Development and interaction of segmented fault systems
Faults in contractional and extensional settings have been studied, both during fieldwork and from interpretation of 2D and 3D seismic sections. Detailed analysis of the geometries of faults, with displacements of a few mm to 100's m, has provided insights into the development of fault networks. Faults evolve from the amalgamation of initially isolated fault segments, and deformation is eventually localised onto 'master' faults. A model for the development of conjugate strike-slip fault zones at Beadnell, Northumbria, and Kilve, Somerset, involves (1) random development of vein arrays, to (2) isolated development of several unconnected conjugate faults, through (3) intersection of a conjugate set of master faults and linkage with minor antithetic faults, and formation of new vein arrays, to (4) breaching of intersection points by dominant faults, and finally (5) propagation towards oversteps that are breached to form a through-going fault. The geometry of the active structures simplifies, as strain is localised along master faults. The conjugate strike-slip faults at Kilve are associated with reversereactivated normal faults, and are related to Alpine-related contraction. Selective reactivation is attributed to: (1) low fault plane frictions, (2) domino block rotation, (3) displacement magnitude, and (4) a percolation model. The style of displacement transfer between the reverse-reactivated normal faults is dependent on the original fault network geometry. Conjugate strike-slip faults accommodated the shortening between two overlapping reverse-reactivated normal faults. Obliquely orientated reversefaults perform a similar role in an underlapping fault zone. The segmented geometries of normal faults, in cross-section and map view, in the Wytch Farm Oilfield are described. The deformation is dominated by recurrent normal fault reactivation, onto which the extension is localised. The localisation is more intense in the Triassic, with a more homogenous strain distribution in the Jurassic. The reactivated faults are identified from multiple growth sequences. Displacement profiles do not show a simple upward decrease in displacement, but have one displacement maxima in the Triassic, and one in the Middle Jurassic. The throws in the Jurassic are often greater than those in the Trias, but the heaves are less due to steeper faults. The geometries are attributed to thin-skinned extension of the Jurassic sequence above a postulated detachment in the Upper Triassic or Lower Jurassic. The existence of a detachment questions the practice of locating the poorly-imaged Sherwood Sandstone reservoir unit on 2D seismic sections from an isopach from the White Lias horizon. Uncertainty in the structure of the Sherwood Sandstone exists due to the quality of a 2D seismic data set, and a gap that exists between two 3D seismic surveys. Fault analysis methods have been applied to the re-interpretation of the Sherwood Sandstone structure, and new maps are presented that represent the uncertainties. These illustrate that the 2D seismic data can be interpreted in several ways, based on variation of throw, geometry and position of the fault system. A comparison of the geometries of the normal fault systems exposed at Kilve with those in Wytch Farm reveals that there are many geometrical similarities. There is, however, no evidence for strike-slip faults, or reverse-reactivated normal faults at Wytch Farm.