Cristin-resultat-ID: 1406934
Sist endret: 1. desember 2016, 10:11
Resultat
Poster
2016

Virtual Outcrop Models to Multiple Point Statistics: Improved Reservoir Modeling From Virtual Outcrops Supported by Digital Field Computing

Bidragsytere:
  • James Robert Mullins
  • John Anthony Howell
  • Christian Kehl og
  • Simon John Buckley

Presentasjon

Navn på arrangementet: AAPG Annual Convention and Exhibition
Sted: Calgary
Dato fra: 19. juni 2016
Dato til: 22. juni 2016

Arrangør:

Arrangørnavn: AAPG

Om resultatet

Poster
Publiseringsår: 2016

Beskrivelse Beskrivelse

Tittel

Virtual Outcrop Models to Multiple Point Statistics: Improved Reservoir Modeling From Virtual Outcrops Supported by Digital Field Computing

Sammendrag

Outcrop analogues play an important role in modeling reservoir heterogeneity. Recent advances in digital outcrop mapping methods, such as lidar and photogrammetry, permit the rapid acquisition of high-resolution 3D “virtual outcrop” models, which capture the detail of the outcrop in the computer. However at present utilizing those data in geocellular reservoir models relies on the manual measurement of object dimensions or variograms. Multiple-point statistics (MPS) is a relatively new property modeling technique which is reliant on defining representative training images (TI). A TI is a conceptual numerical description of geology that encompasses the expected patterns, and variability of heterogeneity in the reservoir under study. To date, TI development is generally based on subjective criteria and the experience of the modeler. Constructing and selecting representative TIs of the reservoir under study and maintaining compatibility with the available borehole and geophysical data – especially in 3D – is a key challenge. Virtual outcrop models provide a critical and underused source of quantitative and qualitative information for generating high quality TIs for constraining MPS simulations. Furthermore, advances in mobile computing (smartphones and tablets) have the potential to allow field-based interpretation and analysis of virtual outcrop models and bridge the gap between geological insight in the field and its inclusion in a 3D reservoir model. Accurate modeling from outcrop data is facilitated when each observation is tied to its correct 3D spatial position while maintaining the 3D integrity and scale of geological features. Surface modelling, volumetric geobody extraction and calibrating the virtual model with auxiliary data in the field can be used to guide TI creation at the conceptual level and provide iterative feedback between the TI and 3D digital outcrop. This research develops a standardised training image database derived from virtual outcrop data across a range of depositional environments. This will significantly improve the prediction of the 3D variation of facies heterogeneity and its impact on reservoir performance. Improved variation predictions are particularly important in less mature and frontier areas where more targeted modelling data is required.

Bidragsytere

James Robert Mullins

  • Tilknyttet:
    Forfatter
    ved University of Aberdeen

John Anthony Howell

  • Tilknyttet:
    Forfatter
    ved University of Aberdeen

Christian Kehl

  • Tilknyttet:
    Forfatter
    ved NORCE Energi og teknologi ved NORCE Norwegian Research Centre AS

Simon John Buckley

  • Tilknyttet:
    Forfatter
    ved NORCE Energi og teknologi ved NORCE Norwegian Research Centre AS
1 - 4 av 4