The Tabarka Injection Complex (TIC) consists of sandstone intrusions sourced from slope sandstone channels and emplaced into deep-marine mudstone and siltstone (Oligo-Miocene) of the Numidian Flysch, Northern Tunisia. The internal and external architecture of these sandstones is extremely complex due to the range of sand remobilisation and intrusion processes that modify depositional units and generate sandstone intrusions. Depositional facies record deposition from gravity flows in sand-rich slope channel complexes. Remobilised facies demonstrate post-depositional dewatering and fluidisation in the parent sandstones with abrupt lateral and vertical transitions into intrusive sandstones emplaced into hydrofractures in host mudstones. Surface marks developed along the walls of intrusions and other sedimentary structures in sandstone intrusions are interpreted to document high velocity erosive sand-laden turbulent flows and the transition to higher concentration laminar flow. Post-injection, sandstone intrusions were modified by dewatering, compaction and local bioturbation. The TIC is 50 m thick and is divided into five stratigraphically constrained units: (i) units 1 to 3 are 4 to 18 m thick and comprise turbiditic channel sandstone, which transition upward and toward the channel margins into extensively de-watered and remobilised sandstone, which in turn feed sandstone dykes and sills; (ii) unit 4 and 5 are 12 and 15 m thick and contain mostly sandstone intrusions that form an intricate complex of sills and dykes (up to 1.2 m thick). Previous outcrop studies on the origin of sandstone intrusions in the Numidian Flysch proposed syn-depositional downward injection or early post-depositional liquefaction and intrusion into adjacent host mudstone. Our data, instead, indicate that the sandstone intrusions formed by post-depositional upward and lateral injection (10's of meters) of fluidised sand from shallowly buried slope channels into the adjacent host mudstone, most likely in two or three injection events. The presence of sandstone sill-dominated units confirms that supra-lithostatic pore-fluid pressure (P f) was attained in the parent sandstone during early burial and dyke strike orientation was controlled by the regional stress field related to the Magrhebian compression.

The giant Johan Sverdrup oil field on the Utsira high in the Norwegian sector of the North Sea is an example of a multibillion barrel oil discovery in a mature and prolific super basin. After 50 yr of exploration in the region, the field had eluded the oil industry. Wells drilled (from 1967 to 2006) had oil shows in both basement and traditional sandstone reservoir rocks. The first significant discovery in the area was not made until 2007 with the Edvard Grieg field, shortly followed by the discovery of the Johan Sverdrup field in 2010. The mainly Jurassic-aged Statfjord and Viking groups constitute the two principal reservoirs in the field containing reserves in the order of 2.2-3.2 BBOE within a productive area of approximately 200 km(2) (similar to 124 mi(2)). Reservoir properties are excellent, with multi-Darcy permeabilities. The stratigraphic succession of the area, from Upper Triassic-Lower Jurassic alluvial deposits at the base to Upper Jurassic-Lower Cretaceous open-marine deposits at the top, records long-term subsidence and marine transgression, although interrupted by significant Middle Jurassic thermal uplift. During this time, accommodation was influenced by phases of extension and the creation of oblique-slip faults, previously not described in this region. The resulting stratigraphic architecture provided the basis for one of the largest oil fields in the North Sea super basin and a major supplier of energy for decades to come.

Giant sand injection complexes form, intricate, basin-scale fluid plumbing systems and document the rem obilisation and intrusion of several tens of cubic kilometres of sand within the shallow crust in stratigraphic units 100's metres thick. This is the first detailed and extensive account of the Panoche Giant Injection Complex (PGIG), aregionally significant outcrop (>300 km(2)) and part of a larger subsurface development (>4000 km(2)) identified in boreholes and on seismic reflection data. Magnificent exposure of the PGIC occurs along the north western margin of the San Joaquin Valley and presents the opportunity to examine the regional geological significance of a giant sand injection complex and its origin in the context of a late Cretaceous - early Paleocene forearc basin. Between 25 and 49 km(3) of sand were remobilised and injected, at least 0.35 km(3) of which extruded onto the paleo-seafloor. large sandstone intrusions often >10 m thick and laterally extensive on a kilometer scale formed saucer-shaped intrusions, wing-like intrusions and a variety of sill geometries along with volumetrically smaller randomly oriented dikes in a 200-300 m thick interval. Dikes prevail below and above this interval, some reaching the paleo seafloor and extruding sand. Networks of propagating hydrofractures form intensely brecciated host strata, some of which were intruded by sand. All intrusions formed in a single pulsed event in which the most intense hydrofracturing caused by supra-lithostatic fluid pressure occurred approximately 600 to 800 m below the paleo seafloor. A crudely orthogonal arrangement of dikes is preserved with most oriented normal, and less commonly oriented parallel to the oceanic trench associated with the late Mesozoic to early Tertiary North Pacific subduction. Dikes orthogonal to the trench opened against the minimum horizontal stress, which was parallel to the trench. Dikes parallel to the trench opened against the regional maximum horizontal stress along minor faults formed in extension caused by shallow crustal deformation. There is no evidence that compressional tectonics influenced the onset of elevated pore fluid pressure necessary to promote sand injection. However, tectonic compression was responsible for creating the basin physiography that locally increased subsidence and accelerated chemical diagenesis in the basin centre. PGIC outcrop, located along the basin margins, was unlikely to have experienced heating above 70 degrees C, equivalent about 2 km burial, so the effects of chemical diagenesis in the host strata of the injection complex had negligible potential to evolve significant pore water volume. In a deeper part of the basin approximately 150 km to the south, lateral equivalents of the host strata were subjected to heating >100 degrees C and would expel significant volumes of water displaced by quartz cementation and clay dehydration that caused lateral pressure transfer to the north and western margin of the basin where the PGIC formed. Estimates of the total volume of water expelled from the deep basin suggest that a fluid volume equivalent to a gross rock volume reduction <1% would have provided a fluid budget sufficient to fluidise and inject the sand that forms the PGIC. In terms of areal and vertical extent, volume and architecture the PGIC shares strong similarity with the regionally developed giant injectite systems of Tertiary age in the North Sea basin. In both cases regional sand injection is genetically linked to pressure transfer toward the basin margin from more rapidly subsiding basin centres. Aqueous fluid is derived from thermally driven chemical diagenesis of thick deep water clastic sandstone and smectitic mudstone or from deeper, stratigraphically older, aquifers.

