Laser Scanning and Geological Modeling: From Outcrop to Flow Simulation

John Howell
University of Bergen, Bergen, Norway

Outcrop analogues have long been used as a means to supplement the limited datasets that are available from subsurface reservoirs. As computer modeling of reservoirs developed through the mid-90s it was logical that outcrops would provide data for the conditioning of reservoir models and also be represented in the software. Initial studies concentrated on the collection of data through traditional means such as sedimentary logging, mapping and the collection of scaled drawings and photo panels.

Recent advances in surveying, especially the development of cm accurate dGPS and laser scanning (lidar) technology, have revolutionized the collection of geological field data. A ground based lidar system can scan a cliff, collecting “point clouds” of surveyed points at rates exceeding 10,000 per second. These points are used to generate extremely detailed and accurate surfaces on which high resolution digital surfaces can be draped producing a “Virtual Outcrop.”

The virtual outcrop has a number of quantitative applications that extend beyond virtual fieldtrips. Large quantities of spatial and geometric data can be collected from the reservoir analogue. These include parameters such as bed thickness and bed geometry. Surfaces can be mapped and the data exported to reservoir modelling software where the geology is recreated in 3-D. Accurate representations of the geology can then used to simulate fluid flow, providing an improved understanding of the impact that heterogeneities have on reservoir performance in the subsurface. The approach will be illustrated with two case studies from the western USA.

The Panther Tongue and Ferron Sandstone include two well-exposed fluvial deltaic units deposited under contrasting base-level conditions. Virtual outcrops were generated for both systems and a database of bed thickness changes was compiled, which illustrates the systematic thinning of delta front clinothems. The virtual outcrops were also used to build two reservoir models which were used to test the importance of Clinoform draping shales upon simulated production.

In the second case study, two fluvial systems from the coal bearing coastal plain deposits of the Blackhawk Formation and the more arid, intra continental deposits of the Colton Formation were surveyed. Virtual outcrops of each were used to build close to deterministic models for the two successions, capturing up to seven- and five-channels bodies in the two cases, respectively. Analysis of the virtual outcrops included detailed mapping of intra channel architecture and led to the building of two, close to deterministic, geocellular models (2x2x0.05 km). These models have been flow simulated and used to test the role of crevasse splay and over bank deposits in providing connectivity between channel bodies.

The science of virtual outcrop geology is still in its infancy. The talk will conclude with a brief look into the future, including the application of ground based hyper spectral methods for the remote mapping of lithology and, oblique aerial lidar (helidar) which allows the very rapid collection of huge datasets from inaccessible cliff sections.

AAPG Search and Discovery Article #90101 © 2010 AAPG Foundation Distinguished Lecturer Series 2009-2010