Sequential Physics-Constrained Neural Operator Forward Modeling for the $\textit{Norne}$ Reservoir System
arXiv:2605.28909v1 Announce Type: new Abstract: We develop a comprehensive mathematical and computational framework for sequential surrogate modeling of three-phase black-oil reservoir dynamics using neural operators, with particular emphasis on Fourier Neural Operators (FNO) and their physics-informed variant (PINO). The application focus is the Norne benchmark reservoir, defined on a heterogeneous $46\times112\times22$ grid ($N=113,344$ cells), with a production history spanning $T=30$ timesteps covering 3298 days. Our theoretical contributions are organized around four interlocking problems
The increasing sophistication of neural operators and physics-informed AI models is enabling their application to complex, real-world engineering problems like reservoir dynamics.
This development indicates a growing capability to use advanced AI for more accurate and efficient simulation and management of critical resource extraction, potentially optimizing output and reducing environmental impact.
The prior reliance on traditional numerical simulators for reservoir modeling can now be augmented or potentially replaced by faster, more physics-aware AI models, changing how energy resources are forecasted and managed.
- · Energy companies
- · AI/ML research labs
- · Computational fluid dynamics experts
- · Reservoir engineering software providers
- · Traditional reservoir simulation software vendors
Improved efficiency and accuracy in modeling oil and gas reservoir performance, leading to more informed drilling and production decisions.
Reduced operational costs and enhanced recovery rates for hydrocarbon assets through better predictive capabilities.
Extension of these AI-driven simulation techniques to other complex geological or industrial processes, potentially accelerating energy transition technologies like geothermal or carbon capture.
This signal links to a primary source. Continuum Brief monitors and indexes it as part of the live intelligence stream — we do not republish source content.
Read at arXiv cs.LG