Houston Advanced Research Center

Accurate geological modeling is essential for reservoir characterization, yet traditional methods struggle with complex subsurface heterogeneity and the conditioning of data to observed values. This study introduces Pix2Geomodel, a novel conditional generative adversarial network (cGAN) framework based on the Pix2Pix architecture, designed to predict key reservoir properties (facies, porosity, permeability, and water saturation) from the Rotliegend reservoir of the Groningen gas field. Utilizing a 7.6 million-cell dataset from the Nederlandse Aardolie Maatschappij, accessed via EPOS-NL, the methodology included data preprocessing, augmentation to generate 2,350 images per property, and training with a U-Net generator and PatchGAN discriminator over 19,000 steps. Evaluation metrics include pixel accuracy (PA), mean intersection over union (mIoU), and frequency-weighted intersection over union (FWIoU). Performance was evaluated in two tasks: (i) masked property prediction and (ii) property-to-property translation. Results demonstrated high accuracy for facies (PA 0.88, FWIoU 0.85) and water saturation (PA 0.96, FWIoU 0.95), with moderate success for porosity (PA 0.70, FWIoU 0.55) and permeability (PA 0.74, FWIoU 0.60), and robust transferability performance (e.g., facies-to-Sw PA 0.98, FWIoU 0.97). The framework captured spatial variability and geological realism, as validated by variogram analysis, and calculated the training loss curves for the generator and discriminator for each property. Compared to traditional methods, Pix2Geomodel provides more accurate and more time- and resource-efficient property mapping. While the current model is trained to perform 2D geomodeling, future work will be developed to involve 3D geomodeling and also consider microstructural heterogeneity in the geology of the area, with extensions to multi-modal inputs planned for Pix2Geomodel v2.0. This study advances the application of generative AI in geoscience, supporting improved reservoir management and open science initiatives.

Substantial volumes of produced water are being generated from oil & gas fields worldwide. If this produced water can be treated for reuse as low salinity injection water, it becomes a game changer to promote sustainability in IOR/EOR projects. In this study, the low salinity treated produced water obtained from zero liquid discharge (ZLD) technology has been used to evaluate the potential of recycled produced water in polymer flooding, gel- and foam-based mobility control processes. Both static and dynamic tests were conducted at ambient and elevated temperatures using high salinity injection water (HSIW) and treated produced water (TPW). Rheometer was used to determine the viscosity characteristics of sulfonated polyacrylamide polymer solutions at 25oC and 75oC. Static glass bottles tests were conducted with gel solutions formulated using 3000 ppm sulfonated polyacrylamide and 150 ppm Cr (III) crosslinker at 95oC to determine the gel strength. Foam half-life times were measured to assess the foam stability. Finally, a core flood was conducted to evaluate the incremental oil recovery potential of using treated produced water in polymer flooding. The results demonstrated that the polymer concentrations are reduced by about 8-times (from 2000 ppm to 250 ppm) to achieve the same viscosity in TPW as HSIW to significantly lower the polymer consumption requirements. The gelation times of the gel in HSIW was one to 2 h, while that of the gel in TPW was one to two days. Such considerable elongation of gelation time obtained with treated produced water would favorably deliver the gel deep into reservoir to achieve more efficient conformance improvement. The foam generated using the treated produced water showed at least 10-times longer foam half-life than that produced using the high salinity injection water. The core flood results conducted using 250 ppm polymer in treated produced water showed about 18 % total incremental oil recovery after high salinity water injection. Also, there was no impact of H2S scavenging chemical derivatives found in TPW on the performance of polymer, polymer gel, and foaming surfactants used in this study. These findings clearly demonstrate the promising potential of treated produced water in different IOR/EOR processes to lower chemical concentrations and achieve better mobility control/ conformance improvement for higher oil recovery. The novelty is that this study evaluates, for the first time, the beneficial impact of using the treated produced water obtained from a ZLD field pilot "as is" in different mobility control processes involving polymer, gels, and foams. There were also not that many studies in the literature that directly evaluated the product water streams obtained from produced water desalination field pilots to determine their synergistic effects with IOR/EOR agents and improved oil recovery. The recycling of produced water evaluated for IOR/EOR processes also differs from most of the reported works in existing knowledge that studied produced water reuse mainly in fracking applications. The promising experimental results suggest that the proposed method of using low salinity treated produced water not only increases oil recovery due to synergistic effects, but also renders major environmental significance to promote circular water economy and sustainability in IOR/EOR projects due to produced water recycle/reuse.

