Karstic network simulation with γ-graphs

Benoıt Thebault and Pauline Collon and Christophe Antoine and Axel Paris and Eric Galin. ( 2022 )
in: 2022 {RING} {Meeting}, pages 18, ASGA

Abstract

For a decade, several methods have been developed to simulate the karstic conduits observed in numerous limestones rocks. They generally succeed in reproducing a specific type of network – branchwork or anastomotic, with fracture influence or not, but a single approach has often difficulty to generate the whole spectrum of karstic patterns. Lots of methods also use a grid, which resolution can strongly limit the quality of the result and cause the apparition of stair-step effects. Recently, in the frame of virtual scene simulations, Paris et al. (2021) have proposed to generate karstic networks using a least-cost path approach and γ-graphs. An adapted point sampling strategy allows to get a grid-free graph, while the γ-parameter allows to control the density of karstic networks. Using cost-computation, the method can handle various geological influencing factors, e.g. inception horizons or fracture orientations. This promising approach has been, however, only demonstrated on simple synthetic examples. In this work, we propose to integrate this simulation method in the SKUA-GOCAD geomodelling software and to confront it to typical geological data. A preliminary analysis allows us to identify the proper settings to generate the karstic network topologies that correspond to the speleogenetic context. Some modifications are also proposed to adapt the method to usual geological data like fracture Rose diagrams. First results demonstrate the potentiality of the approach and pave the way for future developments.

Download / Links

BibTeX Reference

@inproceedings{thebault_karstic_2022,
 abstract = { For a decade, several methods have been developed to simulate the karstic conduits observed in numerous limestones rocks. They generally succeed in reproducing a specific type of network – branchwork or anastomotic, with fracture influence or not, but a single approach has often difficulty to generate the whole spectrum of karstic patterns. Lots of methods also use a grid, which resolution can strongly limit the quality of the result and cause the apparition of stair-step effects. Recently, in the frame of virtual scene simulations, Paris et al. (2021) have proposed to generate karstic networks using a least-cost path approach and γ-graphs. An adapted point sampling strategy allows to get a grid-free graph, while the γ-parameter allows to control the density of karstic networks. Using cost-computation, the method can handle various geological influencing factors, e.g. inception horizons or fracture orientations. This promising approach has been, however, only demonstrated on simple synthetic examples. In this work, we propose to integrate this simulation method in the SKUA-GOCAD geomodelling software and to confront it to typical geological data. A preliminary analysis allows us to identify the proper settings to generate the karstic network topologies that correspond to the speleogenetic context. Some modifications are also proposed to adapt the method to usual geological data like fracture Rose diagrams. First results demonstrate the potentiality of the approach and pave the way for future developments. },
 author = { Thebault, Benoıt AND Collon, Pauline AND Antoine, Christophe AND Paris, Axel AND Galin, Eric },
 booktitle = { 2022 {RING} {Meeting} },
 pages = { 18 },
 publisher = { ASGA },
 title = { Karstic network simulation with γ-graphs },
 year = { 2022 }
}