenes Seamount, to as much as 12 sec TWT in the Levant Basin. In addition,

230 we have access to 20-ultra-deep lines (record length of 14-20 sec TWT) that image

231 crystalline basement; these lines cross the Nile Cone.

232

233 Five wells that penetrate down to ca. 4500 m below seabed were available for this study

234 and provided stratigraphic control on the age of the mapped seismic packages (Mango-1,

235 North Sinai 21-1, Marakia-1, Sidi Barrania-1 and Memphis-1; Figure 1). Mango-1

236 terminated in the Upper Jurassic, North Sinai 21-1, Marakia-1 and Sidi Barrania-1

237 terminated in the Lower Cretaceous, and Memphis-1 terminated in the Upper Cretaceous.

238 Very few wells have published temperature data (Dubille and Thomas, 2012).

239

240 A 220 km long, NW-trending seismic cross section from the Levant Basin was selected for

241 basin and petroleum systems modelling (Figure 1 and Figure 3). The image quality is very

11
242 good to excellent down to ca. 7 sec TWT, at which point image quality deteriorates. The

243 stratigraphy on either side of the section was constrained by well data from the Levant

244 Margin (Figure 3a; (Gardosh & Druckman, 2006a; Gardosh et al., 2010; Gardosh et al.,

245 2011) and Deep Sea Drilling Program (DSDP) wells from the Eratosthenes Seamount

246 (Robertson, 1998c). The central part of the section is based on extrapolated stratigraphy

247 from the margins. However, there is a degree of uncertainty introduced here due to the

248 absence of direct well control below the Turonian and uncertainties associated with

249 interpreting deep stratigraphy across the main basin-bounding faults, some of which have

250 large throws (Figure 3).

251

252 Notwithstanding these uncertainties, this cross section was selected because: (i) it

253 provides a relatively clear, cross-sectional image of the subsurface geology from the Levant

254 Margin to the Eratosthenes Seamount, imaging down to basement depths; this is critical for

255 allowing the construction of a robust geological model that underpins the petroleum

256 systems model; (ii) it ties with well data on the Eratosthenes Seamount and the Levant

257 Margin, which offer excellent constraints on the stratigraphy and lithology of the main

258 depositional units; and (iii) it passes close to the main gas discoveries in the basin, thus

259 allowing a better understanding of the petroleum systems context of these important

260 hydrocarbon accumulations (e.g. Tamar, Leviathan and Aphrodite).

261

262 The seismic profile was depth-converted using velocity data from nearby wells in the Levant

263 Basin (Gardosh and Druckman 2006). A detailed description for the sedimentary fill based

264 on wells and seismic facies data along the seismic profile is presented below. This

265 description outlines the predominant depositional environment for each stratigraphic unit

266 and the likely lithology to be encountered in each, which is the controlling parameter for the

12
267 permeability distribution in the basin and thus the modelled hydrocarbon migration and

268 overpressure build-up.

269

270 Basin and petroleum systems modelling was conducted in PetroModTM. The basin model

271 provided a modelled pressure and temperature (PT) development through geologic time,

272 including the high-resolution modelling of the MSC. Embedded in that PT framework, the

273 evolution of the petroleum systems was modelled (hydrocarbon generation, expulsion,

274 migration, accumulation and preservation). This permitted a PVT-controlled simulation of

275 the reaction of potential pre-Messinian hydrocarbon accumulations and the Oligocene-

276 Miocene biogenic petroleum system to pressure and temperature variations associated with

277 the MSC.

278

279 5. Seismic-stratigraphic description of the 2D model

280 Nine seismic horizons (H1-9) bounding eight seismic-stratigraphic units (SU1-8) were

281 interpreted across the study area (Figure 2 and Figure 3). These horizons, from oldest to

282 youngest, are: (i) H1 - Pre-Jurassic (top basement); (ii) H2 - Mid Jurassic; (iii) H3 - Late

283 Cretaceous (Turonian); (iv) H4 - Late Eocene; (v) H5 - Base Miocene; (vi) H6 - Mid

284 Miocene; (vii) H7 - Base Messinian; (viii) H8 - Top Messinian and (ix) H9 -Sea Bed. The

285 seismic-stratigraphic framework is defined by classic megasequence and sequence based

286 stratigraphy which defines a distinct set of internal seismic facies. Seismic data quality and

287 stratigraphic data from wells are good down to the top Cretaceous surface, thus our

288 interpretation is more certain in the overlying interval. However, interpretations at greater

289 depths within the section, are relatively poorly imaged on seismic data and lack well

290 penetrations, hence are more speculative in terms of absolute age and lithology.

13
291 (i) Deep basin structure

292 The deep basin structure is defined by H1 (top basement). This reflection is mapped across

293 the study area and defines a transition from a chaotic and largely reflection-free package

294 below to a package characterised by continuous, parallel-to-divergent, high-amplitude

295 reflections above (Figure 3b and Figure 3c). The seismic expression of the package below

296 H1 suggests it corresponds to crystalline basement and that H1 therefore defines the base

297 of the sedimentary succession. Along the Levant Margin, the basement is at ca. 6 sec TWT

298 and it deepens westward towards the centre of the basin to ca. 7.5 sec TWT. It then rises

299 north-westwards to ca. 5 sec TWT, immediately beneath the Eratosthenes Seamount

300 (Figure 3c).

301

302 A series of NE-trending horst and grabens, bound by broadly NE-SW-striking normal faults,

303 are developed along H1 (Figure 3c). These structures extend from the Eratosthenes

304 Seamount in the west to the Levant continental shelf in the east. A recent gravity inversion

305 study by Cowie and Kusznir (2012) suggests that the Levant continental crust underwent

306 extreme thinning from ca. 35 km onshore to ca.10 km in the centre of the basin offshore,

307 which is broadly in agreement with other publications (Ginzburg et al., 1979; Makris et al.,

308 1983; Ginzburg et al., 1994; Ben-Avraham et al., 2002).

309

310 (ii) Pre-Jurassic-to-Mid Jurassic (SU1)

311 SU1 is bound below and above by H1 to H2, respectively, and is developed across the

312 entire width of the Levant Basin. SU1 is characterized by high-amplitude, relatively

313 continuous, parallel to divergent reflections, locally changing to a less reflective or

314 reflection-free seismic facies on fault-bound, intra-basin structural highs.

14
315

316 The lower boundary of SU1 is the near-top of the acoustic basement. The upper boundary

317 is correlated with a distinct change in seismic character from a high-amplitude, continuous

318 reflection set below to discontinuous, low-amplitude, occasionally shingled, reflections

319 above. The stratigraphy and age of SU1 is not well established because of the lack of well

320 control.

321

322 Well data from the Levant Margin show that Permian sediments overlie the Precambrian

323 basement (Heletz Deep-l, Gevim-1). It is, therefore, suggested that SU1 is dominated by

324 Permian, Triassic and Lower Jurassic strata (Gardosh & Druckman, 2006a). Based on the

325 onshore lithology from well data and the offshore seismic character, this unit is interpreted

326 as continental to shallow-marine siliciclastic and carbonate deposits.

327 (iii) Mid Jurassic-to-Late Cretaceous (SU2)


328 SU2 is bound below and above by H2 and H3, respectively, and it is thickest at the eastern

329 margin of the basin and reduces in the central and western parts. The eastern margin area

330 is characterised by a relatively discontinuous, low-amplitude seismic reflection series,

331 whereas more continuous, high amplitude reflections dominate the deep part of the basin,

332 further offshore.

333

334 The lower boundary of SU2 is correlated with the Mid Jurassic unconformity. The upper

335 boundary is correlated to a distinct transition of seismic character between several high-

336 amplitude reflections at the top of the unit and the overlying lower-amplitude reflections

337 marked by chaotic and reflection-free zones.

338

15
339 SU2 is penetrated by several onshore and offshore wells in the eastern part of the Levant

340 Basin. An important feature identified in these wells is a pronounced facies change from

341 coarse-grained, shelf-type carbonate units east of the present-day coastline to finer-

342 grained, slope and basin-type carbonate and siliclastics further west (Bein & Gvirtzman,

343 1977; Flexer et al., 1986).

344
345 (iv) Late Cretaceous-to-Late Eocene (SU3)
346 SU3 is bound below and above by H3 and H4, respectively, and it attains its maximum

347 thickness in the central part of the Levant Basin and displays marked thinning towards the

348 eastern and western margins. SU3 is characterised by transparent to medium chaotic

349 reflections.

350 The lower boundary of SU3 is correlated with the Late Cretaceous. The upper boundary is

351 correlated with the top of a relatively continuous higher-amplitude reflection series. The

352 change is seismic character corresponds to an unconformity surface.

353

354 Well data indicate that SU3 is composed of chalk and marl of pelagic to hemipelagic, deep-

355 water origin (Almogi-Labin et al., 1993). Deep-water chalky and marly limestone units

356 intercalated by chert-dominated beds have been observed in the Terbol-1 and Adloun-1

357 coastal wells of Lebanon (Chekka Formation; Nader, 2011; Hawie et al., 2013). The

358 unconformity at the upper boundary of this unit is indicated in wells by a pronounced

359 lithological break and a biostratigraphic hiatus between Eocene carbonates and the

360 overlying Oligocene clastic deposits (Gardosh & Druckman, 2006a).

361

362 (v) Late Eocene-to-Base Miocene (SU4)

16
363 SU4 is bound below and above by H4 and H5, respectively, and attains its maximum

364 thickness in the centre of the basin and displays a minimum thickness over the

365 Eratosthenes Seamount. This seismic package is characterised by moderate-amplitude,

366 discontinuous reflections with some localised chaotic reflections over the pre-existing

367 structural highs.

368 The lower high amplitude boundary of SU4 is correlated with an unconformity representing

369 a transition from Upper Eocene shale and carbonate-dominated sequences to a more

370 clastic-rich sequence in the Oligocene (Gardosh & Druckman, 2006a; Lie et al., 2011). The

371 upper high-amplitude boundary separates Oligocene sediments from Lower Miocene deep-

372 water sediments (Lie et al., 2011).

373

374 This unit is likely to be dominated by distal turbidite sequences composed of alternating

375 sandstones, shales and deep-water carbonates (Lie et al., 2011) and delivered to the basin

376 by the Proto-Nile river system. Over the Eratosthenes Seamount, DSDP wells indicate that

377 Oligocene successions are dominated by chalk and are overlain by a carbonate succession

378 comprising clasts of micritic limestone with reworked planktonic foraminifera of Mid Eocene

379 and Oligocene age (Emeis et al., 1996a).

380 (vi) Base Miocene-to-Mid Miocene (SU5)


381 SU5 is bound below and above by H5 and H6, respectively and it shows more pronounced

382 variation in thickness than SU4. The seismic character of this unit varies from a reflection-

383 free zone very close to the Levant Margin to medium-amplitude sub-continuous layered

384 reflections further west. In the centre of the basin, seismic data exhibits high-amplitude

385 continuous reflections and some mounded reflections, interpreted as siliciclastic, deep-

17
386 water turbidite channels and basin-floor fans, with onlapping geometry upon the anticlines

387 (Gardosh & Druckman, 2006a).

388 The lower boundary is the base Oligocene-Miocene unconformity. The upper boundary

389 corresponds to the Mid Miocene. Along the eastern margin, thick intervals of coarse-

390 grained sandstone and conglomerate were encountered in wells (e.g. Hof Ashdod-1)

391 (Gardosh & Druckman, 2006a). In the offshore, SU5 is composed of pelagic siltstone, deep-

392 water siliclastics and shale (Druckman et al., 1994; Gill et al., 1995). On the western side of

393 the section, over the Eratosthenes Seamount, DSDP results at sites 965 and 966 revealed

394 that shallow-water carbonate deposition was dominant during this time (Robertson, 1998c).

395

396 This unit represents the most prolific reservoir interval in the Levant Basin containing giant

397 gas fields (Aphrodite, Leviathan and Tamar) in the Lower Miocene high quality sands

398 trapped in the Syrian arc compressional structures (Durham, 2013; Needham et al., 2013).

399 (vii) Mid Miocene-to-Late Miocene (SU6)


400 SU6 is bound below and above by H6 and H7, respectively and shows symmetrical thinning

401 on both sides of the basin. This seismic package is characterized by low-amplitude, partly

402 discontinuous reflections, some mounded reflections and reflection-free zones.

403 The upper boundary of the unit, the base Messinian, is represented by a marked transition

404 in seismic facies between the low-amplitude, partly continuous reflection character of SU6

405 below to the chaotic and typically reflection-free seismic package of the Messinian salt

406 (SU7) above. As with SU5, this unit is composed mainly of pelagic siltstone, deep-water

407 siliclastics and shale (Druckman et al., 1994; Gill et al., 1995).

