Managing multi-valued horizons within a structural interpretation framework

Abstract : Horizons are first-class elements of the structural models used for Oil & Gasexploration. When they have a complex shape, they become difficult or even outright impossibleto manage within typical software packages. Handling those so-called “multivalued”horizons indeed requires tackling two problems. First, data models must be found to representthem. A review of the state of the art, complemented by a benchmark, led us to promotethe use of patch systems and triangulated surfaces in order to represent reverse-faulted horizonsand salt domes, respectively. Second, multivalued surfaces must be reconstructed fromsparse and non-uniformly distributed interpretation data. We developed two reconstructionmethods, each targeting a multivalued horizon type. They are based on standard geoscienceinterpolation methods, making them both fast and natural extensions of the state of the art.Using the proposed models along with the reconstruction methods, it is now possible to efficientlytackle the more complex fields were they are typically found.
Complete list of metadatas

Cited literature [119 references]  Display  Hide  Download

https://tel.archives-ouvertes.fr/tel-02140380
Contributor : Abes Star <>
Submitted on : Monday, May 27, 2019 - 11:35:17 AM
Last modification on : Tuesday, May 28, 2019 - 1:24:38 AM

File

BAUDRILLARD_2018_diffusion.pdf
Version validated by the jury (STAR)

Identifiers

  • HAL Id : tel-02140380, version 1

Collections

Citation

Joseph Baudrillard. Managing multi-valued horizons within a structural interpretation framework. Signal and Image processing. Université Grenoble Alpes, 2018. English. ⟨NNT : 2018GREAT108⟩. ⟨tel-02140380⟩

Share

Metrics

Record views

84

Files downloads

36