Sandstone intrusions in giant injection complexes are characterized by texturally immature sand with common micro-fractured framework grains. Individual micro-fractures are distinctive in geometry and unaligned within or between grains, thus differentiating them from micro-fractures formed by shock metamorphism or tectonics. Individual grains preserve histories of multiple impacts. The geometry of micro-fractures and their textural association makes them diagnostic of high-energy inter-granular collisions during sand injection. Mudstone clasts have sand-propped micro-fractures associated with hydraulic fracturing and individual sand grains embedded in clasts by corrasion, which is diagnostic of high grain velocity. Heavy mineral assemblages record abrasion of apatite and hydrodynamic segregation of zircon (both relative to abundance of tourmaline) upward through the injection complex. Granular abrasion and hydrodynamic segregation are consistent with turbulent flow during sand injection. Collectively the petrographic and mineralogical data support the interaction of high-velocity grains in turbulent flow during sand injection in which the granular content is likely to be dilute.

Excellent exposure from part of the Panoche Giant Injection Complex in the San Joaquin Valley is used to examine provenance characteristics of sandstone intrusions with respect to two parent sandstone units that are known to feed the sand-injection complex. The succession is part of the upper Mesozoic to lower Tertiary Great Valley Group, and was deposited in a deep water part of an evolving deep-water forearc basin. The section examined is mudstone-dominated, and the sand injection is constrained to have occurred in the Danian. Sandstones in the Dosados Member (Moreno Fm) are identified as the main parent unit on the basis of total heavy-mineral-assemblage compositions and varietal studies of selected minerals (tourmaline, garnet, titanite, apatite, and zircon). Fluidized sand is emplaced in turbulent flow conditions creating high-velocity inter-grain collisions. Evidence of comminution and diminution of minerals that are less hard than quartz is documented using indices for the relative hardness (TAH) and durability (TAD) of heavy minerals. Preferential settling of high-density zircon relative to lower-density tourmaline produces density-controlled variations of zircon: tourmaline upward through the injection complex. Heavy-mineral dissolution occurred in the most permeable sandstone intrusions and is believed to record the effects of mid-Eocene deep weathering, when subtropical climate prevailed in the study area. Detrital heavy-mineral assemblages, which are dominated by titanite and garnet, record erosion of the Sierran metamorphic terrane with mafic and alkaline plutonic rocks. Zircon with U/Pb ages of c. 140-160 Ma and c. 90-110 Ma, consistent with earlier independent analyses, record erosion of Sierran granitoids. On the paleo-seafloor, enrichment of Ca-amphibole and epidote is indicative of Sierran provenance concurrent with sand extrusion. The presence of Na-amphibole in the Uhalde Sandstone supports earlier work that suggested sediment input from obducted seafloor to the west.

Current recovery from the Statfjord Group in the majority of the fields on the Tampen Spur is less than 50%. A contributing factor to this is an incomplete understanding of multiscale heterogeneities, their distributions within a range of fluvial geobodies and their lateral extent and morphology in inter-well areas. Sedimentary heterogeneities have been modelled, together with petrophysical parameters, at a variety of scales. The modelled properties at a given scale were upscaled to the next level of heterogeneity, thus better honouring effective property values. The use of outcrop analogues is still a key tool for understanding facies relationships and the stratigraphic development of subsurface hydrocarbon-bearing reservoirs. The Lourinha Formation, Portugal, was used as an analogue to collect both qualitative and quantitative data consistently following a three-phase workflow to capture data at various scales of heterogeneity. Traditional field data collection techniques have been supplemented with the collection of LiDAR data. A digital workflow utilizing interlinked datasets facilitates rapid data analysis and better data visualization with results that are more easily utilized in multiscale modelling studies. These scaled models were used to increase our understanding of the effect on flow of litho-facies and facies association distributions together with internal architectural elements and heterogeneities.