408 (viii) Messinian (SU7)

18
409 SU7 is bound below and above by H7 and H8, respectively. This unit generally displays a

410 chaotic seismic character with some locally bright internal reflectors that are interpreted to

411 represent re-deposited shelf deposits and different evaporite lithologies (Gardosh &

412 Druckman, 2006a; Lie et al., 2011).

413 The lower boundary of SU7 corresponds to the base Messinian and the upper boundary

414 corresponds to the top Messinian. SU7 is a distinct seismic package identified throughout

415 the basin, composed of a thick evaporitic series of Late Miocene age (Hsü, 1972; Neev,

416 1975; Ryan, 1978,2009). In the Levant Basin this unit is about 1-2 km thick (0.5-1 s TWT,

417 assuming a velocity of 4000 m/sec through salt), extending consistently across most of the

418 basin except close to the eastern margin and Eratosthenes Seamount where it pinches out.

419

420 No Messinian salt has been encountered by DSDP wells over the Eratosthenes Seamount.

421 Shallow-water carbonates (of Burdigalian age) at sites 965 and 966 have been reported by

422 Robertson (1998b) to contain extensive solution porosity. This is attributed to flushing by

423 meteoric waters during the MSC suggesting that the Eratosthenes Seamount was exposed

424 sub-aerially (Robertson, 1998c).

425

426 (ix) Pliocene to Pleistocene (SU8)


427 SU8 is bound below and above by H8 and H9, respectively. Throughout the basin, SU8 is

428 characterized by continuous, high to low amplitude reflections. At the eastern Levant

429 Margin, where the unit attains its maximum thickness, it is dominated by prograding

430 sigmoidal reflections.

431 The lower boundary of SU8 is a composite unconformity. In the basin centre, it corresponds

432 to top Messinian salt whereas on the Levant Margin, where the Messinian salt is absent, it

19
433 correlates with the Messinian Unconformity (Gardosh & Druckman, 2006a). The upper

434 boundary of SU8 corresponds to the sea-bed. Well data (e.g. Yam West-1; Figure 1) show

435 that the Plio-Pleistocene section, termed the Yafo Formation, is a mud-dominated unit

436 composed primarily of claystone and siltstone (Gvirtzman & Buchbinder, 1978).

437

438 6. Basin and Petroleum Systems Modelling

439 Basin and petroleum systems modelling simulates the evolution of a sedimentary basin

440 through time as it fills with sediments that may eventually generate or contain hydrocarbons

441 (Hantschel and Kauerauf, 2009). Building such a model requires going through two steps;

442 (i) constructing the basin model; and (ii) adding the petroleum systems components. In the

443 first step, the present day basin structure and stratigraphy is defined based on observations

444 from seismic and well data, often with the help of outcrop and analogues from similar

445 basins. The structural history and the evolution of heat flow and surface temperatures

446 through time then need to be defined. The main outcomes of this step are modelled

447 physical properties such as rock stress, pore pressure, effective stress, porosity reduction

448 by mechanical compaction, heat flow, temperature and burial history. In the second step,

449 the petroleum systems model is appended to the basin model by adding source rock

450 properties to simulate hydrocarbon generation, expulsion, migration, accumulation and

451 preservation. This simulation is fully controlled by the physical pressure and temperature

452 framework calculated in the first step. This is particularly important for the phase behaviour

453 of the modelled accumulations, which forms the main focus of this part of the study.

454

20
455 6.1 Basin Modelling

456 6.1.1 Structural evolution

457 Starting with the cross section as described above, the structural evolution of the basin

458 during pre-rift, syn-rift and post-rift times has been constrained (Figure 4), using the

459 backstripping technique (e.g., Hantschel and Kauerauf, 2009) The pre-rift basin geometry

460 is assumed to represent a broad and flat continental to shallow marine environment. This is

461 supported by the strata of the Triassic Ramon Group and the Lower Jurassic Arad Group

462 found in the southern Levant which are composed of limestone, dolomite, siliciclastic, and

463 evaporite successions (Goldberg & Friedman, 1974), suggesting that the southern Levant

464 region was located in a wide continental to shallow marine platform during this time

465 (Gardosh et al., 2011).

466

467 Syn-rift structures associated with Early-to-Mid Jurassic rifting event have been

468 reconstructed as a series of horst-grabens systems (Figure 4a). This was followed by post-

469 rift thermal subsidence, which gradually led to the development of deep-water conditions by

470 the Late Cretaceous (Figure 4b and Figure 4c). Evidence for deep-water conditions is

471 provided by pelagic carbonates that commonly contain chert of replacement origin, together

472 with localised organic-rich sediments recovered from hole (967) of the Ocean Drilling

473 Program (ODP) - Leg 160 drilled in the vicinity of Eratosthenes Seamount (Robertson,

474 1998c). In the centre of the basin (between the Eratosthenes Seamount and the Levant

475 Margin), the maximum depth was estimated to be ca. 1000 m during the Late Cretaceous

476 (Figure 4b) to Eocene (Figure 4c) as inferred from the planktonic foraminifers and

477 calcareous macrofossils present. However, carbonates growing on the intra-basinal highs

478 were apparently able to aggrade and keep pace with the thermal subsidence of the basin,

21
479 thus maintaining shallow water conditions compared to their basinal surroundings. Syrian

480 Arc folding which initiated in the Late Cretaceous was incorporated in these reconstructions

481 by developing a series of anticlines superimposing the intra-basinal highs.

482

483 During the Oligocene-to-Miocene (Figure 4d), rapid loading of deep-water turbidite sands,

484 transported by the proto-Nile River (Steinberg et al., 2011), accelerated the subsidence in

485 the centre of the basin. The Eratosthenes Seamount, however, was reconstructed to be a

486 prominent structural high during the Miocene. This is based on the Robertson (1998c)

487 interpretation that > 1000 m of vertical uplift is required to account for the shallow water

488 Miocene carbonates reported by Emeis et al. (1996b).

489

490 Basin reconstructions from the MSC to present are shown in Figure 4 e-f. The Eratosthenes

491 Seamount was exposed sub-aerially during the maximum drawdown (Figure 4e). At the end

492 of the salt deposition (5.60-5.55 Ma, Figure 4f) the water depth was assumed to close to 0

493 m, with an installation of a salt flat. At the end of the MSC (5.33 Ma, Figure 4g), the

494 subsidence in the basin accelerated due to the refilling of the basin by the Zanclean Flood

495 until reached its present aspect (Figure 4h), characterised by thick Plio-Pleistocene

496 sediments on the margin that thins basinward.

497

498 6.1.2 Modelling the MSC

499 An important part of the palaeo-geometrical evolution is the modelling of the water

500 drawdown during the MSC. As mentioned in the geological setting section (section 5.3), the

501 magnitude of the Messinian drawdown and the timing of the evaporite deposition are

502 disputed. In this model, we have applied a Messinian drawdown of ca. 2070 m. The drop in

503 sea level was assumed to have started at 5.96 Ma (Figure 5) based on high-resolution
22
504 astrochronology for the Pissouri Basin in Cyprus (Krijgsman et al., 2002). The duration of

505 the drawdown phase is assumed to be 360 kyr (ca. 5.96-5.60 Ma), and the subsequent

506 deposition of the bulk of the evaporites was assumed to have occurred within 45 kyr (5.60-

507 5.55 Ma) (Roveri et al., 2008a), resulting in a salt flat. The onset of Lago Mare deposition

508 has been modelled to be at 5.53 Ma based on Roveri et al. (2008b), accompanying a

509 progressive increase in water depth arbitrarily set to 1000m towards the beginning of the

510 Zanclean flooding (5.33 Ma), This event marks the abrupt filling of the basin with water and

511 the end of the MSC.

512

513 6.1.3 Thermal Boundary Conditions

514 The standard thermal boundary conditions used in basin modelling are the basal heat flow

515 and the sediment-water interface temperature (SWIT) through geologic time. The basal

516 heat flow in particular is poorly constrained and often introduces large uncertainties. This is

517 also the case in this study, with only limited access to direct thermal calibration data such

518 as corrected borehole temperatures and thermal maturity indicators such as vitrinite

519 reflectance.

520

521 The basal heat flow determines the amount of heat influx applied to the base of the model.

522 It is estimated through geological time based on the amount of lithospheric stretching, the

523 type and thickness of the crust, and other considerations such as deep mantle processes

524 (Allen & Allen, 2013). In our model, the heat flow history was estimated using the McKenzie

525 uniform stretching model (McKenzie, 1978). This model uses the subsidence, timing and

526 duration of the syn-rift (in our case, 176 to 145 Ma) and post-rift (145 Ma to present) phases

527 to estimate the stretching factor (and generate a palaeo-heat flow map.

23
528

529 The SWIT prediction through geological time was based on a model by Wygrala (1989),

530 which takes into account palaeo-latitude and associated palaeo-climate evolution, coupled

531 with the plate tectonic setting. The model then adjusts the SWIT for palaeo-water depth. For

532 the Levant Basin case, the predicted SWIT has been corrected further for specific

533 conditions (Figure 5) namely; (i) present-day elevated sea bed temperature of ca. 13.2 °C,

534 obtained from the Deep Sea Drilling Program (DSDP) down hole temperature measurement

535 at site 376, west of Cyprus (Figure 1) and (ii) the arid climate that characterised the

536 Messinian times. The SWIT is expected to have been much higher during the drawdown

537 than present due to the overall Messinian arid climate (Hsü, 1972; Ryan, 2009). A

538 Messinian lake floor temperature of > 35-45 °C was proposed by Hsu (1972b and 1984);

539 we therefore elected to use a Messinian SWIT of 40°C in our model.

24
540 6.1.4 Pressure and Temperature Evolution

541 The modelled present-day overpressure and isotherms is shown in Figure 6. Overpressure

542 is predicted in the Messinian Salt and the Eocene sediments due to their low permeability,

543 whereas the rest of the basin is normally pressured. Due to its very high thermal

544 conductivity, the salt also heavily perturbs the temperature, causing huge lateral variations

545 of the thermal gradient. An extraction at a pseudo-well location, roughly corresponding to

546 the lateral position of the Tamar field (Figure 6b), shows the modelled geothermal gradient.

547 The gradient in the post-salt section is > 40 ℃ km−1 , < 10 ℃ km−1 within the salt and about

548 35 ℃ km−1 in the pre-salt Oligocene-Miocene section. The temperature at ca. 4800 m

549 burial depth, which is the approximate depth of the Tamar field reservoir, is ca. 94 ℃ with

550 an overlying salt thickness of ca. 1200 m. This corresponds to temperatures published data

551 for the Tamar field (Dubille and Thomas, 2012). Closer to the margin, temperatures at the

552 same burial depth are higher (Figure 6c) This shows the importance of salt thickness in

553 controlling the local thermal gradient and, therefore, in predicting reservoir temperature and

554 source rock maturation state.

555

556 A plot of pore pressure variations within the uppermost sub-Messinian sediments around

557 Messinian times (Figure 7b) shows that the modelled pore pressure had an instantaneous

558 response to MSC-related water unloading, the deposition of salt and the flooding. The

559 pressure profiles at any particular depth exhibit similar trends for both the rate and the

560 magnitude (Figure 7c). During the drop in sea level, pressure at all depths decreased by the

561 same amount (ca. 20 MPa, equivalent to ca. 2900 Psi). Similarly, rapid salt deposition (ca.

562 0.03 m/kyr) between 5.60-5.55 Ma caused a rapid increase in pore pressure. However, this

563 pressure increase varies from ca. 10 MPa (equivalent to 1450 Psi) in the shallow sediments

25
564 to ca. 20 MPa (2900 Psi) in the deeper basin. This is interpreted to be related to different

565 permeabilities and thus different flow rates during the dewatering: whereas pore water can

566 more easily escape in the shallow Miocene sediments, this is not the case for the well-

567 compacted deeper sediments, where slow water flow is modelled to cause additional

568 overpressure (ca. 10 MPa; 1450 Psi). The deep signatures of the progressive refill and the

569 Zanclean flooding are again instantaneous 5.55-5.33 Ma).

570

571 The temperature evolution of the shallow sediments during the MSC (Figure 7b) follow the

572 SWIT, evolving from temperatures of about 10℃ to 20℃ after the sea level drop and >40℃

573 after the deposition of the salt. In contrast to pressure, the temperature response to this

574 event with increasing depth is modelled to be of transient nature (Figure 7d). Heat

575 transmission is attenuated and delayed progressively with depth, both during heating

576 accompanying the water removal and in the cooling during refill. Also the magnitude of heat

577 peak gets reduced with depth to a point that the effect is insignificant at top basement.

578

579 6.2 Petroleum Systems Modelling

580 Building up on the PT framework of the basin model described above, petroleum systems

581 modelling allows to analyse the evolution of the biogenic and thermogenic petroleum

582 systems through geologic time. Therefore, source rocks need to be assigned to the basin

583 model. A review of the petroleum systems elements; source rocks, reservoir rocks and seal

584 rocks presented in this section.

