3d Modelling of (Upper Triassic) Continental Mixed Fluvial Systems Integrating “Lidar” Digital Outcrop Images with High-Resolution Sedimentology: High Atlas , Morocco

Fabuel-Perez, Ivan¹, Jonathan Redfern¹, David Hodgetts¹ (1) University of Manchester, Manchester, United Kingdom

The Oukaimeden Sandstones Formation (F5) is an Upper Triassic fluvial/aeolian formation located in the High Atlas Range of Morocco. These outcrops are an important analogue for the TAGI formation (Triassic Argilo-Gréseux Inférieur) representing one of the most productive siliciclastic reservoirs in central Algeria.

The Oukaimeden Sandstones Formation crops out in narrow tectonic belts with an ENE-WSW orientation along the major faults of the Tertiary High Atlas range. The study area offers extensive outcrops in excess of 200 m high and extending for many kilometres. The formation is composed of stacked lenticular shaped sand bodies often containing cross-bedding and current and oscillation ripples, interbedded with more tabular shape fluvial bars showing erosional bases with conglomeratic lags containing rounded pebbles of quartz, chert and felsic volcanics. Minor amounts of interbedded siltstones and mudstones are also present showing bioturbation and some lamination.

The application of high resolution 3D laser mapping (LIDAR) to these outcrops combined with Differential Global Positioning System (DGPS) mapping and traditional sedimentological data have provided a very powerful and detailed dataset from which several ‘close to deterministic' reservoir models have been produced. 31 separate LIDAR scan positions containing over a billion points has been collected covering a total area of 5 km long by 250 m high, along with a high resolution field dataset consisting of 12 sedimentary logs 3 gamma ray sections and analyses of the different geo-objects.

New approaches and methodologies have been developed in order to integrate the different data into a 3D georeferenced framework from which geostatistical information to on object geometries and facies distributions to populate the final reservoir model can be obtained.