Éditeur(s) : Elsevier Science Bv
Résumé : Fluid escape structures on the Nile Deep Sea Fan were investigated during the MEDIFLUX MIMES expedition in 2004. Mud volcanoes, pockmarks and authigenic carbonate structures were surveyed for the first time with a high-resolution deep-towed 75 kHz sidescan sonar and a 2-8 kHz Chirp sediment echosounder. In combination with existing multibeam bathymetry and detailed seafloor in situ geological observations, these new data allowed detailed seep analyses. About 60 gas flares were detected acoustically in the water column from the sidescan sonar raw data at water depths from 770 to 1700 m. These gas flares coincide at the seabed with 1) the centres of the mud volcanoes where mud is also extruded, 2) the borders of the mud volcanoes where the emitted gases contribute to the precipitation of authigenic carbonates, and 3) to the edges of broad sheets of authigenic carbonates. Subsurface sediments are commonly disturbed by ascending fluids throughout the delta, with an abundance of seep-related carbonate structures on the seafloor. The feeder channels below mud volcanoes, similar to the gas conduits below the widespread carbonate crust structures and pockmarks, are relatively narrow and, for the vast majority of them, do not exceed a few metres in diameter. The seeps on the Nile Deep Sea Fan clearly follow lineations on the seafloor that we can relate to faults. (C) 2010 Elsevier B.V. All rights reserved.
Marine Geology (0025-3227) (Elsevier Science Bv), 2010-09 , Vol. 275 , N. 1-4 , P. 1-19

Droits : 2010 Elsevier B.V. All rights reserved.
http://archimer.ifremer.fr/doc/00013/12457/9766.pdf
DOI:10.1016/j.margeo.2010.04.003
http://archimer.ifremer.fr/doc/00013/12457/