585 (i) Source rocks


586 Source rocks have been assigned to the petroleum systems model at five stratigraphic

587 levels: pre-rift (Late Triassic-to-Early Jurassic), syn-rift (Mid-to-Late Jurassic) and post-rift

26
588 (Late Cretaceous, Oligocene-to-Mid Miocene and Late Miocene; Figure 2). Because there

589 is no publically available information any possible source rock in the basin, the source rock

590 intervals listed above are selected based on the available information from nearby active

591 source rock analogues around the Neo-Tethyan Margin (Ala & Moss, 1979; Carrigan et al.,

592 1995; Alsharhan & Abd El-Gawad, 2008; Moretti et al., 2010; Bou Daher et al., 2014). This

593 is considered to be a reasonable assumption as most of the organic rich intervals are

594 believed to be laterally continuous on a basin scale. A summary of these source rocks in

595 terms of the depositional environment, kerogen type, total organic carbon (TOC) and

596 hydrogen index (HI) is given in Table 1. As can be seen from the Table 1, more than one

597 analogue is provided for each stratigraphic interval. This is to show that the source rock

598 within that particular stratigraphic interval is present in different places around the margin of

599 the basin. The kinetics used for the modelling of source rock kerogen to hydrocarbon

600 transformation are shown in Table 1. These are selected based on published kinetics

601 available in the PetroMod library that best match the Kerogen type, TOC and HI ranges of

602 source rock analogues.

603 A schematic palaeo-facies map that represents a conceptual model for the facies

604 distribution in the eastern Mediterranean Basin during the Cretaceous and the Oligocene-

605 Miocene is shown in Figure 8. During the Cretaceous, the basin was dominated by

606 carbonate platforms on the margin, and the deposition of deep-water organic-rich

607 mudstones is postulated in the basin centre during the Late Cretaceous. During the

608 Oligocene-to-Miocene, however, a deep-water turbidite system depositing high-quality

609 reservoir sands and biogenic source rocks became established over much of the basin. The

610 Oligocene source rocks are interpreted as disseminated, largely land plant organic matter

611 derived form the Nile Delta plain when Northern Africa was experiencing a more humid

612 climate during this time and transported into the deepwater Eastern Mediterranean basin
27
613 via slope turbidite channels (Vilinski, 2013). Fossilised remnants of widespread Oligocene

614 forests have been described for the onshore Nile Delta area (Dolson et al, 2001). The

615 Oligocene source has been shown by Vilinski (2013) to be the major contributor to both the

616 thermogenic and biogenic charge systems in the Nile Delta and it is likely that similar

617 sediments have been responsible for the biogenic charge in the offshore Levant.

618

619 The authors acknowledge that these parameters carry a degree of uncertainty. However,

620 the focus of this study is neither a detailed hydrocarbon mass balance, nor a quantitative

621 prediction of charge volumes. Therefore, these parameters have not been varied for

622 sensitivity analysis. It should be noted that the kinetic assigned to the Oligocene-Miocene

623 biogenic source rocks is based on a simple temperature-dependent reaction.

624

625 (ii) Regional carrier beds and reservoirs


626 Recent Lower Miocene gas discoveries in the Levant Basin (i.e. Tamar, Leviathan and

627 Aphrodite) have revealed that the Oligocene-Miocene section is composed of reservoir-

628 quality deep-water sandstones (Skiple et al., 2012; Durham, 2013). These sandstones

629 occur in laterally continuous sheets, expressed on seismic data as packages of continuous

630 and parallel reflections, extending regionally from as far south as the Nile Delta to as far

631 north as northern Lebanon (Roberts & Peace, 2007; Steinberg et al., 2011). The high

632 lateral continuity of these sandstones means they may act as effective carrier beds that

633 assist hydrocarbon migration into traps. Reservoirs have been assigned at three

634 stratigraphic intervals; Mid Jurassic, Late Cretaceous and Oligocene-Miocene (Figure 2). In

635 the Jurassic and the Cretaceous, carbonates of speculative reservoir qality were dominant

636 either in the form of carbonate buildups (e.g. Gialo filed in central Sirt Basin; Gumati &

28
637 Schamel, 1988) in the basinal highs or as carbonate platforms on the margin (Gardosh et

638 al., 2011).

639
640
641 (iii) Regional Seals
642 The Messinian Salt is considered to be the regional seal for the entire eastern

643 Mediterranean region, forming a barrier between the dominantly biogenic, post-Messinian

644 Pliocene petroleum systems and the dominantly thermogenic, pre-Messinian system in the

645 Nile Delta. Although the Messinian Salt is a very effective regional seal, it has been

646 deposited very recently and can therefore only act as a seal for hydrocarbons that have

647 migrated in the last 5 Ma.

648 For the pre-Messinian section, intra-formational shales may act as very effective seals as

649 they have been encountered in many plays around the Mediterranean. For example, shales

650 of the Kheir Formation in Libya and Souar-Cherahil Formation in Tunisia form a very good

651 seal for Palaeocene and Lower Eocene carbonate reservoirs (Macgregor & Moody, 1998).

652 In our model, we have assigned seals within the Late Cretaceous and Oligocene-Miocene

653 stratigraphic intervals. Dolson et al. (2001) state that the widespread transgression during

654 the development of the Late Cretaceous passive margin led to regional deposition of shales

655 that acted both as source rocks and seals. In the Oligocene-Miocene, the thin sheet-like

656 source rocks, inter-bedded within the turbidite sands, are assumed to act also as high

657 capillary entry pressures seal rock.

29
658 Table 1: A summary of the five source rocks used in the petroleum systems model based on analogues around the Neo-Tethyan Margin

Initial TOC Initial HI


Source rock Analogue Depositional Environment Kerogen Type
(weight %) (mgHC/g TOC)

 Type I  5-15%  >600

Miocene
Kupferschiefer Formation, North Sea  Stratified lake with restricted communication with open

Upper
(5) water systems.
(Glennie, 1998) (Glennie, 1998) (Glennie, 1998)
 No kinetics has been assigned as this source rock is
thermally immature

 Information about Kerogen type, TOC and HI are available  Based on Villinski (2013) findings from drilling (i.e.  Biogenic  1-2%  150-300
from drilling results of the Oligocene-Miocene successions Satis-1 and Satis -3) of the Oligocene successions
offshore Nile (Villinski, 2013) (Villinski, 2013) (Villinski, 2013) (Villinski, 2013)
offshore Nile Delta
Oligocene-
Miocene

 Niger or Mississippi systems could also be feasible analogue.


 Sandy turbidite basin floor fans delivered by proto-Nile
(4)
Post –rift

river.
 Kinetic is based on biogenic reaction scheme following a
Gaussian distribution (mean Temperature: 50 °C; standard
deviation 10 °C)
 Both reservoir successions and organic matters were
carried by these turbidite systems (Villinski, 2013).

 (3.1) Chekka Formation (northern Lebanon) (Bou Daher et  Broad, isolated bathymetric lows forming restricted  Type II  2%  413-610
al., 2014) euxinic basins flanked by carbonate platforms.
Cretaceous

(Bou Daher et al., 2014) (Bou Daher et al., 2014) (Bou Daher et al., 2014)
 (3.2) Sirte Shale (Libya)(Hallett, 2002)
Upper

 Deposition occurred in an outer-shelf setting.


(3)

 (3.3) Bahloul Formation (Tunisia) (Klett, 2001)

 (3.4) Brown Limestone (Gulf of Suez) (Alsharhan, 2003)

 Kinetics is based on di Primio and Horsfield (2006)

 Laminated organic-rich lime mudstone deposited in a  Type II  3%  600-800


relatively short lived intra-shelf basin.
 (2.1) Hanifa and Tuwaiq Formation (Saudi Arabia) (Carrigan (Carrigan et al., 1995) (Carrigan et al., 1995) (Carrigan et al., 1995)
et al., 1995)  The basin was partially separated from neo-Tethys
Ocean by flanking paleo-highs composed of grainstone
Upper Jurassic

shoal/ barrier island facies.


Syn-rift

(2)

 Composed mainly of shales, sandstones with coal seams  Type I/II  0.11-3.5%  36-766
and minor limestones deposited in a deltaic
 (2.2) Khatabta Formation (South Alamein Western Desert) environment. (Alsharhan & Abd El- (Alsharhan & Abd El- (Alsharhan & Abd El-
(Alsharhan & Abd El-Gawad, 2008) Gawad, 2008) Gawad, 2008) Gawad, 2008)

 Kinetics used for the Upper Jurassic is based on Ungerer


(1990)
Upper Triassic

 Kurra Chine Formation (eastern Syria) (Jassim & Goff, 2006)  Restricted intra-shelf basins with deposition of  Type II and III  3%  500
condensed sections during the early transgressive phase.
Pre-rift

 Kinetics used in the modeling is based on (Ungerer, 1990) (Jassim & Goff, 2006) (Jassim & Goff, 2006) (Jassim & Goff, 2006)
(1)

30
659 6.2.1 Source rock maturity

660 Due to the very high uncertainty of the kinetics (describing the kerogen to

661 hydrocarbon transformation) chosen for the modelling of the thermogenic

662 source rocks, we prefer describing general thermal maturity windows, based

663 on source rock kinetic-independent vitrinite reflectance calculation (Sweeney

664 and Burnham, 1990). The present-day thermal maturity as modelled is shown

665 in Figure 9a. Any potential Upper Triassic source rocks are assessed to be in

666 the “gas window”, as are some of the Lower Jurassic source beds. Upper

667 Jurassic source rocks are modelled to be currently in the “oil window” and, if

668 present, might be the most effective thermogenic source rocks at present-day.

669 The model suggests that potential Oligocene to Miocene source rocks are

670 immature for thermogenic oil and gas generation. The modelled evolution of

671 the thermal maturity of the source rocks intervals (Upper Triassic/Lower

672 Jurassic, Upper Jurassic, Upper Cretaceous, Lower Miocene and

673 UpperMiocene (Messinian)) is shown in Figure 9b. The rift-related heat peak

674 (black line) affects mainly the Upper Triassic (pink line) and marginally the

675 Jurassic (blue line) interval. General burial-related temperature increase is

676 modelled in post-rift times, accentuated during the Upper Miocene and, in

677 particular, during the Pliocene-Pleistocene.

678 As discussed above, the combined effect of elevated surface temperatures

679 and rapid burial by salt deposition during the MSC triggered a subsurface

680 temperature increase, which became less significant with depth. The

681 petroleum systems modelling, indicates that the thermogenic source rocks are

682 only marginally affected by the MSC (Figure 9c) since it is limited to a
31
683 relatively minor increase in thermal maturity in the still immature Oligocene-

684 Miocene. The impact of the subsurface temperature increase is however

685 much more significant on the biogenic system. Since the simple temperature-

686 dependent kinetic applied to the biogenic source rocks is very sensitive to

687 changes in the range 30 to 70°C, which corresponds the thermal window the

688 shallow Miocene sediments are in during the MSC. Dependent on the pre-

689 Messinian state of maturity, kerogen to gas transformation ratios can increase

690 by up to 80% during the MSC. Deeper Oligocene source rocks are not

691 modelled to be affected since they have been buried deep enough and

692 therefore have gone through the narrow biogenic gas generation window prior

693 to the MSC.

32
694 6.2.2 Shallow gas accumulations

695 The high sensitivity of biogenic source rocks to the rapid heating of the shallow sub-

696 surface during the Messinian sea-level drop lead to a massive gas generation rate

697 (Figure 9d). The modelling results further suggest that, after some of the newly

698 generated gas was adsorbed by the source rocks and saturated their pore space as

699 free gas, the majority of it was expelled rapidly from the source beds. In addition, the

700 sea-level drop destabilized possibly pre-existing gas accumulations, since the pressure

701 drop lead to a gas density decrease (gas volume increase), a rapid desorption of gas

702 from the source rocks and exsolution of gas previously dissolved within the pore water.

703 All those sudden changes are modelled to have had as a common consequence the

704 gas flooding of the sub-Messinian strata during the MSC. At the same time, the storage

705 capacity for that gas went dramatically down, with the load of the precipitating salt

706 leading to a rapid compaction of the sub-Messinian sediments.

707 Those effects have not been further quantified, since too many important unknowns

708 exist in particular around the characterization of the biogenic source rock (spatial

709 distribution, kinetic, richness, adsorption capacity). However, it can be stated that the

710 combined effect of newly generated biogenic gas, the destabilization of previously

711 accumulated gas and the modification of the physical environment lead to a massive

712 gas charge in a setting, which cannot easily contain the additional gas volumes.

713 The consequence of the gas flood is assumed to have been regional fracturing of

714 stratigraphically concentrated around the biogenic source rocks and gas-bearing layers,

715 triggering accelerated gas expulsion and migration. Whereas the gas has been

716 modelled to have migrated predominantly vertically prior to and during the sea-level

33
717 drop, with an accelerated surface leakage, the load of the Messinian salt fundamentally

718 changed that pattern. Since in the model the Messinian salt has been assumed

719 impermeable, the underlying sediments became quickly overpressured. From that

720 moment on, the gas was modelled to migrate laterally and escape to the sea floor along

721 the depositional edges of the salt. Similarly to the temperature effect, the overall

722 influence of the MSC on the shallow gas system is modelled to diminish with depth.