Almost two hundred years of research is reviewed that focuses on the physical characteristics of sandstone intrusions. It is concerned with mechanisms of sand injection, particularly with fluid-grain transport and sedimentation processes during the remobilization, injection and extrusion of sand. Outcrop and subsurface studies in combination with laboratory experimental data are drawn on to present the state-of-the-art of sand injection. The text covers 1) geometry, internal structure, and microtexture of deformed parent units, injected and extruded sandstones, 2) host-strata and their seal characteristics that contribute to basin-wide overpressure generation, 3) common trigger mechanisms for sand injection such as high magnitude seismicity and the rapid injection of large volumes of fluids, 4) fluid types that drive sand into fractures, 5) hydrofracture mechanisms that induce regional-scale seal failure, 6) liquefaction and fluidization processes that transport sand into fractures, 7) sedimentation processes in fractures, 8) the flow regime of fluidized sand during injection, 9) post-sand-injection fluid flow and diagenesis, 10) porosity and permeability characteristics of injected sandstones and 11) post-sand-injection fluid-flow over geological timescales. Processes of sand remobilization, injection, and extrusion are complex and depend on many interrelated factors including: fluid(s) properties (e.g. pressure, volume, composition), parent unit and host-strata characteristics (e.g. depositional architecture, grain size and distribution, clay-size fraction, thickness, permeability) and burial depth at the time of injection. Many studies report erosional contacts between host strata and injected sands and these record high-velocity, erosive flow during injection. The flow regime is poorly constrained and similar features are interpreted as records of laminar and turbulent flow, or both, during injection. Internal structures are common in sandstone intrusions and can be accounted for by a variety of processes. The interpretational limits largely result from a lack of laboratory experiments that focus on developing analogues for sand injection. The relationship between grain fabric developed during injection and its control on permeability in sandstone intrusions is poorly understood and failure to advance this field of research will hinder the quantitative characterization of sandstone intrusions as fluid-flow conduits during basin evolution. We conclude that future research should focus on: 1) quantification of sediment transport modes under different flow conditions in different fracture dimensions with laboratory data relevant to sand injection; 2) estimation of the effect of injection on the bulk permeability of otherwise low-permeability seals (host strata) so that their effect on fluid flow can be assessed at all scales; and 3) incorporation of sand injection into quantitative basin models. Although an enormous amount of data have arisen from existing studies there remains a need to advance many fields of research related to sand injection so that the significance of these important structures can be fully appreciated in the geological record.

The Yellowbank Creek Injectite Complex (YCIC) has excellent exposure of the external geometries and internal structures in a large-scale sandstone injectite. The YCIC has an irregular geometry that can be defined neither as a sandstone dike nor as a sill. Substantial erosion occurred along the margins between the YCIC sandstone and host mudstones that includes scoured basal margins and a scalloped upper margin that cuts up to 5 in into the overlying host mudstones. Erosion of host mudstone has incorporated significant volumes of mudstone clasts into sandstones at the margins of the injectite. Using microtextural data corrasion is identified as a typical process by which mudstone is eroded during sand injection. Corrasion occurs when particles of mudstone are either cut or broken from a cohesive bed as the result of high-velocity turbulent flow of sand grains acting as tools. Corrasion together with scouring and the general erosion of the host mudstone imply high-velocity turbulent flow conditions and confirm that the fluidized sand had a sufficiently low viscosity, during sand injection, to allow turbulence.Internal structures include laminae of more tightly packed sand grains than occur in the surrounding sandstones, and are interpreted to have formed by flow and deposition of injecting sands during waning of flow, thus reflecting an overall reduction of pore-fluid pressure. Deformation of laminae into folds defines bands, which are aligned parallel to the margins of the sandstone. The force driving deformation is likely to have been shear stresses induced between the injected sandstone and the more rigid host mudstone. The principal style of deformation is hydroplastic. Oversteepened laminae and pipes are interpreted to have formed due to post-sand-injection consolidation and resulted in only minor mobilization of sand.

An architectural hierarchy of clastic sills is recognized in the Panoche Giant Injection Complex in which staggered, stepped with erosive top surfaces, and multi-layered geometries occur in that stratigraphic order upward. Genetic relationships between parent depositional sand bodies, the sand injections, a zone of hydraulic fracture and a palaeo sea floor are seen at a scale previously observed only by using seismic data. Sills and randomly oriented dykes intrude into a hydraulically fractured shale unit above and below which dykes predominate. Erosive surfaces (scallops) are identified on sills that, along with smaller erosional features, record low-viscosity turbulent flow during sand injection. Sand extrusions occur where dykes reach the palaeo sea floor.