Éditeur(s) : HAL CCSD
Résumé : Sand intrusions (or injectites) are most often the product of post-depositional remobilization of sand leading to its injection into the surrounding rocks. While injectites were recognized for the first time nearly 200 years ago, their emplacement process has been studied for a couple of decades only, since the concepts of deep sea depositional environments have allowed us to better understand their emplacement processes. However, these processes are still relatively poorly understood. Our approach is based on the study of injectites in the Lower Congo Basin from seismic and well data, which we compare to a fossil system in the SE France basin. We have shown that: (1) In buried turbidite channel systems draping deposits on the channel flanks and terraces of channels have the same geophysical signature as 'wing-like' injectites. Finally, the only criterion for identifying seismic injectites is the presence of bedding-discordant seismic reflections, and in the best case the associated uplift of the overlying seismic reflectors. (2) Seismic-scale conical and saucer-shaped sand injectites have been identified in the Lower Congo Basin. The remobilization is likely due to overpressuring induced by the buoyancy effect of hydrocarbons trapped in the margins of a lobe buried underneath 160 m of sediment, followed by the sudden injection of fluidized sand associated with fault reactivation of faults (with a possible role of nearby salt diapirs). (3) A network of injectites (dykes, sills/wings and laccoliths) was formed in the Vocontian basin during the late Albian and/or early Cenomanian, from a lower-middle Albian turbidite channel. The emplacement is probably due to the early compartmentalization of the channel during its burial and the increase of the sedimentation rate generating overpressure; and the subsequent large influx of deep fluids triggering injection. The injection of sand was polyphased: a first episode formed the sills and another emplaced the dykes. Sills/wings and dykes propagated about 2 km laterally away from the parent sand body and about 200 m up to the surface, revealing a much more extended lateral than vertical reach, contrary to the classically accepted idea from the interpretation of seismic data. (4) The emplacement of this large network of injectites was governed by hydrofracturing. Therefore, its morphology is dependent on the host rock heterogeneity (isotropy, fractures), the paleo-stress orientation (ó3 = NW-SE) and the burial depth of the source (300-600 m) at the time of injection. The study of this fossil network allows us to define the relationship between morphology of the injected network and stress state at the time of injection. This relationship can be extrapolated to constrain the morphology of subsurface networks beyond seismicvisibility. (5) Sands injected into low permeability lithologies bear evidence to a major event of fluid escape in the studied basin, but also channeled fluids long after their formation. In this way, injectites both attest to specific episodes of fluid migration in sedimentary basins and contribute to long-lived re-routing of migrating fluids once emplaced. The injection of sand is associated with the sudden escape of fluids, probably resulting from a significant tectonic and/or sedimentary event; in addition, the architecture of injectite networks is governed by the local paleo-stress and heterogeneity in the host rock. Consequently, characterizing injectite networks is an important step in understanding the plumbing systems of continental margins, i.e. the post-depositional evolution of sedimentary basins.
Les intrusions sableuses (ou injectites) sont le plus souvent le produit de la remobilisation post-dépositionnelle des sédiments et de l'injection du sable dans les roches environnantes. Bien que reconnues pour la première fois il y a près de 200 ans, elles ne sont réellement étudiées que depuis quelques dizaines d'années, depuis que les concepts sur les environnements de dépôt dans les domaines marins profonds nous permettent de mieux comprendre les processus de mise en place. Cependant, ces processus restent encore aujourd'hui relativement mal compris. Notre approche repose sur l'étude d'injectites dans le bassin du Bas-Congo a partir de données de sismique et de puits que nous comparerons a un système fossile dans le bassin du SE de la France. Nous avons montre que : (1) Dans des systèmes de chenaux turbiditiques enfouis, les dépôts de drapage sur les marges et terrasses de chenaux présentent la même signature géophysique que les injectites de type "wing". Finalement, le seul critère sismique d'identification des injectites est la présence de réflexions sismiques sécantes vis-a-vis de la stratigraphie associée dans le meilleur des cas au soulèvement des réflecteurs sismiques sus-jacents. (2) Des injectites d'échelle sismique en forme de cône et d'assiette ont été identifiées dans le bassin du Bas-Congo. La remobilisation résulte probablement des pressions anormales induites par l'effet de flottabilité des hydrocarbures piégés dans les marges d'un lobe enfoui sous 160 m de sédiment, puis de l'injection soudaine du sable fluidise associée a la réactivation de failles (possible rôle des diapirs de sel a proximité). (3) Un réseau d'injectites (dykes, sills/wings et laccolites) s'est formé dans le bassin Vocontien entre la fin de l'Albien supérieur et/ou le début du Cénomanien, depuis un chenal turbiditique de l'Albien inférieur-moyen. La mise en place résulte probablement de la compartimentalisation précoce du chenal au cours de son enfouissement et de l'augmentation du taux de sédimentation générant la surpression et de l'apport ulterieur d'importante quantité de fluides profonds déclenchant l'injection. L'injection du sable a été polyphasée : une première injection a formé des sills et une suivante des dykes. Les sills/wings et les dykes se sont propagés latéralement au chenal source sur environ 2 km et vers la surface sur environ 200 m, mettant en évidence une forte remobilisation latérale plutôt que verticale, contrairement a l'idée classiquement admise a partir de l'interprétation des données sismiques. (4) La formation de ce large réseau d'injectites a été gouvernée par des mécanismes d'hydrofracturation. Par conséquent, sa morphologie a été dépendante des hétérogenéités de la roche hôte (milieu isotrope, fracture), des directions de paléo-contraintes (ƒÐ3 = NWSE) et de la profondeur d'enfouissement de la source (300-600 m) au moment de l'injection. L'étude de ce réseau fossile permet de définir les relations entre morphologie du réseau injecté et état de contraintes au moment de l'injection. Cette relation peut être extrapolée de façon à contraindre la morphologie des réseaux de subsurface au-delà de la visibilité sismique. (5) Les sables injectés dans des lithologies de faible perméabilité témoignent d'un épisode d'échappement de fluide important dans les bassins étudiés mais ont aussi guide les fluides longtemps après leur formation. Les injectites contribuent ainsi a l'initiation épisodique et la pérennisation de migrations de fluides dans les bassins sedimentaires. Le processus d'injection est associé a l'échappement brutal de fluides, résultant vraisemblablement d'un évènement tectonique et/ou sédimentaire important, et l'architecture des réseaux d'injectites est gouverné par les paléo-contraintes locales et les hétérogénéités de la roche hôte. Par conséquent, la caractérisation des réseaux d'injectites est une étape importante dans la compréhension de la plomberie des marges, c'est-a-dire l'évolution post-dépôt des bassins sédimentaires.
https://tel.archives-ouvertes.fr/tel-01011486

tel-01011486
https://tel.archives-ouvertes.fr/tel-01011486
https://tel.archives-ouvertes.fr/tel-01011486/document
https://tel.archives-ouvertes.fr/tel-01011486/file/These_Monnier_2013_banalisee.pdf