723 Sediments of the Upper Miocene are more exposed to the MSC, which is due to several

724 factors. Firstly, shallow biogenic source rocks had more gas generation potential prior to

725 the MSC as opposed to already deeper buried biogenic source rocks, which were

726 already partially to totally transformed. In addition, the temperature changes due to the

727 sea-level drop, which triggered the gas generation according to the simple kinetic used

728 for modelling, are of higher magnitude in those shallow sediments (Figure 7d).

729 Furthermore, porosity reduction due to the rapid compaction by the salt (Figure 7b), and

730 related fluid migration are accentuated in the shallow sediments, since they were

731 relatively freshly deposited, as opposed to partially compacted buried rocks. Finally,

732 since the gas escape occurred predominantly vertically, upper sediments were not only

733 exposed to the destabilization effects of “local” biogenic gas, but also received all the

734 fluids the deeper sediments could not hold.

735 Figure 10 presents the modelled present day hydrocarbon saturation and the migration

736 vectors of the last timestep. The general basin-wide fluid escape system can be

737 observed, with regional migration trends from the centre towards the eastern and

738 western edges of the Messinian salt, where fluids leak to the sea floor. Biogenic gas

739 accumulations have been modelled in the structural highs in the Miocene and Oligocene

34
740 levels, with saturations which progressively increase with depth. It can therefore be

741 interpreted that, as a result of the above-mentioned processes, the “preservation

742 window” of biogenic gas accumulation starts at a depth of approximately 1km below the

743 base of the Messinian salt. This “line” divides the Oligocene to Miocene into a domain

744 where the MSC had rather destructive effects on the potential for biogenic gas and a

745 domain where these effects where attenuated. All current discoveries are located below

746 that line when projected on the modelled cross section (Figure 10b).

747

748

749 6.2.3 Impact of the MSC on existing oil accumulations

750 Possible Mesozoic source rocks, yet to be proven, are modelled to be in the oil window

751 or below (Figure 9a) mostly before the MSC (Figure 9b). Thermogenic generation of

752 hydrocarbon is modelled, and the migration results indicate deep oil accumulations in

753 the Cretaceous focussed on the structural highs of the basin (Figure 10a). At the crest

754 of those structures and at the basin edges, the oil is modelled to leak into the Cenozoic.

755 That suggests the possibility of a mixture of thermogenic hydrocarbon, predominantly

756 oil, and biogenic gas in the Oligocene to Miocene sediments. In the following section,

757 hypothetical accumulation cases and their possible reaction to the MSC are discussed.

758 The authors insist on the fact that the model has been set up in a way to create those

759 “critical mixtures” to test and showcase their general behaviour during the MSC, and to

760 propose a methodology of risking the effect of the MSC on possible accumulations. In

761 no case, the model accumulations discussed below should be confused with existing

762 fields, discoveries or prospects.

35
763

764

765 (i) Phase Change

766 The phase of a sub-surface accumulation, i.e. which hydrocarbon components (oil, gas)

767 are in liquid or vapour state, is dependent on the in situ pressure and temperature (PT;

768 Dandekar, 2013). Dependent on the initial composition of an oil accumulation, PT

769 changes during production often causes phase changes. When the reservoir pressure is

770 depleted, gas previously contained in the oil (gas component in liquid phase, or

771 “associated gas”) can exsolve to form a gas cap. The same phenomena, at different

772 time scales, occur during the geological history. Generally, changes in pressure and

773 temperature are very progressive at geological timescales. However, the sea-level drop

774 associated to the MSC occurred can be considered as a geological “moment” due to its

775 very short duration (5.96 to 5.6 Ma).

776

777 Figure 11a shows the same PT extraction for one model cell (Upper Miocene at the

778 vicinity of the projected Tamar field) as in the time extraction Figure 7b, but in a PT

779 diagram. The order of the extraction points reflect the geological time aspect, each point

780 represents a 10 kyr modelling step. The following trends can be seen: pressure drop

781 and heating (1) during the sea-level fall, pressure increase and further heating by the

782 burial of the salt (2), followed by a progressive pressure increase and cooling phase

783 during the deposition of the Lago Mare Formation (3) and a very rapid pressure

784 increase as a response to the flooding event (4), and finally the post-MSC heating at a

36
785 near-constant pressure due to the post-MSC burial (5). On the same PT diagram, any

786 phase diagram of a possible analogue oil, in that particular case from the Al Shaheen

787 field offshore Qatar (Lindeloff et al., 2008, Al Ghafri et al., 2014), can plotted. The Qatar

788 crude oil has been selected as an analogue because it represents one of the Neo-

789 Tethyan margin basins that shares a very similar tectonic evolution history to the

790 eastern Mediterranean. More importantly, this crude is produced from the Cretaceous

791 carbonate reservoirs, which are speculated to be active, but as yet untested, in the

792 eastern Mediterranean. The bubble point curve of that phase diagram separates the

793 liquid domain from the mixed liquid and vapour domain, On this combined plot it can be

794 analysed when the PT path of the model cell crosses the bubble point line, and

795 therefore when and by how much the corresponding oil mixture is expected to change

796 phase, i.e., starts to build a gas cap.

797

798 PT extractions of other model cells, either at different locations along the model section

799 or at different depth, can be plotted on that diagram. Those locations could correspond

800 to existing fields, discoveries or prospects. None of those accumulations need to be

801 specifically simulated, since only the results of the pressure and temperature modelling

802 are extracted from the model. The PT plot on Figure 11b illustrates two important points.

803 Firstly, the first-order control on phase change during the MSC is not temperature

804 variation, but pressure changes. The pressure drop depends directly on the amount of

805 sea level drawdown assumed in the model. The two-step pressure increase is due to

806 the loading by the salt followed by the refill and flooding. Secondly, and even more

807 importantly, is the observation that not all existing pre-Messinian hydrocarbon

37
808 accumulations are sensitive to the modelled PT changes. Deep accumulations are

809 much more likely to be ‘protected’ from gas exsolution than shallow ones. However, the

810 main variable is the hydrocarbon composition and the corresponding phase diagram.

811 Neither a typical black oil containing hardly any associated gas (Dandekar, 2013,

812 pp.135) nor a pure gas accumulation already in the vapour phase would react to the

813 MSC with a phase change. Only specific hydrocarbon mixtures are concerned, and the

814 closer the reservoir PT conditions were to the bubble point curve prior to the MSC, the

815 more likely it was to react by gas exsolution (“bubble point oil”;Dandekar, 2013).

816 Obviously, many different possible analogue oils can be plotted in the diagram, together

817 with an uncertainty envelope.

818

819 (ii) Impact of the Phase Change

820 The combined PT plot (Figure 11b) can provide a good understanding whether a

821 possible accumulation in given model location, or trap, might have been subject to

822 phase change or not. However, a phase change can have drastically different

823 consequences on that trap, dependent on controlling parameters of the trap fill. To

824 investigate how phase changes could have affected pre-Messinian hydrocarbon

825 accumulations, three trap scenarios have been modelled, where each represents a trap

826 fill-controlling parameter: (A) amount of charge, or “charge limited”, (B) seal strength, or

827 “seal limited”, and (C) trap size, or “trap limited” (Figure 12). Those traps have been

828 filled by a “near bubble point” hydrocarbon mixture of mainly thermogenic oil with some

829 biogenic gas has been created to fill the trap, referred to as “reactive accumulations”.

38
830 The oil has essentially been sufficiently enriched with gas that those accumulations can

831 undergo gas exsolution during the MSC and its consequences can be modelled. The

832 migration method used in the simulation is Invasion Percolation (Wilkinson, 1984,

833 Hantschel and Kauerauf, 2009).

834

835 Scenario A: Charge-limited


836
837 A charge-limited accumulation represents an under-filled trap because of the little

838 amount of hydrocarbon charge it received with respect to the available pore volume,

839 which is assumed to be much larger than the charge amount in that scenario. The seal

840 is assumed to be very strong. Scenario A showcases the evolution of a reactive

841 accumulation within a charge-limited trap which experiences a phase separation and a

842 related increase in volume and column height during drawdown. At the peak of the

843 MSC, two phases could conceivably coexist because the fluid expansion was

844 accommodated by the under-filled and large trap. After the peak of drawdown, the

845 accumulation volume is modelled to decrease and the two phases to merge again into a

846 single liquid phase. Since the trap is large enough and the seal is effective, no

847 hydrocarbon losses are modelled to occur, and the post-Messinian composition is

848 comparable to the composition prior to the MSC.

849
850 Scenario B: Seal-limited
851
852 A seal-limited accumulation represents an under-filled trap because the buoyancy

853 pressure of the hydrocarbon column is in equilibrium with the capillary entry pressure of

854 the seal, and no further increase of the column height is supported. Scenario B

39
855 describes how a reactive accumulation within a seal-limited trap experiences a phase

856 separation and a related increase in volume and buoyancy pressure, leading to leakage

857 through the top seal. Theoretically, with preferential leakage of vapour gas, there is a

858 corresponding relative enrichment in the liquid oil component. However, because the

859 rates of buoyancy increase imposed by the rapidly falling sea level are abnormally high,

860 the maximum rate of capillary leakage is likely to be reached relatively quickly. If the

861 seal cannot release the vapour gas as fast as the buoyancy increases, it fractures and

862 there could be catastrophic seal failure.

863
864 Scenario C: Trap size-limited
865
866 A trap size-limited accumulation represents a trap which is filled to spill, with the

867 hydrocarbon charge amount exceeding the available pore space and a very strong seal.

868 Scenario C shows a reactive accumulation within a size-limited trap that is filled-to-spill

869 prior to the MSC. The accumulation experienced a phase separation and a related

870 increase in volume and column height during drawdown. The vapour cap (mostly gas

871 components) expels the liquid “leg” (mostly oil components) downwards via the

872 structural spill point and relative enrichment of vapour gas components occurs. Once

873 the lake level rises again, the increase in pressure compressed the gas cap, resulting in

874 an under-filled trap with higher gas content, possibly with a remaining gas cap if not

875 enough oil is present any more, than prior to the MSC.

876

877 It should be noted that these scenarios represent end-member cases, whereas in

878 nature, the controlling parameters on trap fill are likely to evolve. However, the three

40
879 modelling scenarios demonstrate that those parameters need to be individually

880 investigated to increase the understanding of the present day composition and column

881 height of fields and discoveries, and in the assessment of leads and prospects.

882

883 7. Discussion

884 The presented modelling results suggest that the sea-level fall associated to the

885 beginning of the MSC triggered a large drop in sub-surface pore pressure, immediately

886 and deeply penetrating the basin, accompanied by a surface temperature increase,

887 which attenuated with depth. These changes in PT conditions are modelled to

888 dramatically impact both the shallow biogenic and deeper thermogenic petroleum

889 systems, resulting in sudden fluid-related fracturing events which are stratigraphically

890 concentrated (around biogenic source rocks and gas bearing layers) and structurally

891 controlled (centred around the crest of paleo-accumulations). In particular, the effect on

892 traps with seal-limited “bubble point oil” accumulations are modelled to have triggered

893 catastrophic fluid escape events. Evidence for such fluid events is present in our data,

894 and has been documented in the field from several circum-Mediterranean locations

895 (Iadanza et al., 2013)

896

897 Pockmarks are geomorphic expression of focused subsurface fluid flow and common

898 features on the ocean floor (King & MacLean, 1970; Schroot et al., 2005; Judd &

899 Hovland, 2007). Ancient (i.e. buried) (e.g. Bizarro, 1998; Gemmer et al., 2002;

900 Andresen et al., 2008) pockmarks are also observed in many basins, and they are

901 commonly indicative of the escape of gas-rich fluids to the palaeo-seafloor. Ancient

41
902 pockmarks are thus critical elements in establishing the fluid flow history of a

903 sedimentary basin (e.g. Dimitrov, 2002; Loncke et al., 2004; Dupré et al., 2010). Buried

904 depressions occur in our dataset at the base Messinian surface, offshore Western

905 Desert, Egypt (section BB’, Figure 1) in present water depths of ca. 3000 m. In cross

906 section (Figure 13), the depressions are characterised by symmetrical ellipsoid, ca. 2-3

907 km wide and about ca. 500 m deep, which truncate underlying reflections. Although

908 sub-salt imaging can be challenging and is often associated with acoustic distortions

909 effects that can distort the primary geometry of underlying seismic reflections, such

910 acoustic disturbance has been documented above the South Arne field in the North Sea

911 where they are interpreted to be related to fluid escape features (Andresen et al., 2008).

912 Based on their similar geometry to the depressions observed in North Sea and

913 elsewhere (Pilcher & Argent, 2007; Andresen & Huuse, 2011; Ostanin et al., 2012), we

914 interpret that these depressions are pockmarks that are caused by focussed fluid flow

915 from deeper, pre-Messinian structures.

916

917 Our interpretation suggests that these pockmarks most likely formed in response to

918 focused fluid flows supported by the observation that other base Messinian pockmarks

919 are observed in 3D seismic data from the nearby Levant Basin (Lazar et al., 2012;

920 Bertoni et al., 2013). More than 35 pockmarks, ranging in diameter from 100 to 2000 m,

921 have been mapped at the base of the pre-Messinian Afiq Canyon by Bertoni et al.

