Petroleum Geoscience - Volume 1, Issue 2, 1995

The electrical behaviour of fully electrolyte-saturated reservoir rock samples has been represented through verified type curves as a continuum across a range of electrolyte resistivity. The type curves can be used to facilitate the well-log interpretation of clastic reservoirs for which the Archie porosity exponent is known or can be realistically assumed. They allow an immediate appreciation of the electrical characteristics of reservoir rocks and how these relate to other, previously studied formations. The type curves can be used with core or log data to infer whether a reservoir is behaving as effectively clean or shaly. In the case of effectively shaly sands, a single-point value of excess conductivity can be evaluated from log data alone. More generally, the type curves allow different reservoirs to be compared in terms of their electrical behaviours. Where these are similar, it is reasonable to assume that the petrophysical interpretation procedures will be similar for the reservoirs under consideration. Thus, the type curves allow a given reservoir to be characterized and evaluated within a generic framework of electrical behaviour. This, in turn, allows experience gained elsewhere to be brought to bear on the current appraisal, especially in terms of the identification of appropriate procedures for the well-log evaluation of hydrocarbons-in-place. Through this insight, the method constitutes a potential basis for greater confidence in petrophysically-supported reservoir models at an earlier stage in the evaluation of a field.

The full waveforms recorded by an array of receivers in a modern full-waveform sonic tool contain secondary arrivals that are reflected from near-borehole structural features. These arrivals are used to form an image of the near-borehole structural features in a manner similar to seismic migration. Possible uses of this technique include horizontal well logging; structural dip and contour determination; fault, salt dome, pinnacle reef, and fracture zone imaging; and EOR steam-flood monitoring. Examples are given for a deviated well penetrating a North Sea reservoir, and for a horizontal well penetrating a thin (10m thick) reservoir. For the horizontal well case, full-waveform sonic data are successfully used to image the reservoir top as the well penetrates into the reservoir and continues near-parallel to the reservoir boundary.

Reservoir characterization of the Sherwood Formation in the Wytch Farm Field depended on control from 48 wells, approximately half of which had been cored. Lithotypes were denned by detailed study of a near-by outcrop, where photo mosaics were mapped to establish lithotype geometry for the reservoir model. Probe permeameter measurements were taken both at outcrop and in the laboratory to supplement existing core plug data. The key to achieving an optimum well trajectory through the reservoir is appropriate definition of the kv/kh ratio. Lithological data were assimilated to derive characteristics for input to flow simulation models. These models were developed at the single-well scale and reservoir characterizations then 'up-scaled' to full-field reservoir flow models. These studies indicated that lenses of mudstone and calcrete conglomerates have a significant effect on reservoir flow characteristics. Thus, to optimize individual well performance, a trajectory at 85 degrees to 87 degrees would be more effective than a horizontal well.

The South Kavala Field will have produced 670 x 10 6 Sm 3 of its original recoverable gas reserves of 1.0 x 10 9 Sm 3 by the end of 1994. Gas storage of imported gas is planned, and to provide continuity of supply and to reduce the risk of induced seismicity caused by pressure variation in the reservoir, the remaining 300 x 10 6 Sm 3 will be used as a cushion. the field lies nearby seismically active areas so the seismic hazard is evaluated. Even though the seismicity of the area is low, it is considered prudent to install a dense network of seismographs in order to monitor the effects of small magnitude (M s < or = 2.5) earthquakes.

In the Romanian East Carpathians overthrust relations between the Tarcau, the Marginal folds and the Subcarpathian nappes and the Moldavian and Moesian platforms have been deduced from seismic profiles and confirmed by wells. The down steps of the platform, under the nappes, have created many faulted blocks. South of the Trotus Valley, the Pralea-Focsani Depression has developed on the platform basement. Between the Slanic and Dambovita valleys, in the Mio-Pliocene zone, the Subcarpathian nappe, the diapir folds and the Oligocene deep structure have been delineated on seismic profiles. The presence of all the conditions for oil generation and entrapment encourages new investment in geological and geophysical exploration.

The Villeperdue oil field is a composite stratigraphic and structural trap. The study focuses on the stratigraphic control, i.e. the relationship between stratigraphy and sedimentology, and a sedimentological model constrained by core data, outcrop analogues, and thorough petrographic analyses. The Callovian oolite limestones are subdivided into three stratigraphic units each with its own oolitic shoal facies; the upper two units form two separate complex reservoirs with porosity zones which are hard to correlate though they are probably interconnected.

In the interpretation of structures from fragmentary subsurface geological data, the modelling attribute most commonly employed is that of model geometry or form. Other model attributes can provide insights to exploration-related problems associated with a structure. The nature of these attributes depends on the modelling approach (i.e. materials, boundary conditions, initial conditions); several are unique to a particular approach. The integration of data from several modelling approaches, each aimed at specific attributes, allows the most complete understanding of a structure. This leads to the development ol conceptual models of the origin, evolution and expression of structures as a function of initial conditions and boundary conditions. Such models provide 'rules of thumb' that guide interpretations in structurally complex areas. Laboratory and mathematical modelling attributes have been used here to improve our understanding of the deformation of the hanging wall above a thrust ramp.

Scaled analogue models of 2D and 3D extensional fault systems provide powerful templates for the seismic interpretation of complex fault structures. 2D sandbox models of extensional faults that link to a lower detachment surface are characterized by roll-over structures together with crestal collapse graben and antithetic fault systems. The geometry of the roll-overs are controlled by the shape of the main detachment surface. 3D extensional fault systems in sandbox models of orthogonal and oblique rift systems are characterized by segmented planar fault systems that grow along strike, link, and overlap forming relay ramp structures and accommodation zones. In orthogonal rifts the border faults and the intra-rift faults are oriented at high angles to the extension direction. In oblique rifts the border faults are parallel to the zone of rifting whereas the intra-rift faults are oriented at a high angle to the extension direction.

Scaled experiments, whose design was based on recently published low values for the viscosity of natural rocksalt, suggest that under typically low geological extension rates, a salt layer 500 m thick or thicker decouples the brittle overburden from the faulted basement. Because of its low viscosity, salt cannot transmit the large differential stresses necessary for basement faults to propagate upward as faults or abrupt forced folds into the brittle overburden, unless the salt layer is thin or has been depleted. Thick salt diffuses localized deformation in the basement by flowing away from rising basement blocks, and toward sinking basement blocks. Basement extension is not transmitted to the overburden vertically but is laterally transmitted from the margins as the salt basin widens. Rather than being controlled by basement faults, the distribution of faults and diapirs in the overburden is controlled by lateral changes in overburden stratigraphy or thickness, preexisting faults, and faults that propagate laterally into the salt basin.

Dynamically scaled analogue models with an overburden of cohesive sand and a viscous "salt" layer were deformed in a centrifuge to mimic real salt structures in the Nordkapp Basin. Like their natural prototypes, model diapirs were aligned in rows parallel to the basin margins. In profile, model diapirs were asymmetric, suggesting that the real diapirs will possess asymmetric geometries. Like many of the real salt structures, model diapirs pierced without developing a pillow stage because they rose along basin margin faults which propagated up through the overburden from the basement during thick-skinned extension. Once their overburden was weakened by faulting, differential loading forced model diapirs to rise as long as buoyant material was supplied. Some real salt diapirs initially rose as conformable pillows during the early Triassic, became diapiric during middle Triassic and spread broad overhangs during slow sedimentation in late Triassic and Jurassic times. Later, the overhangs reactivated asymmetrically when buried by Cretaceous and Tertiary sediments.