Robustness and Transferability of Pix2Geomodel for Bidirectional Facies Property Translation in a Complex Reservoir

2026-05-05

AI Summary

This study addresses the challenge of capturing nonlinear relationships between lithofacies and petrophysical properties in complex reservoirs, where data sparsity and strong heterogeneity hinder conventional modeling approaches. To this end, the authors propose a bidirectional image-to-image translation model based on the Pix2Pix framework, demonstrating for the first time its cross-domain transfer capability in vertically low-resolution and highly heterogeneous reservoir settings. The model employs a U-Net generator and a PatchGAN discriminator, enhanced with geometric augmentation and spatial continuity constraints to enable mutual translation among lithofacies, porosity, permeability, and clay volume (VCL). Experimental results show that the model effectively preserves geological structures and spatial continuity, achieving a pixel accuracy of 0.9326 in the lithofacies-to-porosity task and a mean Intersection over Union (IoU) of 0.7049 in the VCL-to-lithofacies task, thereby significantly improving the practicality and robustness of reservoir modeling in complex geological environments.

Abstract

Reservoir geomodeling is central to subsurface characterization, but it remains challenging because conditioning data are sparse, geological heterogeneity is strong, and conventional geostatistical workflows often struggle to capture nonlinear relationships between facies and petrophysical properties. This study evaluates the robustness and transferability of Pix2Geomodel on a different and more complex reservoir dataset with reduced vertical support. The new case includes a heterogeneous reservoir-quality classification and only 54 retained layers, providing a stricter test of whether Pix2Pix-based image-to-image translation can preserve facies-property relationships under constrained data conditions. Facies, porosity, permeability, and clay volume (VCL) were extracted from a reference reservoir model, exported as aligned two-dimensional slices, augmented using consistent geometric transformations, and assembled into paired image datasets. Six bidirectional tasks were evaluated: facies to porosity, facies to permeability, facies to VCL, porosity to facies, permeability to facies, and VCL to facies. The Pix2Pix model, consisting of a U-Net generator and PatchGAN discriminator, was evaluated using image-based metrics, visual comparison, and variogram-based spatial-continuity validation. Results show that the model preserves the dominant geological architecture and main spatial-continuity trends. Facies to porosity achieved the highest pixel accuracy and frequency-weighted intersection over union of 0.9326 and 0.8807, while VCL to facies achieved the highest mean pixel accuracy and mean intersection over union of 0.8506 and 0.7049. These findings show that Pix2Geomodel can transfer beyond its original case study as a practical framework for rapid bidirectional facies-property translation in complex reservoir modeling.

Problem

• reservoir geomodeling
• facies-property translation
• geological heterogeneity
• sparse conditioning data
• nonlinear relationships

Innovation

• Pix2Pix
• bidirectional translation
• reservoir geomodeling
• facies-property relationship
• spatial continuity