922 (2013). Bertoni et al. (2013) interpreted these pockmarks as to have formed in response

923 to gas venting occurring during the Messinian drawdown (cf. Fraser et al. (2011)).

42
924 The reaction of hydrocarbon to changes in overburden load has been described in the

925 Barents Sea, offshore northern Norway (Henriksen et al., 2011), where regional uplift

926 occurred due to the isostatic rebound following Plio-Pleistocene deglaciations. Fluid and

927 gas escape to the seafloor is manifested by two giant pockmarks ca. 1-2 km wide in

928 addition to > ca. 300 pockmarks are mapped beneath the seabed on the Upper

929 Regional Unconformity. Six of these pockmarks are sufficiently large (ca. 1-2 km in

930 diameter) to be classified as ‘giant’. Nyland et al. (1992) and Skagen (1993) argue that

931 the pressure release during overburden unloading had a direct impact on existing

932 hydrocarbon accumulations in Hammerfest Basin. Similar to the size-limited trap

933 scenario presented in this study, the combined effect of gas exsolution from the oil

934 column and resulting gas cap expansion, is thought to have displaced the oil via the

935 structural spill point. This would have resulted in spilling of significant amounts of oil

936 from existing traps (Nyland et al., 1992; Henriksen et al., 2011). Residual oil columns

937 ca. 100 m below the present-day oil-water contact has been reported from the Snøhvit

938 gas field in the Hammerfest Basin (Henriksen et al., 2011) and might be related to the

939 gas cap expansion and oil spill mechanism.

940

941 Successions of vertically stacked pockmarks have also been documented in the Lower

942 Congo Basin, offshore West Africa (Andresen & Huuse, 2011). The timing of formation

943 of these pockmarks appears to coincide with periods of relative sea-level fall, leading

944 Andresen and Huuse (2011) to suggest that water load removal and subsurface

945 depressurisation was the key trigger for fluid escape and pockmark formation (Andresen

946 & Huuse, 2011).

43
947

948 Outcrop evidence for MSC-related palaeo-fluid escape comes from a Late Messinian

949 limestone-bearing unit (Brecciated Limestones Unit) exposed in Maiella, central Italy.

950 Iadanza et al. (2013) suggest upward migration of hydrocarbon-rich fluid through the

951 Messinian succession accompanied by brecciation was triggered by the sudden

952 depressurization associated with Messinian drawdown. The migration of hydrocarbon-

953 charged fluids is proven by distinct tar impregnation in the carbonate microfacies,

954 suggesting that the area experienced two main phases of fluid flow (Iadanza et al.,

955 2013). The early phase was characterised by light hydrocarbon fluid migration

956 ascending at high flow rates; this induced widespread brecciation of the overlying

957 sedimentary column. During the Late phase, oil and heavier hydrocarbons migrated

958 upwards, presumably assisted by the previously formed fluid flow pathways (Iadanza et

959 al., 2013).

960

961 The lack of general understanding of biogenic gas generation and its relatively weak

962 implementation in petroleum systems modelling software limits the predictions. The

963 “temperature only”-dependent source rock kinetic presents an extreme simplification.

964 The biogenic gas kinetic published by Middelburg et al. (1991), which mainly depends

965 on sedimentation rate, has been applied to the model. However, due to the lack of high-

966 resolution dating of the sub-Messinian sediments, which are necessary to define a

967 sedimentation rate at the resolution of the modelled process, the modelling results have

968 been judged not trustable. In addition, even if high resolution dating of the Miocene

969 sediments were available, since those are in great burial depth at present day, the

44
970 decompaction necessary to restore depositional sedimentation rate adds another big

971 uncertainty, which might be well above the resolution of the results. It should be noted

972 that the kinetic by Middelbourg (1995) does not have any temperature dependency.

973 However, even though the impact of the MSC on biogenic gas generation is

974 questionable due to a simplified kinetic, the other mechanisms modelled and described

975 above should be sufficient enough to heavily destabilise the biogenic gas accumulations

976 during the MSC.

977 Another big uncertainty, heavily disputed in literature, are the amplitude of the sea level

978 drop, the relative chronology and absolute timing of the MSC. The way the MSC has

979 been implemented in the basin model (Figure 5) presents only one possible simplified

980 scenario, with uncertainties are at many levels. Where no fundamental impact on the

981 modelling results has been assumed, e.g. a single phase sea-level drop instead of a

982 cyclic sea-level drop, those simplifications were taken into account without further

983 sensitivity analysis. One major uncertainty to be further investigated concerns the

984 relative timing between the sea-level drop and the salt precipitation, since the sea-level

985 drop triggers fluid escape, whereas the sealing salt is more likely to retain some of the

986 fluids, or at least complicate their way to the basin floor. In reality, the Messinian “salt” is

987 not as clean and sealing as assumed in the modelling, and recent observations of intra-

988 Messinian fluid escape structures (Eruteya et al., 2015) suggest a more complex

989 migration pattern, where fluid chimneys initiated during the MSC were reused until

990 recent times. However, the overall impact of the MSC on both the shallow biogenic

991 system and deeper oil accumulations remains undisputable and should be considered

992 in the individual understanding and risking of leads and prospects in the eastern

45
993 Mediterranean Sea, and possibly in other basins where similar events are reported to

994 have occurred.

995

996 8. Conclusions and implications for oil and gas exploration

997 In this study, 2D basin and petroleum systems modelling has been conducted to

998 demonstrate the impact of the geologically instantaneous Messinian Salinity Crisis

999 (MSC), with a water drawdown, subsequent salt deposition and re-flooding of the basin,

1000 on the proven biogenic and speculative thermogenic petroleum systems of the eastern

1001 Mediterranean Sea. The study demonstrates that significant changes in subsurface

1002 pressure and temperature are likely to have occurred. The sea level fall of 2070 m is

1003 assumed to have taken place in an extraordinarily short time period (i.e 360 kyr) and

1004 caused the in situ pore pressure of underlying sediments to drop by ca. 20 MPa

1005 (equivalent to ca. 2900 Psi) The pressure drop is modelled to be instantaneous and to

1006 affect even deeply buried layers. In contrast, the model suggests that the effect of the

1007 MSC on temperature was more transient, i.e. slightly delayed in time and attenuated

1008 with increasing burial depth. The deposition of ca. 1500 m of Messinian salt is assumed

1009 to have followed the sea-level drop, partly reloading and therefore pressurising,

1010 compacting and heating the underlying sediments very rapidly (45 kyr). The basin has

1011 been refilled with water in two steps, firstly a progressive fill while depositing the Lago

1012 Mare Formation, and finally the abrupt Zanclean flooding event, which marks the end of

1013 the MSC.

1014

46
1015 These changes in pressure and temperature are modelled to have severely affected the

1016 shallow sub-Messinian biogenic petroleum system. Rapid compaction and heating, an

1017 accelerated gas generation and expulsion, gas density changes and a decrease of

1018 adsorption capacity and water solubility are assumed to have led to fluid brecciation

1019 stratigraphically concentrated around biogenic source rocks and gas-bearing layers and

1020 triggered massive fluid escape events. It appears that the Upper Miocene is far more

1021 affected than the deeper Miocene and Oligocene, indicating a “preservation window” of

1022 biogenic gas accumulations starting at ca. 1km below the Messinian salt. All recent gas

1023 discoveries (Tamar, Leviathan, Aphrodite) are located within this window. Furthermore,

1024 the deeply penetrating pressure drop is modelled to have caused shallower, pre-

1025 Messinian hydrocarbon accumulations with a specific “bubble point oil” composition to

1026 undergo a phase change, i.e. to build a gas cap by exsolution from the oil. However,

1027 even though affected similarly by the pressure drop, deeper accumulations with similar

1028 composition are more likely to have been preserved from such phase changes, since

1029 accumulations in highly pressured depths are assumed to be always in the liquid PT

1030 domain. The impact of the phase change on the pre-Messinian hydrocarbon

1031 accumulations was found to be dependent on the trap-fill controlling parameters: seal-

1032 limited, size-limited and charge-limited. In the former case, catastrophic seal failure has

1033 been modelled as the most likely consequence of the rapidly increasing gas column

1034 triggered by the MSC. Size limited-traps, in comparison, were predicted to experience

1035 tertiary hydrocarbon migration, with the growing gas cap pushing the liquid oil down to

1036 spill out of the trap. In the latter case of charge-limited traps, the accumulations are

1037 likely to be of similar size and composition after the MSC.

47
1038

1039 This research focused on the Levant Basin, although it has implications for many of the

1040 Mediterranean basins affected by the MSC (e.g. Sirt Basin, offshore Libya). The authors

1041 strongly recommend that any future exploration activity should integrate the MSC in

1042 basin and petroleum systems models and prospect risk analysis. A comprehensive

1043 methodology has been suggested with the use of pressure and temperature extractions

1044 at prospect locations, in conjunction with phase diagrams of possible oil analogues.

1045 Shallow traps estimated to be charged prior to the MSC and capped by weak seals are

1046 predicted to be very high-risk prospects, especially if there are associated with palaeo-

1047 pockmarks in the overburden. Mapping these palaeo-pockmarks may therefore be a

1048 very useful tool to delineate the high and medium-risk prospects. Medium risk prospects

1049 are expected in similar traps within the vicinity of these palaeo-pockmarks but not

1050 directly beneath them. In contrast, deep accumulations are classified as low risk

1051 prospects as they are less likely to have been affected by the MSC drawdown.

1052

1053 Acknowledgments

1054 Petroleum Development Oman (PDO) is thanked for providing the funding for the first

1055 author PhD project at Imperial College London. British Petroleum Egypt is thanked for

1056 providing the data and allowing this study to be published. Schlumberger is also

1057 thanked for providing Petrel and PetroMod software. In particular, we would like to thank

1058 Bjorn Wygrala, Thomas Hantschel, Armin Kauerauf, Thomas Fuchs, Michael Fuecker

1059 and Adrian Kleine (Schlumberger Technology Centre at Aachen) for useful discussion.

1060 Saif Al-Ghafri from the department of chemical engineering at Imperial College is also

48
1061 acknowledged for providing the phase envelope for the Qatari crude oil we used in this

1062 study and for the useful discussion on the phase behaviour of reservoir fluids.

1063

1064

1065

1066

1067
1068
1069 Figure Captions

1070

1071 Figure 1: Location map of the eastern Mediterranean Basin showing the study area, the

1072 database available for this study and the distribution of the hydrocarbon fields (mainly

1073 gas) in the Nile Delta region. Blue lines show the seismic profiles we have used. Profile

1074 A-A’: 235 km long seismic profile that has been used to construct the petroleum

1075 systems model. The recent gas discoveries (Tamar, Leviathan and Aphrodite) in the

1076 vicinity of this seismic profile are indicated by the red squares numbered 1, 2 and 3

1077 respectively. Profile B-B’: 35 km long seismic profile showing evidence for pockmarks at

1078 base Messinian. Blue circles show the location of the Deep Sea Drilling Program

1079 (DSDP) holes drilled over the Eratosthenes Seamount and to the west of Cyprus.

1080

1081

1082 Figure 2: Tectono-stratigraphic chart summarising the pre-rift, syn-rift and post-rift

1083 geologic history of the present-day Levant Basin. The key seismic horizons (H1-H9)

1084 bounding the major seismic units (SU1-SU8) as described in this paper are shown in

49
1085 the chart. The stratigraphic intervals of the petroleum systems elements (source,

1086 carrier/reservoir and seal) assigned to the petroleum systems model are also shown.

1087

1088 Figure 3: (a) uninterpreted, (b) partially interpreted and (c) interpreted seismic reflection

1089 profile and (d) the model used in this study. The approximate location of well control

1090 from the Eratosthenes Seamount and the Levant Margin are shown in (a). Horizons with

1091 dashed lines and yellow question marks in (b) indicate areas where the data quality

1092 deteriorates and hence the interpretation is less certain.

1093

1094 Figure 4: A series of structural reconstructions describing the evolution of the Levant

1095 Basin through time with particular emphasis on the MSC period (reconstructions from

1096 (d) to (g)). (a) In the Late Jurassic, the basin is dominated by shallow marine carbonates

1097 with carbonate build-ups on the structural highs. (b) In the Late Cretaceous,

1098 convergence inverted some of the earlier rift structures and caused the development of

1099 anticlines. (c) In the Late Cretaceous to Eocene, organic rich source rock and chalk

1100 were deposited. (d) At the start of the MSC (5.96 Ma), the basin was about 3 km deep.