Éditeur(s) : Elsevier
Résumé : We report on a multidisciplinary study of cold seeps explored in the Central Nile deep-sea fan of the Egyptian margin. Our approach combines in situ seafloor observation, geophysics, sedimentological data, measurement of bottom-water methane anomalies, pore-water and sediment geochemistry, and Th-230/U dating of authigenic carbonates. Two areas were investigated, which correspond to different sedimentary provinces. The lower slope, at similar to 2100 m water depth, indicates deformation of sediments by gravitational processes, exhibiting slope-parallel elongated ridges and seafloor depressions. In contrast, the middle slope, at similar to 1650 m water depth, exhibits a series of debris-flow deposits not remobilized by post-depositional gravity processes. Significant differences exist between fluid-escape structures from the two studied areas. At the lower slope, methane anomalies were detected in bottom-waters above the depressions, whereas the adjacent ridges show a frequent coverage of fractured carbonate pavements associated with chemosynthetic vent communities. Carbonate U/Th age dates (similar to 8 kyr BP), pore-water sulphate and solid phase sediment data suggest that seepage activity at those carbonate ridges has decreased over the recent past. In contrast, large (similar to 1 km(2)) carbonate-paved areas were discovered in the middle slope, with U/Th isotope evidence for ongoing carbonate precipitation during the Late Holocene (since similar to 5 kyr BP at least). Our results suggest that fluid venting is closely related to sediment deformation in the Central Nile margin. It is proposed that slope instability leads to focused fluid flow in the lower slope and exposure of 'fossil' carbonate ridges, whereas pervasive diffuse flow prevails at the unfailed middle slope.
Marine Geology (0025-3227) (Elsevier), 2009-06 , Vol. 261 , N. 1-4 , P. 92-104

Droits : 2009 Elsevier B.V. All rights reserved.
http://archimer.ifremer.fr/doc/2009/publication-6586.pdf
DOI:10.1016/j.margeo.2008.10.008
http://archimer.ifremer.fr/doc/00000/6586/

Éditeur(s) : HAL CCSD Springer-Verlag
Résumé : International audience
With hull-mounted multibeam echosounder data, wereport for the first time along the active Ecuadorian margin,acoustic signatures of water column fluid emissions and seep-relatedstructures on the seafloor. In total 17 flare-shaped acousticanomalies were detected from the upper slope (1250 m) to the shelfbreak (140 m). Nearly half of the flare-shaped acoustic anomaliesrise 200–500 m above the seafloor. The base of the flares is generallyassociated with high-reflectivity backscatter patchescontrasting with the neighboring seafloor. We interpret these flaresas caused by fluid escape in the water column, most likely gases.High-resolution seismic profiles show that most flares occur closeto the surface expression of active faults, deformed areas, slopeinstabilities or diapiric structures. In two areas tectonic deformationdisrupts a Bottom Simulating Reflector (BSR), suggesting thatburied frozen gas hydrates are destabilized, thus supplying free gasemissions and related flares. This discovery is important as it opensthe way to determine the nature and origin of the emitted fluids andtheir potential link with the hydrocarbon system of the forearcbasins along the Ecuadorian margin.
ISSN: 0033-4533

insu-01267238
https://hal-insu.archives-ouvertes.fr/insu-01267238
DOI : 10.1007/s00024-015-1230-7

