Observation-driven AI framework for integrated geological–hydrodynamic reservoir modelling
SOCAR Proceedings · 2026 · (2) · 110-115
DOI: 10.5510/ogp20260201200 ↗Kurz gefasst
Fachartikel in der Zeitschrift SOCAR Proceedings, 2026. Verfasst von M. A. Huseynov und 1 weiteren Autor:innen.
Abstract (Originalsprache)
This study presents an observation-driven approach for integrated geological and hydrodynamic reservoir modelling using multi-source data. Reservoir systems are complex due to heterogeneous structures, nonlinear flow behaviour, and limited direct access to subsurface information. These factors make accurate modelling a challenging task in reservoir engineering. Conventional modelling approaches usually follow a sequential workflow from geological interpretation to numerical simulation. This often increases computational cost and introduces uncertainty in model construction. To address these limitations, the study explores the use of artificial intelligence methods together with physics-informed and hybrid modelling concepts. In the proposed framework, geological, petrophysical, and production data are combined in a unified processing scheme. A simple encoder–decoder structure is used to extract latent features that describe reservoir behaviour. A bridge mechanism is introduced to connect these features with physically meaningful reservoir properties such as connectivity, flow capacity, and pressure response. The method does not replace traditional reservoir simulation but aims to support it by improving data integration and reducing computational effort. A synthetic validation study shows that the proposed approach can reproduce main reservoir behaviour trends with reasonable accuracy. Overall, the results suggest that data-driven methods, when guided by physical understanding, can be a useful complement to classical reservoir modelling workflows. Keywords: reservoir modelling; artificial intelligence; data integration; physics-informed methods; hybrid modelling; surrogate models.
Zitieren
APA
Huseynov, M. A., & Mammadov, R. R. (2026). Observation-driven AI framework for integrated geological–hydrodynamic reservoir modelling. SOCAR Proceedings, 110-115. https://doi.org/10.5510/ogp20260201200
Vancouver
Huseynov MA, Mammadov RR. Observation-driven AI framework for integrated geological–hydrodynamic reservoir modelling. SOCAR Proceedings. 2026:110-115. doi:10.5510/ogp20260201200
BibTeX
@article{Huseynov2026,
author = {Huseynov, M. A. and Mammadov, R. R.},
title = {Observation-driven AI framework for integrated geological–hydrodynamic reservoir modelling},
journal = {SOCAR Proceedings},
year = {2026},
number = {2},
pages = {110--115},
doi = {10.5510/ogp20260201200}
}
EUVIMED