1101 (e) At the time of the maximum drawdown (5.60 Ma), the water depth was only about 1

1102 km and Messinian shale has already been deposited. The Eratosthenes Seamount was

1103 exposed subaerially during this time. (f) At end of salt deposition (5.55 Ma), the water

1104 depth was maintained at about 1 km and the Eratosthenes Seamount is in subaqueous

1105 condition because of the subsidence associated with salt deposition. (g) At the end of

1106 the MSC (5.33 Ma), the basin continued subsiding due to the water loading associated

50
1107 with the Zanclean flood. (h) At present-day, the Messinian salt is overlain by Plio-

1108 Pleistocene sediments that are thicker on the margin and thinner in the basin centre.

1109

1110

1111

1112 Figure 5: Schematic diagram showing how the MSC was incorporated in the modelling.

1113 Sketch sections from (1) to (4) summarise the sequential order of the MSC

1114 development. (1) Is the interpretation of the pre-Messinian setting in the basin with sea

1115 level at normal Atlantic levels. (2) The MSC is modeled to have started at 5.96 Ma

1116 where sea level started dropping below Atlantic level. The maximum drawdown of c.

1117 2070 m was reached at 5.60 Ma and we assumed that the water depth is c.1000 at the

1118 time of maximum drawdown. During this time (5.96-5.60 Ma), as the circulation in the

1119 basin became restricted, the Messinian shale was modelled to have been depositing as

1120 indicated by the thin black layer. (3) The Messinian salt deposition (c. 1500 m) was

1121 modeled to have occurred between 5.60-5.55 Ma. During the salt deposition, we

1122 maintained the water depth at c. 1000 m by steadily increasing the Messinian lake level

1123 as indicated by the red arrows. After Messinian salt deposition, the lake level continued

1124 to rise, and the deposition of the Lago-Mare facies was modeled to take place between

1125 5.6-5.33 Ma. (4) The end of MSC which was modelled by the Zanclean flooding to be at

1126 5.33 Ma. The Zanclean flooding was incorporated in our model by filling the basin

1127 instantaneously as indicated by the very steep slope of Messinian sea level curve.

1128 Subsequent Plio-Pleistocene deposition occurred from 5.33 Ma to present-day.

51
1129 Pressure and SWIT conditions associated with these steps are indicated on the

1130 diagram.

1131

1132 Figure 6: (a) The present-day overpressure and temperature as predicted by our basin

1133 model. Overpressure field as shown by the overlay (scale is in bottom left of the figure)

1134 predicts that the Messinian salt and Eocene are overpressured. Temperature variation

1135 across the profile is shown by the isolines (in yellow). To compare the effect of the

1136 Messinian salt on temperature across the profile, geothermal gradients have been

1137 extracted at the basin centre (b) and the basin margin (c). In the basin centre (b) where

1138 the salt is about 1200 m thick, the temperature is predicted to be c. 85°C at an

1139 approximate depth of c. 4500 m. At the basin margin (c) where the salt is thinner, the

1140 temperature is predicted to be c. 125°C for the same burial depth.

1141

1142 Figure 7: Pressure (c) and temperature (d) time extraction plots between 6.5 - 4.5 Ma

1143 at four different depths: Basement (7000m); red, Top Mesozoic (3650m); green,

1144 Oligocene/Miocene (Tamar: 1800m); brown and Top Miocene (close to surface);

1145 orange. On diagram (b), temperature (red line) shows a transient response to water

1146 unloading and evaporite loading whereas pressure (blue line) shows an instantaneous

1147 response.

1148

1149 Figure 8: Schematic palaeo-facies maps for the post rift megasequence in the eastern

1150 Mediterranean. (a) Carbonate dominated passive margin during the Cretaceous. (b)

1151 Clastic dominated margin (Oligocnene-Miocene). Source rocks are predicted in the

52
1152 Cretaceous deep-water deposits and Oligocene organic rich turbidite sands. The

1153 Oligocene and later Miocene Nile derived turbidites also act as major lateral carrier

1154 beds in the basin.

1155

1156 Figure 9: Present-day thermal maturity as modelled in this study. (a) An overlay of

1157 vitrinite reflectance (key in the bottom left) shows the maturation stages within various

1158 stratigraphic intervals. (b) The thermal maturity evolution. The black solid line shows the

1159 heat flow variation (input to the modeling) associated with the Middle Jurassic -Early

1160 Cretaceous rifting based on McKenzie stretching model. The modelled thermal

1161 evolution (dotted line: temperature; solid line: vitrinite reflectance is shown at the

1162 stratigraphic level of potential source rocks (i.e. Upper Triassic: pink, Upper Jurassic:

1163 blue, Upper Cretaceous: green, Miocene-Oligocene: orange, Upper Miocene

1164 (Messinain): purple. The location of the extractions along the cross section are shown in

1165 Figure 6. (c) A zoomed in view of plot (b) showing the Late Miocene (6.0-5.30 Ma)

1166 interval. (d) Transformation ratios for biogenic source beds. Kinetics modelled after

1167 Middleburg et al. 1991.

1168

1169 Figure 10: (a) The modelled thermogenic generation of hydrocarbons in the Levant

1170 Basin. The migration results indicate deep oil accumulations in the Cretaceous focused

1171 on the structural highs in the basin. At the crest of these structures the oil is modeled to

1172 leak vertically into the Cenozoic. (b) Modelled biogenic accumulations in the Levant

1173 Basin sitting above the 1km sub base Miocene salt preservation window.

1174

53
1175 Figure 11: (a) shows the modelled pressure-temperature extraction for one model cell

1176 (Upper Miocene at the vicinity of the projected Tamar field) as in the time extraction

1177 Figure 7b, but in a PT diagram. The order of the extraction points reflect the geological

1178 time aspect, each point represents a 10 kyr modelling step. The following trends can be

1179 seen: pressure drop and heating (1) during the sea-level fall, pressure increase and

1180 further heating by the burial of the salt (2), followed by a progressive pressure increase

1181 and cooling phase during the deposition of the Lago Mare Formation (3) and a very

1182 rapid pressure increase as a response to the flooding event (4), and finally the post-

1183 MSC heating at a near-constant pressure due to the post-MSC burial (5). On the same

1184 PT diagram, any phase diagram of a possible analogue oil, in that particular case from

1185 the Al Shaheen field offshore Qatar (Lindeloff et al., 2008, Al Ghafri et al., 2014), can

1186 plotted. The bubble point curve of that phase diagram separates the liquid domain from

1187 the mixed liquid and vapour domain, On this combined plot it can be analysed when the

1188 PT path of the model cell crosses the bubble point line, and therefore when and by how

1189 much the corresponding oil mixture is expected to change phase, i.e., starts to build a

1190 gas cap.

1191

1192 Figure 12: Models showing the impact of phase change for different trap geometries.

1193 Three trap scenarios have been modelled, where each represents a trap fill-controlling

1194 parameter: (A) amount of charge, or “charge limited”, (B) seal strength, or “seal limited”,

1195 and (C) trap size, or “trap limited”. Those traps have been filled by a “near bubble point”

1196 hydrocarbon mixture of mainly thermogenic oil with some biogenic gas has been

1197 created to fill the trap, referred to as “reactive accumulations”. The oil has essentially

54
1198 been sufficiently enriched with gas that those accumulations can undergo gas

1199 exsolution during the MSC and its consequences can be modelled. The migration

1200 method used in the simulation is Invasion Percolation (Wilkinson, 1984, Hantschel and

1201 Kauerauf, 2009).

1202

1203 Figure 13: Pockmark and fluid escape evidence from offshore Western Desert (refer to

1204 Figure 1 for location). The pockmark appears as a sub circular depression, 2-3 km wide

1205 and about 500 m deep, surrounded by truncational unconformities on either side. The

1206 chaotic distorted seismic character beneath and surrounding the pockmarks could be

1207 attributed to focussed fluid flow from the deeper pre-Messinian structures during the

1208 water drawdown which disrupted the overburden sediments.

1209

1210

1211 References

1212

1213 Al-Balushi, A., Fraser, A.J., Allen, P.A. & Jackson, C.A.-L. 2013. Quantifying the impact

1214 of the Messinian Salinity Crisis on the Eastern Mediterranean petroleum systems AAPG

1215 European Regional Conference, Exploring the Mediterranean: New Concepts in an

1216 Ancient Seaway, April 8-10, 2013, Barcelona, Spain.

1217 Al-Balushi, A., Fraser, A.J., Jackson, C.A.-L. & Allen, P.A. 2012. The tectono-

1218 stratigraphic evolution of the Eastern Mediterranean and implications for the

1219 development of petroleum systems, offshore Nile Delta and Levant Margin. Petroleum

1220 Geoscience Research Collaboration Showcase, Earls court, London.

55
1221 Al Ghafri, S.Z., Maitland, G.C. & Trusler, J.P.M. 2014. Experimental and modeling study

1222 of the phase behavior of synthetic crude oil + CO2. Fluid Phase Equilibria, 365, 20-40,

1223 doi: http://dx.doi.org/10.1016/j.fluid.2013.12.018.

1224 Ala, M.A. & Moss, B. 1979. Comparative petroleum geology of southeast Turkey and

1225 northeast Syria. Journal of Petroleum Geology, 1, 3-27.

1226 Allen, P.A. & Allen, J.R. 2013. Basin Analysis: Principles and Application to Petroleum

1227 Play Assessment. Wiley.

1228 Almogi‐Labin, A., Bein, A. & Sass, E. 1993. Late Cretaceous upwelling system along

1229 the southern Tethys margin (Israel): interrelationship between productivity, bottom water

1230 environments, and organic matter preservation. Paleoceanography, 8, 671-690.

1231 Alsharhan, A.S. 2003. Petroleum geology and potential hydrocarbon plays in the Gulf of

1232 Suez rift basin, Egypt. AAPG Bulletin, 87, 143-180, doi: 10.1306/062002870143.

1233 Alsharhan, A.S. & Abd El-Gawad, E.A. 2008. Geochemical characterization of Potential

1234 Jurassic/Cretaceous source rocks in the Shushan Basin, northern Western Desert,

1235 Egypt. Journal of Petroleum Geology, 31, 191-212, doi: 10.1111/j.1747-

1236 5457.2008.00416.x.

1237 Andresen, K.J. & Huuse, M. 2011. ‘Bulls-eye’ pockmarks and polygonal faulting in the

1238 Lower Congo Basin: Relative timing and implications for fluid expulsion during shallow

1239 burial. Marine Geology, 279, 111-127, doi:

1240 http://dx.doi.org/10.1016/j.margeo.2010.10.016.

1241 Andresen, K.J., Huuse, M. & Clausen, O. 2008. Morphology and distribution of

1242 Oligocene and Miocene pockmarks in the Danish North Sea–implications for bottom

1243 current activity and fluid migration. Basin Research, 20, 445-466.

56
1244 Bartol, J. & Govers, R. 2009. Flexure due to the Messinian-Pontian sea level drop in the

1245 Black Sea. Geochemistry, Geophysics, Geosystems, 10, Q10013, doi:

1246 10.1029/2009gc002672.

1247 Behar, F., Vandenbroucke, M., Tang, Y., Marquis, F. & Espitalie, J. 1997. Thermal

1248 cracking of kerogen in open and closed systems: determination of kinetic parameters

1249 and stoichiometric coefficients for oil and gas generation. Organic geochemistry, 26,

1250 321-339, doi: http://dx.doi.org/10.1016/S0146-6380(97)00014-4.

1251 Bein, A. & Gvirtzman, G. 1977. A Mesozoic fossil edge of the Arabian plate along the

1252 Levant coastline and its bearing on the evolution of the Eastern Mediterranean.

1253 Structural History of the Mediterranean Basins, B. Diju-Duval and L. Montadert, eds, 95-

1254 110.

1255 Ben-Avraham, Z. 1989. Multiple opening and closing of the eastern Mediterranean and

1256 south China basins. Tectonics, 8, 351-362.

1257 Ben-Avraham, Z., Ginzburg, A., Makris, J. & Eppelbaum, L. 2002. Crustal structure of

1258 the Levant Basin, eastern Mediterranean. Tectonophysics, 346, 23-43.

1259 Ben-Avraham, Z., Woodside, J., Lodolo, E., Gardosh, M., Grasso, M., Camerlenghi, A.

1260 & Vai, G.B. 2006. Eastern Mediterranean basin systems. Geological Society, London,

1261 Memoirs, 32, 263-276, doi: 10.1144/gsl.mem.2006.032.01.15.

1262 Ben-Gai, Y., Ben-Avraham, Z., Buchbinder, B. & Kendall, C.G.S.C. 2005. Post-

1263 Messinian evolution of the Southeastern Levant Basin based on two-dimensional

1264 stratigraphic simulation. Marine Geology, 221, 359-379, doi:

1265 http://dx.doi.org/10.1016/j.margeo.2005.03.003.

57
1266 Bertoni, C., Cartwright, J. & Hermanrud, C. 2013. Evidence for large-scale methane

1267 venting due to rapid drawdown of sea level during the Messinian Salinity Crisis.

1268 Geology, doi: 10.1130/g33987.1.

1269 Bertoni, C. & Cartwright, J.A. 2005. 3D seismic analysis of circular evaporite dissolution

1270 structures, Eastern Mediterranean. Journal of the Geological Society, 162, 909-926, doi:

1271 10.1144/0016-764904-126.

1272 Bizarro. 1998. Subcircular features and autotracking artefacts in 3D seismic

1273 interpretation; a case study from the central North Sea. Petroleum Geoscience, 4, 173-

1274 179.