Éditeur(s) : Geological Soc Amer, Inc
Résumé : Pockmarks are seafloor depressions commonly associated with fluid escape from the seabed and are believed to contribute noticeably to the transfer of methane into the ocean and ultimately into the atmosphere. They occur in many different areas and geological contexts, and vary greatly in size and shape. Nevertheless, the mechanisms of pockmark growth are still largely unclear. Still, seabed methane emissions contribute to the global carbon budget, and understanding such processes is critical to constrain future quantifications of seabed methane release at local and global scales. The giant Regab pockmark (9 degrees 42.6' E, 5 degrees 47.8' S), located at 3160 m water depth near the Congo deep-sea channel (offshore southwestern Africa), was investigated with state-of-the-art mapping devices mounted on IFREMER's (French Research Institute for Exploitation of the Sea) remotely operated vehicle (ROV) Victor 6000. ROV-borne micro-bathymetry and backscatter data of the entire structure, a high-resolution photo-mosaic covering 105,000 m(2) of the most active area, sidescan mapping of gas emissions, and maps of faunal distribution as well as of carbonate crust occurrence are combined to provide an unprecedented detailed view of a giant pockmark. All data sets suggest that the pockmark is composed of two very distinctive zones in terms of seepage intensity. We postulate that these zones are the surface expression of two fluid flow regimes in the subsurface: focused flow through a fractured medium and diffuse flow through a porous medium. We conclude that the growth of giant pockmarks is controlled by self-sealing processes and lateral spreading of rising fluids. In particular, partial redirection of fluids through fractures in the sediments can drive the pockmark growth in preferential directions.
Geology (0091-7613) (Geological Soc Amer, Inc), 2014-01 , Vol. 42 , N. 1 , P. 63-66

Droits : 2013 Geological Society of America
http://archimer.ifremer.fr/doc/00186/29731/28293.pdf
DOI:10.1130/G34801.1
http://archimer.ifremer.fr/doc/00186/29731/

Éditeur(s) : HAL CCSD Elsevier
Résumé : The seismic characteristics of focused fluid-related features such as pockmarks and carbonate build-ups on the Nigerian continental slope have been investigated using complementary seismic data sets (reprocessed 3D exploration data and 2D Very High Resolution hull-mounted and near-bottom seismic data) coupled with the results of previous studies (sedimentological, geotechnical and geochemical analyses).The results show different types of fluid-related features within the hemipelagic phase of long duration (∼1 Myr) Turbidite/Hemipelagic cycles. They are the product of the disturbance of the hosting sediments following two main distinct processes: (1) no or low sediment deposition caused by fluid escape associated with seafloor settlement (benthic fauna, carbonate precipitation, gas hydrate formation) and (2) post-depositional erosion, caused either by fluid release (dewatering from underlying features such as channels and upward migration from a deep source) or by collapse (destabilisation of gas hydrates). The temporal evolution reveals a decrease in fluid flow during the hemipelagic deposition as witnessed by the decrease in methanogenic activity and by gas hydrate destabilisation. The spatial distribution of fluid-related features seems to be controlled by the type of fault (small fault or structural lineament).
ISSN: 0264-8172

hal-01685513
https://hal.archives-ouvertes.fr/hal-01685513
DOI : 10.1016/j.marpetgeo.2017.10.004

Éditeur(s) : HAL CCSD Elsevier
Résumé : International audience
Large sand intrusions often exhibit conical morphologies analogous to magmatic intrusions such as saucer-shaped or cup-shaped sills. Whereas some physical processes may be similar, we show with scaled experiments that the formation of conical sand intrusions may be favoured by the pore-pressure gradients prevailing in the host rock before sand injection. Our experiments involve injecting air into a permeable and cohesive analogue material to produce hydraulic fractures while controlling the pore pressure field. We control the state of overpressure in the overburden by applying homogeneous basal pore pressure, and then adding a second local pore pressure field by injecting air via a central injector to initiate hydraulic fractures near the injection point. In experiments involving small vertical effective stresses (small overburden, or high pore fluid overpressure), the fracturing pressure (lambda(fract)) is supralithostatic and two dipping fractures are initiated at the injection point forming a conical structure. From theoretical considerations, we predict that high values of lambda(fract), are due to strong cohesion or high pore fluid overpressure distributed in the overburden. Such conditions are favoured by the pore pressure/stress coupling induced by both pore pressure fields. The dips of cones can be accounted for elastic-stress rotation occurring around the source. Contrary to magmatic chamber models, the aqueous fluid overpressure developed in a parent sandbody (and prevailing before the formation of injectites) may diffuse into the surrounding overburden, thus favouring stress rotation and the formation of inclined sheets far from the parent source. For experiments involving higher vertical effective stresses (thick overburden or low pore fluid overpressure), the fracturing pressure is lower than the lithostatic stress, and a single fracture is opened in mode I which then grows vertically. At a critical depth, the fracture separates into two dilatant branches forming a flat cone. We make use of a P.I.V. (Particle Imaging Velocimetiy) technique to analyse plastic deformation, showing that these inclined fractures are opened in mixed modes. Close to the surface, they change into steep shear bands where fluids can infiltrate. The final morphology of the fracture network is very similar to the common tripartite architecture of various injection complexes, indicating that different mechanisms may be involved in the formation of dykes. Feeder dykes under the sill zones may open as tensile fractures, while overlying dykes may be guided by the deformation induced by the growth of sills. These deformation conditions may also favour the formation of fluid escape structures and pockmarks. (C) 2011 Elsevier B.V. All rights reserved.
ISSN: 0012-821X

