EAEG/EAPG/EAGO Joint Multidisciplinary Workshop - Developing New Reservoirs in Europe

Progressive increase in cost of discovery and development of new oil and gas fields that depends on composition of oil and gas reservoirs require the new technologies which provide detailed studying of geological peculiarities of the objects with optimal decreasing of volume of expensive drilling.

The topic under discussion is mathematical models and computer techniques developed at VNIGRI for simulation of reaI exploration process which provide basis for sound strategie decisions of future oil and gas exploration efforts. Decisions are based on the results of multivariant geo-economic analyses, comparisons and their conformity to certain geological limitation and conceptions. Validity of the above mentioned technologies was testedby results of exploration in many basins of Russia.

The South Central Trough of the North Sea contains numerous oil and gas fields in the Upper Cretaceous Chalk section. Many of these fields occur along the northwest-southeast trending Lindesnes Ridge. Amoco Norway Oil Company is the operator of the Valhall and Hod Fields that are located on the structural highs of this ridge, where the Chalk has been fractured and deformed during uplift.

We have designed a new procedure that combines fast horizon tracking and edge detection with precise 3-D editing to significantly improve the speed and quality offault interpretation on 3-D seismic workstations. The procedure works as foIIows: • First, we auto-track horizons at speeds 5 to 10 times faster than currently available on equivalent workstations. We achieve this speed improvement through the use of a proprietory technology. • Second, we detect horizon edges; an edge being the locus of points where the horizon dip and amplitude variation along the dip direction exceed thresholds that are interactively specified. • Third, we use a 3-D interactive editor that zooms in and edits automatically detected horizon edges with great precision. Editing is done on projections of the fault trace on three orthogonal planes. These planes are dynamically linked, so any change in one plane is concurrently updated in the other two planes, and also in a perspective rendering of horizons and fault edges. • Fourth, we assign edge segments belonging to the same fault to a common fault set. • Fifth, we augment the fault set of horizon edges with fauIt segments interpreted on vertical seismic sections. • Finally, we fit a surface through segments in the common fault set and render the surfacetogether with horizons in a 3-D perspective view. In order to avoid any ambiguity due to multi-valued surfaces, we use a data-adaptive coordinate system to fit the surface. In the case of subtIe faults, which are sometimes only visible as discontinuities on an attribute map, we perform edge detection on the attribute map instead of on the horizon dip and amplitude maps.

Seismic data represents an elastic impression of given distributions of reservoir parameters. The relationship between the seismic response and the distribution of reservoir parameters are typically non-unique. Additional constraints are therefore needed if reservoir parameters should be predicted from the seismic data. These constraints are normally model-dependent and might as such not be universally valid. To select the appropriate constraints for the problem at hand often requires expert knowledge. Sometimes this expert knowledge span multiple domains as seen from Figure 1.

A novel technique for 3-D constrained, parametric, post-stack seismic inversion is presented. Deterministic integration is performed of seismic data, log data (for instance sonic and density) derived from either vertical or deviated wells, formation tops and time interpretations. One horizon in depth is used as a reference surface for the tirne-to-depth conversion. This study shows the potential benefits of the method. The properties of this inversion method are exemplified by a feasibility study using data from the Snorre Field in the North Sea. The goal of this study is to optimally define the geological model for the reservoir formations using this 3-D inverse modelling tooI. As a result of this inversion study detailed sonic and density models are derived, both in two-way travel time and depth.

In a complex structural environment where several generations of faults were at play, interpretation is a long iterative process. Structural models are slowly built from sections and checked for consistency by means of 3D displays. Unlike horizons, fault pieking is usually done manually, since faults are not controlled by seismic attributes. 3D surface displays provide valuable insight into the data structure and enable model QC and validation, as well as determination of model inconsistencies. In the latter case, model construction has to be done over again, This model construction phase is definitely a bottle-neck during an interpretation process. Use of 3D modelling to insure 3D consistency of the model, and to reduce the interpretation time, is illustrated on a 3D data set from the North Sea.

An accurate assessment of oil and gas reservoirs requires a detailed interpretation of the 3D seismic data. These data can be transformed into maps of acoustic impedances which facilitate the interpretation of reservoir structure and distribution, and provide more accurate estimates of average reservoir properties such as porosity. But neglecting nonseismie knowledge in such a process could he highly detrimental to the accuracy of acoustic impedance estimates.

The prospect under study is a shale dominated submarine fan of late Jurassic age located in the UK sector of the North Sea, North-West Europe. The trap consists of a horst block which contains a dipping wedge of Jurassic sediments, truncated and sealed by the erosional unconformity of the Base Cretaceous. The objective of this study is to identify those parts of the fan where the net sand content is highest and thereby choose targets for appraisal drilling that maxi mise the potential recovery from the prospect.

This paper presents the results of a joint investigation between the Geological Institute of the Azerbaijan Academy of Sciences (GIA) and the BP and Statoil Alliance into the stratigraphy and sedimentology of the Neogene Productive Series at outerop on the Apsheron Peninsula in Eastern Azerbaijan. The Productive Series is essentially Kimmerian in age, but also includes some sediments of earliest Akchagylian age.