1275 Bou Daher, S., Nader, F., Strauss, H. & Littke, R. 2014. Depositional environment and

1276 source rock characterisation of organic‐matter rich Upper Santonian–Upper Campanian

1277 carbonates, Northern Lebanon. Journal of Petroleum Geology, 37, 5-24.

1278 Carrigan, W., Cole, G., Colling, E. & Jones, P. 1995. Geochemistry of the Upper

1279 Jurassic Tuwaiq Mountain and Hanifa Formation Petroleum Source Rocks of Eastern

1280 Saudi Arabia. Petroleum source rocks. Springer, 67-87.

1281 Dandekar, A.Y. 2013. Petroleum Reservoir Rock and Fluid Properties. CRC Press.

1282 Dewey, J.F., Pitman, W.C., Ryan, W.B.F. & Bonnin, J. 1973. Plate Tectonics and the

1283 Evolution of the Alpine System. Geological Society of America Bulletin, 84, 3137-3180,

1284 doi: 10.1130/0016-7606(1973)84<3137:ptateo>2.0.co;2.

1285 di Primio, R. & Horsfield, B. 2006. From petroleum-type organofacies to hydrocarbon

1286 phase prediction. AAPG Bulletin, 90, 1031-1058, doi: 10.1306/02140605129.

1287 Dimitrov, L.I. 2002. Mud volcanoes—the most important pathway for degassing deeply

1288 buried sediments. Earth-Science Reviews, 59, 49-76.

58
1289 Dubille, M., Thomas D., 2012. Petroleum System Assessment of the Offshore Lebanon

1290 Thorugh 3D Basin Modeling. Lebanese International Petroleum Exploration Forum

1291 (LIPE), Beirut, Lebanon.

1292 Dolson, J.C., Shann, M.V., Matbouly, S., Harwood, C., Rashed, R. & Hammouda, H.

1293 2001. AAPG Memoir 74, Chapter 23: The Petroleum Potential of Egypt.

1294 Druckman, Y., Conway, B., Eshet, Y., Gill, D., Grossowicz, L., Lipson, S., Moshkovitz,

1295 S., Rosenfled, A. & Siman-Tov, R. 1994. The stratigraphy of the Yam Yafo 1 borehole.

1296 Report GSI/28/94. 25 pp. Israel Geological Survey. Jerusalem, Israel.

1297 Dupré, S., Woodside, J., Klaucke, I., Mascle, J. & Foucher, J.-P. 2010. Widespread

1298 active seepage activity on the Nile Deep Sea Fan (offshore Egypt) revealed by high-

1299 definition geophysical imagery. Marine Geology, 275, 1-19.

1300 Durham, L. 2013. Levant Basin Brings Potential to New Areas. EXPLORER May 2013.

1301 Emeis, Robertson, A.H.F. & Richter, C., et al.,. 1996a. Proc. ODP, Init.Repts., 160:

1302 College Station, TX (Ocean Drilling Program).

1303 Emeis, K., Robertson, A.H.F. & (Eds.), C.e.a.R. 1996b. Proceedings of the Ocean

1304 Drilling Program, Initial Reports 160. College Station, TX (Ocean Drilling Program), 972

1305 pp.

1306 Eruteya, O. E., Waldmann, N., Schalev, D., Makovsky, Y., Ben-Avraham, Z. B. 2015.

1307 Intra- to post-Messinian deep-water gas piping in the Levant Basin,

1308 SE Mediterranean. Marine and Petroleum Geology.

1309 Flexer, A., Rosenfeld, A., Lipson-Benitah, S. & Honigstein, A. 1986. Relative sea level

1310 changes during the Cretaceous in Israel. AAPG Bulletin, 70, 1685-1699.

59
1311 Fraser, A., Guderjahn, C.G., Allen, H. & Al-Balushi, A. 2011. The Messinian Salinity

1312 Crisis: Impact on Prospectivity in the E. Mediterranean. AAPG International Conference

1313 and Exhibition, Milan, Italy, 23-26 October 2011.

1314 Gardosh, M. & Druckman, Y. 2006. Seismic stratigraphy, structure and tectonic

1315 evolution of the Levantine Basin, offshore Israel. Tectonic development of the Eastern

1316 Mediterranean region, 201–227.

1317 Gardosh, M., Weimer, P. & Flexer, A. 2011. The sequence stratigraphy of Mesozoic

1318 successions in the Levant margin, southwestern Israel: A model for the evolution of

1319 southern Tethys margins. AAPG Bulletin, 95, 1763-1793, doi: 10.1306/02081109135.

1320 Gardosh, M.A., Garfunkel, Z., Druckman, Y. & Buchbinder, B. 2010. Tethyan rifting in

1321 the Levant Region and its role in Early Mesozoic crustal evolution. Geological Society,

1322 London, Special Publications, 341, 9-36, doi: 10.1144/sp341.2.

1323 Garfunkel, Z. 1998. Constrains on the origin and history of the Eastern Mediterranean

1324 basin. Tectonophysics, 298, 5-35.

1325 Garfunkel, Z. & Derin, B. 1984. Permian-early Mesozoic tectonism and continental

1326 margin formation in Israel and its implications for the history of the Eastern

1327 Mediterranean. Geological Society, London, Special Publications, 17, 187-201, doi:

1328 10.1144/gsl.sp.1984.017.01.12.

1329 Gemmer, L., Huuse, M., Clausen, O.R. & Nielsen, S.B. 2002. Mid‐Palaeocene

1330 palaeogeography of the eastern North Sea basin: integrating geological evidence and

1331 3D geodynamic modelling. Basin Research, 14, 329-346.

60
1332 Gill, D., Conway, B., Eshet, Y., Lipson, S., Perelis-Grossowicz, L., Rosenfeld, A. &

1333 Siman-Tov, R. 1995. The stratigraphy of the Yam West 1 borehole. Report GSI/13/95.

1334 Geol. Surv. Isr., Jerusalem.

1335 Goldberg, M. & Friedman, G.M. 1974. Paleoenvironments and paleogeographic

1336 evolution of the Jurassic system in southern Israel. State of Israel, Ministry of

1337 Commerce and Industry, Geological Survey.

1338 Gumati, Y. & Schamel, S. 1988. Thermal maturation history of the Sirte Basin, Libya.

1339 Journal of Petroleum Geology, 11, 205-218.

1340 Gvirtzman, G. & Buchbinder, B. 1978. The late Tertiary of the coastal plain and

1341 continental shelf of Israel and its bearing on the history of the Eastern Mediterranean.

1342 In. Rep. Deep Sea Drill. Proj., vol. 42 (2). U.S. Govt., Printing Office, Washington, DC,

1343 pp. 1195–1222.

1344 Hallett, D. 2002. Petroleum geology of Libya [electronic resource]. Elsevier Science &

1345 Technology Books.

1346 Hawie, N., Gorini, C., Deschamps, R., Nader, F.H., Montadert, L., Granjeon, D. &

1347 Baudin, F. 2013. Tectono-stratigraphic evolution of the northern Levant Basin (offshore

1348 Lebanon). Marine and Petroleum Geology, 48, 392-410, doi:

1349 http://dx.doi.org/10.1016/j.marpetgeo.2013.08.004.

1350 Henriksen, E., Bjørnseth, H.M., Hals, T.K., Heide, T., Kiryukhina, T., Kløvjan, O.S.,

1351 Larssen, G.B., Ryseth, A.E., Rønning, K., Sollid, K. & Stoupakova, A. 2011. Chapter 17

1352 Uplift and erosion of the greater Barents Sea: impact on prospectivity and petroleum

1353 systems. Geological Society, London, Memoirs, 35, 271-281, doi: 10.1144/m35.17.

61
1354 Hilgen, F., Kuiper, K., Krijgsman, W., Snel, E. & van der Laan, E. 2007. Astronomical

1355 tuning as the basis for high resolution chronostratigraphy: the intricate history of the

1356 Messinian Salinity Crisis. Stratigraphy, 4, 231–238.

1357 Hilgen, F.J. & Langereis, C.G. 1993. A critical re-evaluation of the Miocene/Pliocene

1358 boundary as defined in the Mediterranean. Earth and Planetary Science Letters, 118,

1359 167-179.

1360 Hirsch, F., Flexer, A., Rosenfeld, A. & Yellin Dror, A. 1995. Palinspastic and crustal

1361 setting of the eastern Mediterranean. Journal of Petroleum Geology, 18, 149-170.

1362 Hodgson, N. 2012. The Miocene hydrocarbon play in Southern Lebanon. first break

1363 volume 30, December 2012.

1364 Hsü, K.J. 1972. Origin of saline giants: A critical review after the discovery of the

1365 Mediterranean Evaporite. Earth-Science Reviews, 8, 371-396.

1366 Hsü, K.J., Montadert, L., Bernoulli, D., Cita, M.B., Erickson, A., Garrison, R.E., Kidd,

1367 R.B., Mèlierés, F., Müller, C. & Wright, R. 1977. History of the Mediterranean salinity

1368 crisis. Nature, 267, 399-403.

1369 Iadanza, A., Sampalmieri, G., Cipollari, P., Mola, M. & Cosentino, D. 2013. The

1370 “Brecciated Limestones” of Maiella, Italy: Rheological implications of hydrocarbon-

1371 charged fluid migration in the Messinian Mediterranean Basin. Palaeogeography,

1372 Palaeoclimatology, Palaeoecology, 390, 130-147, doi:

1373 http://dx.doi.org/10.1016/j.palaeo.2013.05.033.

1374 Jassim, S.Z. & Goff, J.C. 2006. Geology of Iraq. Geological Society of London.

1375 Jennifer Villinski. 2013. Unusual but Effective Petroleum Systems – Disseminated

1376 Terrestrial Organic Matter of the Upper Oligocene as the Primary Source Rock,

62
1377 Offshore Nile Delta, Egypt. AAPG European Regional Conference, Barcelona, Spain, 8-

1378 10 April 2013.

1379 John Healy, S., Jack Sanford, S., Kerby Dufrene, S. & Josh Fink, D.R. 2013. Design,

1380 Installation, and Initial Performance of Ultra-High-Rate Gas Deepwater Completions-

1381 Tamar Field, Offshore Israel.

1382 Judd, A.G. & Hovland, M. 2007. Seabed fluid flow: the impact of geology, biology and

1383 the marine environment. Cambridge University Press.

1384 King, L.H. & MacLean, B. 1970. Pockmarks on the Scotian shelf. Geological Society of

1385 America Bulletin, 81, 3141-3148.

1386 Klett, T.R. 2001. Total Petroleum Systems of the Pelagian Province, Tunisia, Libya,

1387 Italy, and Malta—The Bou Dabbous– Tertiary and Jurassic-Cretaceous Composite. U.S.

1388 Geological Survey Bulletin 2202-D.

1389 Krijgsman, W., Blanc-Valleron, M.M., Flecker, R., Hilgen, F.J., Kouwenhoven, T.J.,

1390 Merle, D., Orszag-Sperber, F. & Rouchy, J.M. 2002. The onset of the Messinian salinity

1391 crisis in the Eastern Mediterranean (Pissouri Basin, Cyprus). Earth and Planetary

1392 Science Letters, 194, 299-310.

1393 Lazar, M., Schattner, U. & Reshef, M. 2012. The great escape: An intra-Messinian gas

1394 system in the eastern Mediterranean. Geophysical research letters, 39, L20309, doi:

1395 10.1029/2012gl053484.

1396 Lie, Ø., Skiple, C. & Lowrey, C. 2011. New Insights into the Levantine Basin.

1397 GeoExPro. GeoPublishing Ltd, London, United Kingdom, 24-27.

1398 Lindeloff N , S., Mogensen K , S., van Lingen P, S., H. Do. S., S., FrankS. , S., Maersk

1399 Oil, & Noman R. , S., Qatar Petroleum. 2008. Fluid-Phase Behaviour for a Miscible-

63
1400 Gas-Injection EOR Project in a Giant Offshore Oil Field With Large Compositional

1401 Variations. SPE Annual Technical Conference and Exhibition, 21-24 September 2008,

1402 Denver, Colorado, USA.

1403 Loncke, L., Mascle, J. & Parties, F.S. 2004. Mud volcanoes, gas chimneys, pockmarks

1404 and mounds in the Nile deep-sea fan (Eastern Mediterranean): geophysical evidences.

1405 Marine and Petroleum Geology, 21, 669-689.

1406 Longacre, M., Bentham, P., I.Hanbal, J.Cotton & R.Edwards. 2007. New Crustal

1407 Structure of the Eastern MEditerranean Basin: Detailed Integration and Modeling of

1408 Gravity, Magnetic, Seismic Refraction, and Seismic Reflection Data. EGM 2007

1409 International Workshop, Capri, Italy.

1410 Macgregor, D.S. 2012. The development of the nile drainage system: integration of

1411 onshore and offshore evidence. Petroleum Geoscience, 18, 417-431, doi: DOI

1412 10.1144/petgeo2011-074.

