Applications of digital outcrop models: two fluvial case studiesfrom the Triassic Wolfville Fm., Canada and OukaimedenSandstone Fm., Morocco

摘要

The application and benefits of employing digital outcrop models (DOMs) are discussed using two Triassic fluvial case studies todemonstrate data collection and integration methods. Developments in data analysis techniques are examined to demonstratetheir utility for collecting meaningful and reliable statistical information needed to build realistic stochastic reservoir models. Toestablish a significant geostatistical dataset a large number of accurate observations are required. It is difficult to get thenecessary statistics using subsurface data alone, due to the limited resolution and/or areal coverage of respectively seismic andwell data. Outcrop studies are, therefore, commonly utilized to provide analogue statistical information (e.g. channel width, length,thickness and thickness vs. width ratio). Traditional data collection methods used in the field are however largely restricted to areaswith (easy) physical access, or using remote observations with limited accuracy, such as photographic methods.Digital data collection techniques such as LiDAR (Light Detection and Ranging) and differential GPS allow more accuratemeasurements, as well as from previously inaccessible locations, to be taken of sedimentary architecture. The techniquegenerates much larger volumes of measurements, as the area from which accurate data can be extracted is increased. This offers amore meaningful statistical dataset, hence reducing uncertainty in the final reservoir model.Both case studies, the Oukaimeden Sandstone Formation (OSF), Morocco and Wolfville Formation, Canada, are from LateTriassic braided fluvial systems. The OSF dataset has been used to illustrate how geometric information of channel width versusthickness relationships (W:T) are collected using a projection plane technique. The results show W:T variations between 3.49:1in the Lower Oukaimeden member, 1.54:1 in the Middle Oukaimeden member and 3.75:1 in the Upper Oukaimeden member,demonstrating the observed architectural evolution of the fluvial system. The Wolfville Formation case study shows how DGPSin combination with LiDAR data has been used to more accurately map faults to obtain statistical information on faultorientation (NE–SW) and length (mean = 38.3 m and median = 18.2 m). Another applied analysis technique utilizes a faciesclassified point-cloud to aid surface correlations between sedimentary logs and construct a log based correlation panel fromwhich estimates of facies frequencies are derived.