Lower and higher diamondoids persist from early oil generation into high-maturity condensates and can fingerprint fluids when classical biomarkers are depleted. However, routine measurement of higher diamonoids has been limited by their trace abundances and poor fragmentation. A method based on pseudo-Multiple Reaction Monitoring (pMRM) on gas chromatography coupled to triple quadrupole tandem mass spectrometry (GC-MS/MS) was developed to detect and quantify diamondoids using nine diverse petroleum and rock-extract samples. In pMRM, the molecular ion was monitored in the first and third quadrupoles while a modest collision energy was applied in the collision cell to suppress interferences. Relative to selected ion monitoring, signal-to-noise for higher diamondoids increased markedly, enabling confident peak detection in the late-elution window. Optimal collision energies followed cage size, approximately 20 electron-volts for triamantane, 40 for tetramantanes, and 50 for pentamantanes and cyclohexamantane. The limit of detection (LOD) and limit of quantitation (LOQ) of the method were assessed to be 0.01 and 0.05 ng ml-1, respectively. Utilizing pMRM, many crude oils were suitable for direct injection ("dilute and shoot"), whereas rock extracts, heavy and biodegraded crudes, and some condensates benefited from a silica-gel preparation step. In those cases, approximately 60 mg of sample was sufficient to prepare a saturated, n-alkane free fraction suitable for detecting higher diamondoids using pMRM. The pMRM method demonstrated high analytical reproducibility across chemically diverse matrices including rock extracts, crude oils, and condensates, with relative standard deviations for normalized higher diamondoids typically below 10% over a three month monitoring period, confirming the stability and robustness of Quantitative Extended Diamondoid Analysis (QEDA) measurements using the pMRM approach. These results indicate that pseudo multiple reaction monitoring enables routine, high sensitivity measurement of higher diamondoids and provides molecular fingerprints that support oil-oil, oil-condensate, and condensate-source rock correlations, which were previously difficult to achieve due to the inherent compositional variability among these sample types.

This study presents the preparation of calcium carbonate (CaCO3) ceramics using vaterite derived from recycled concrete powder (RCP) through a novel in-situ polymorph transformation-enhanced cold sintering process. The resulting chemically bonded CaCO3 ceramics consists of 100 % calcite and achieve high compressive strength and a relative density of up to 80.5 %. The initial transformation from vaterite to calcite occurs at particle surfaces, decreasing porosity between particles and gradually forming a core-shell structure with a dense outer shell and a porous interior. The fusion of these shells at the contact points of adjacent particles enhances the interparticle chemical bonding. Later polymorph transformations increase pore size and volume and promote particle fusion to form a more homogeneous microstructure. This increases strength by up to 40 % compared to CaCO3 ceramics produced by conventional cold sintering. The research highlights the potential of utilizing waste concrete for sustainable and high-value CaCO3 ceramic production.

We propose a novel physics-guided deep learning framework for geophysical inversion that integrates Langevin Monte Carlo (LMC) sampling to quantify uncertainties in model parameters. A statistical sampling strategy is employed to enhance computational efficiency by reducing the number of required samples while preserving diversity and informativeness. The training data for the supervised learning networks are iteratively expanded with outputs from a stochastic sampler and their corresponding observed responses, ensuring representative coverage of the model space. The Jensen-Shannon divergence is adopted as the loss function for training the network model, in which the Gaussian assumption is applied to enable analytical computation. The developed workflow is evaluated on reservoir porosity inversion, where it successfully reconstructs porosity patterns in the subsurface, yielding results that closely match the reference model. Compared to traditional LMC algorithm applied to the entire data cube, the proposed approach attains substantial computational efficiency by leveraging an active learning strategy that identifies and utilizes a limited yet representative subset of the observations. The results demonstrate the effectiveness of the proposed method, highlighting its potential for application to a wide range of geophysical inverse problems.