1413 Macgregor, D.S. & Moody, R.T. 1998. Mesozoic and Cenozoic petroleum systems of

1414 North Africa. Geological Society, London, Special Publications, 132, 201-216.

1415 Mart, Y., Ryan, W.B.F. & Lunina, O.V. 2005. Review of the tectonics of the Levant Rift

1416 system: the structural significance of oblique continental breakup. Tectonophysics, 395,

1417 209-232, doi: http://dx.doi.org/10.1016/j.tecto.2004.09.007.

1418 McKenzie, D. 1978. Some remarks on the development of sedimentary basins. Earth

1419 and Planetary Science Letters, 40, 25-32.

1420 Meijer, P.T. & Krijgsman, W. 2005. A quantitative analysis of the desiccation and re-

1421 filling of the Mediterranean during the Messinian Salinity Crisis. Earth and Planetary

1422 Science Letters, 240, 510-520.

64
1423 Moretti, I., Kerdraon, Y., Rodrigo, G., Huerta, F., Griso, J.J., Sami, M., Said, M. & Ali, H.

1424 2010. South Alamein petroleum system (Western Desert, Egypt). Petroleum

1425 Geoscience, 16, 121-132, doi: 10.1144/1354-079309-004.

1426 Middelburg, J.J., Calvert, S.E., Karlin, R., 1991. Organic-rich transition facies in silled

1427 basins: Response to sea-level change. Geology 19: 679-682.

1428 Nader, F.H. 2011. The petroleum prospectivity of Lebanon: an overview. Journal of

1429 Petroleum Geology, 34, 135-156, doi: 10.1111/j.1747-5457.2011.00498.x.

1430 Needham, D., Hosler, J., Nowak, S., Christensen, C. & Ffrench, J. 2013. The Tamar

1431 Field from Discovery to Production. AAPG European Regional Conference, Barcelona,

1432 Spain, 8-10 April 2013.

1433 Neev, D. 1975. Tectonic evolution of the Middle East and the Levantine basin

1434 (easternmost Mediterranean). Geology, 3, 683-686, doi: 10.1130/0091-

1435 7613(1975)3<683:teotme>2.0.co;2.

1436 Nyland, B., Jensen, L., Skagen, J., Skarpnes, O. & Vorren, T. 1992. Tertiary uplift and

1437 erosion in the Barents Sea: magnitude, timing and consequences. Structural and

1438 tectonic modelling and its application to petroleum geology, 1, 153-162.

1439 Patton, T.L., Moustafa, A.R., Nelson, R.A. & Abdine, S.A. 1994. Tectonic evolution and

1440 structural setting of the Suez Rift. Interior Rift Basins: AAPG Memoir 59, 9.

1441 Robertson, A.H. 1998a. Mesozoic-Tertiary tectonic evolution of the easternmost

1442 Mediterranean area: integration of marine and land evidence. Proceedings of the Ocean

1443 Drilling Program, Scientific Results, Vol. 160; Chapter 54.

65
1444 Robertson, A.H. 1998b. Miocene shallow-water carbonates on the Eratosthenes

1445 Seamount, easternmost Mediterranean Sea. Proceedings of the Ocean Drilling

1446 Program, Scientific Results, Vol. 160; Chapter 33.

1447 Robertson, A.H. & Dixon, J. 1984a. Introduction: aspects of the geological evolution of

1448 the Eastern Mediterranean. Geological Society, London, Special Publications, 17, 1-74.

1449 Robertson, A.H.F. & Dixon, J.E. 1984b. Introduction: aspects of the geological evolution

1450 of the Eastern Mediterranean. Geological Society, London, Special Publications, 17, 1-

1451 74, doi: 10.1144/gsl.sp.1984.017.01.02.

1452 Roveri, M., Bertini, A., Cosentino, D., Di Stefano, A., Gennari, R., Gliozzi, E., Grossi, F.,

1453 Iaccarino, S.M., Lugli, S. & Manzi, V. 2008. A high-resolution stratigraphic framework for

1454 the latest Messinian events in the Mediterranean area. Stratigraphy, 5, 327–345.

1455 Roveri, M., Bertini, A. & Cosentino, D.e.a. 2008b. A high-resolution stratigraphic

1456 framework for the latest Messinian events in the Mediterranean area. Stratigraphy, 5(3–

1457 4), 323–342.

1458 Ryan, W.B. 1976. Quantitative evaluation of the depth of the western Mediterranean

1459 before, during and after the Late Miocene salinity crisis. Sedimentology, 23, 791-813,

1460 doi: 10.1111/j.1365-3091.1976.tb00109.x.

1461 Ryan, W.B.F. 1978. Messinian badlands on the southeastern margin of the

1462 Mediterranean Sea. Marine Geology, 27, 349-363, doi: http://dx.doi.org/10.1016/0025-

1463 3227(78)90039-7.

1464 Ryan, W.B.F. 2009. Decoding the Mediterranean salinity crisis. Sedimentology, 56, 95-

1465 136, doi: 10.1111/j.1365-3091.2008.01031.x.

66
1466 Ryan, W.B.F. & Cita, M.B. 1978. The nature and distribution of Messinian erosional

1467 surfaces — Indicators of a several-kilometer-deep Mediterranean in the Miocene.

1468 Marine Geology, 27, 193-230, doi: http://dx.doi.org/10.1016/0025-3227(78)90032-4.

1469 Schroot, B.M., Klaver, G.T. & Schüttenhelm, R.T. 2005. Surface and subsurface

1470 expressions of gas seepage to the seabed—examples from the Southern North Sea.

1471 Marine and Petroleum Geology, 22, 499-515.

1472 Scotchman, I.C., Gilchrist, G., Kusznir, N.J., Roberts, A.M. & Fletcher, R. 2010. The

1473 breakup of the South Atlantic Ocean: formation of failed spreading axes and blocks of

1474 thinned continental crust in the Santos Basin, Brazil and its consequences for petroleum

1475 system development. Geological Society, London, Petroleum Geology Conference

1476 series, 7, 855-866, doi: 10.1144/0070855.

1477 Segev, A., Rybakov, M., Lyakhovsky, V., Hofstetter, A., Tibor, G., Goldshmidt, V. & Ben

1478 Avraham, Z. 2006. The structure, isostasy and gravity field of the Levant continental

1479 margin and the southeast Mediterranean area. Tectonophysics, 425, 137-157.

1480 Skagen, J.I. 1993. Effects on hydrocarbon potential caused by Tertiary uplift and

1481 erosion in the Barents Sea. In: Vorren, T.O., Bergsager, E., Dahl-Stamnes, Ø.A.,

1482 Holter, E., Johansen, B., Lie, E., Lund, T.B. (Eds.), Arctic Geology and Petroleum

1483 Potential. Norwegian Petroleum Society, Special Publication 2, pp. 711e719.

1484 Skiple, C., Anderson, E. & Fürstenau, J. 2012. Seismic interpretation and attribute

1485 analysis of the Herodotus and the Levantine Basin, offshore Cyprus and Lebanon.

1486 Petroleum Geoscience, 18, 433-442, doi: 10.1144/petgeo2011-072.

1487 Stampfli, G., Marcoux, J. & Baud, A. 1991. Tethyan margins in space and time.

1488 Palaeogeography, Palaeoclimatology, Palaeoecology, 87, 373-409.

67
1489 Steckler, M.S., Feinstein, S., Kohn, B.P., Lavier, L.L. & Eyal, M. 1998. Pattern of mantle

1490 thinning from subsidence and heat flow measurements in the Gulf of Suez: Evidence for

1491 the rotation of Sinai and along‐strike flow from the Red Sea. Tectonics, 17, 903-920.

1492 Steckler, M.S., Lofi, J., Mountain, G., Ryan, W.B.F., Gorini, C. & Berne´, S. 2003.

1493 Reconstruction of the gulf of Lion margin during the Messinian salinity crisis. EGS-AGU-

1494 EUG Joint Meeting in Nice. Geophysical Research Abstract 5. 07319.

1495 Steinberg, J., Gvirtzman, Z., Folkman, Y. & Garfunkel, Z. 2011. Origin and nature of the

1496 rapid late Tertiary filling of the Levant Basin. Geology, 39, 355-358, doi:

1497 10.1130/g31615.1.Sweeney, J. J., and A. K. Burnham, 1990, Evaluation of a simple

1498 model of vitrinite reflectance based on chemical kinetics: AAPG Bulletin, v. 74, no. 10,

1499 p. 1559–1570.

1500 Ungerer, P. 1990. State of the art of research in kinetic modelling of oil formation and

1501 expulsion. Organic geochemistry, 16, 1-25, doi: http://dx.doi.org/10.1016/0146-

1502 6380(90)90022-R.

1503 Urgeles, R., Camerlenghi, A., Garcia-Castellanos, D., De Mol, B., Garcés, M., Vergés,

1504 J., Haslam, I. & Hardman, M. 2011. New constraints on the Messinian sealevel

1505 drawdown from 3D seismic data of the Ebro Margin, western Mediterranean. Basin

1506 Research, 23, 123-145, doi: 10.1111/j.1365-2117.2010.00477.x.

1507 Woodside, J. 1977. Tectonic elements and crust of the eastern Mediterranean Sea.

1508 Marine Geophysical Researches, 3, 317-354.

1509 Wygrala, B.P. 1989. Integrated Study of an Oil Field in the Southern Po Basin, Northern

1510 Italy [PhD]. University of Cologne.

1511

68
1512

69
Figure 1 Click here to download Figure Fig_1.pdf
Figure 2 Click here to download Figure Fig_2.pdf
Figure 3 Click here to download Figure Fig_3.pdf
Figure 4a Click here to download Figure Fig_4a.pdf
Figure 4b Click here to download Figure Fig_4b.pdf
Figure 4c Click here to download Figure Fig_4c.pdf
Figure 5 Geologichere
Click Time (Ma)
to download Figure Fig_5_new.pdf
6.0 5.9 5.8 5.7 5.6 5.5 5.4 5.3 5.2
1 Pre-Messinian
0

1 -20 Mpa
2900 psi Sea water
2 Sea level fall
13°C
2 13°C 55°F
55°F
40°C
104°F Messinian
3 salt
3
Salt deposition

3 – Salt deposition
4
1 – Pre-Messinian

2 – Sea level fall

4 – Re-flooding
Pre-
Messinian
4 Re-flooding 5
Depth
(km)

Geologic Time (Ma)


Figure 6 Click here to download Figure Fig_6_new.pdf

94°C
Figure 7 Click here to download Figure Fig_7_new.pdf
Geologic Time (Ma)
(a)
6.0 5.9 5.8 5.7 5.6 5.5 5.4 5.3 5.2

Depth [km]
1 -20 Mpa
2900 psi Sea water

13°C
2 8°C 55°F
46°F
40°C
104°F Messinian
3 salt

3 – Salt deposition
4
1 – Pre-Messinian

2 – Sea level fall

4 – Re-flooding
Pre-
Messinian
5

(b)
60
Porosity

Temperature [°C]
Pressure [MPa]

60

Porosity [%]
40
40
Pressure
20 20
Temperature
0 0

(c)

200
Top Basement
Pressure [MPa]

150 (7000m)
Top Mesozoic
100 (3650m)
Oligocene/Miocene
50 (Tamar; 1800m)
Top Miocene
(close to surface)
0

(d)
250
Temperature [°C]

Top Basement
200 (7000m)
150 Top Mesozoic
(3650m)
100 Oligocene/Miocene
(Tamar; 1800m)
50 Top Miocene
(close to surface)
0
Figure 8 Click here to download Figure Fig_8.pdf
Figure 9a,b,c Click here to download Figure Fig_9a_c.pdf
Figure 9d Click here to download Figure Fig_9d_new.pdf

(d)
Figure 10 Click here to download Figure Fig_10_new.pdf

(a)

Messinian Salt
Depth [km]

0 Hydrocarbon Saturation [%] 100

3
(b)

4
Depth [km]

Aphrodite Leviathan Tamar


Figure 11 Click here to download Figure Fig_11_new.pdf

(a) 100
Pressure [MPa]

5 - Burial until d
present day Liqui
Phase envelope of oil accumulation
4 - Re-flooding Example from Al Shaheen Field
1 – Sea-level (offshore Qatar)
drop
3 – Progressive por Lindeloff et al., 2008
drowning Liquid & va Al Ghafri et al., 2014
2 - Salt deposition
0
0 Temperature [°C] 400

(b) 100
Top Jurassic (4500m) high

Top Cretaceous (3600m)

accumulation being
affected by MSC
Pressure [MPa]

Probability of
Oligocene (2000m)

Top Miocene low


d
(close to surface) Liqui ???

Phase envelope of oil accumulation


Example from Al Shaheen Field
???
(offshore Qatar)

por Lindeloff et al., 2008


Liquid & va Al Ghafri et al., 2014

0
0 Temperature [°C] 400
Figure 12 Click here to download Figure Fig_12_new.pdf

Con$nuous oil charge from Mesozoic

Con$nuous oil charge from Mesozoic

Con$nuous oil charge from Mesozoic

Con$nuous oil charge from Mesozoic


Figure 13 Click here to download Figure Fig_13.pdf

You might also like