hal-00745640
https://hal.archives-ouvertes.fr/hal-00745640
DOI : 10.1016/j.epsl.2011.10.029

Éditeur(s) : HAL CCSD Elsevier
Résumé : International audience
The Giant Gjallar Vent (GGV), located in the Vøring Basin off mid-Norway, is one of the largest (~ 5 × 3 km) vent systems in the North Atlantic. The vent represents a reactivated former hydrothermal system that formed at about 56 Ma. It is fed by two pipes of 440 m and 480 m diameter that extend from the Lower Eocene section up to the Base Pleistocene Unconformity (BPU). Previous studies based on 3D seismic data differ in their interpretations of the present activity of the GGV, describing the system as buried and as reactivated in the Upper Pliocene. We present a new interpretation of the GGV's reactivation, using high-resolution 2D seismic and Parasound data. Despite the absence of geochemical and hydroacoustic indications for fluid escape into the water column, the GGV appears to be active because of various seismic anomalies which we interpret to indicate the presence of free gas in the subsurface. The anomalies are confined to the Kai Formation beneath the BPU and the overlying Naust Formation, which are interpreted to act as a seal to upward fluid migration. The seal is breached by focused fluid migration at one location where an up to 100 m wide chimney-like anomaly extends from the BPU up to the seafloor. We propose that further overpressure build-up in response to sediment loading and continued gas ascent beneath the BPU will eventually lead to large-scale seal bypass, starting a new phase of venting at the GGV.
ISSN: 0025-3227

hal-01054142
https://hal.archives-ouvertes.fr/hal-01054142
DOI : 10.1016/j.margeo.2014.03.006

Éditeur(s) : HAL CCSD Elsevier
Résumé : International audience
Type I and Type II eclogite xenoliths from the Roberts Victor kimberlite (South Africa) show marked differences in terms of microstructures, mineralogy, major-and trace-element compositions and oxygen-isotope compositions. The unequilibrated microstructures of Type I eclogites, their typical accessory assemblages (phologopite, diamond, sulphides, fluid inclusions) and the ubiquitous presence of "melt pockets" in garnets provide strong evidence of metasomatism. Type II eclogites systematically lack such features and are microstructurally more equilibrated. Type I eclogites are more magnesium-rich than most Type II (mean Mg# = 0.56 vs. 0.46), while Type II eclogites are generally more Ca-rich (mean CaO = 9 vs. 12 wt%) and Fe-rich (mean FeO = 10 vs. 12 wt%). Type I eclogites are systematically enriched in LREE, Sr, Ba, alkali elements, HFSE, Th and U compared to the more depleted Type II eclogites. Calculated trace-element patterns of fluids in equilibrium with Type I eclogites are closely similar to those of volatile-rich small-volume mantle melts in the carbonatite-kimberlite spectrum commonly inferred to be responsible for mantle metasomatism. Although oxygen isotopes are often used to argue for a subduction origin of mantle eclogites, correlations between delta O-18 of garnet and typical metasomatic tracers suggest that the metasomatic process also has shifted the oxygen-isotope compositions of the Type I eclogites toward heavier values. Roberts Victor Type I eclogites thus carry the imprint of a metasomatic process that strongly modified their major-element, trace-element and isotopic compositions, while the more pristine Type II eclogites escaped this modification. Therefore, attempts to constrain the origin of Roberts Victor eclogites should not be based on the much more abundant Type I eclogites, which retain little geochemical memory of their protoliths. The most suitable materials for such investigations may be the less metasomatised, but more rare, Type II eclogites.
ISSN: 0016-7037

hal-00669867
https://hal.archives-ouvertes.fr/hal-00669867
DOI : 10.1016/j.gca.2011.08.035