Petrobras (Brazil)

Early-stage offshore field development requires screening of technical alternatives, robust scenario evaluation, and seamless interaction between multidisciplinary domains such as reservoir engineering, subsea architecture design, and economic assessment. Despite significant advancements in digital engineering, workflows connecting these disciplines often remain fragmented, relying heavily on manual data transfer, spreadsheet consolidation, and expert-driven interpretation. This fragmentation introduces delays, increases the risk of inconsistency, and limits the number of development pathways that can realistically be explored within Front-End-Loading stages. The increasing complexity of offshore reservoirs, combined with tighter economic margins and the need for transparent, auditable decision-making, highlights the necessity for integrated, automated workflows. By bridging simulation environments and conceptual engineering tools through programmable interfaces, it becomes possible to unify data flows, enforce engineering logic, and reduce turnaround time without compromising technical consistency. This paper presents a fully integrated framework that bridges the gap between numerical reservoir simulation and automated subsea concept generation. By leveraging a sensitivity cube approach derived from physics-based simulations, the proposed workflow facilitates the rapid synthesis of realistic field development architectures. This integration allows for the direct coupling of dynamic production profiles with rigorous economic assessment, enabling the calculation of Capital Expenditure (CAPEX), Operational Expenditure (OPEX), and Net Present Value (NPV) across a vast design space. Unlike traditional fragmented workflows, this system supports holistic iterative optimization between subsurface and facilities domains. The result is a substantial enhancement in engineering consistency, process autonomy, and evaluation speed, critical attributes for robust decision-making during early project framing and Front-End Loading (FEL) phases.

Giant saline aquifers (defined here as aquifers that cover areas larger than 10.000 km²) are promising candidates to scale up geological CO2 storage. However, they present significant simulation challenges due to their vast extent, heterogeneity, and limited subsurface data. This study introduces a reliable multiscale modeling framework which is designed for these fields. The method is also applied to assess CO2 storage in the Ponta Aguda saline aquifer (Santos Basin, Brazil, 40000 km2 area) to demonstrate its applicability in real field environments.Our multiscale strategy is formulated such that it delivers reliable quantification of the trapped and mobile mass of CO2, i.e., the plume migration under complex hysteretic transport physics. Of particular interest is to preserve reliable quantification of the plume dynamics from near wellbore region (in the order of 10m horizontal resolution) all the way to the far field zones (with 1000m horizontal resolution).Consistency checks are applied to make sure that the results from different scales are representative of the same realization and storage conditions. Our novel multiscale strategy benefits from the local saturation and global pressure physics. More precisely, the best global pressure representation is provided on the largest scale and therefore is used to provide local boundary conditions (using methods such as Fetkovich model) to the higher resolutions (smaller sub-domains). On the other hand, saturation distribution is first resolved from the smallest sub-domains (highest resolution) and upscaled to the large-scale domains. Through these analyses, it is found that classical upscaling approaches systematically overestimate the trapped amount of CO2 on coarser models. This motivates the development of advanced reliable multiscale strategies which are efficient but also accurate while the system is being represented on coarser-resolution grids.We present three different methods and compare them based on their accuracy of trapped amount of CO2 in the field-scale model. These are namely: Local Grid Refinement (LGR), Effective Values (EV), and Algebraic Dynamic Multilevel (ADM). The results indicate that ADM is the most stable and robust approach among all the approaches considered for real-field applications. Especially, LGR and EV are found limited in their scopes since they depend on a matching procedure (against a reference solution) for their upscaled parameters, before any new simulations. As a result, their tuned parameters cannot be transferred from one model to another. ADM, on the other hand, does not require any upscaling procedure, as the multiscale basis functions allow for consistent mapping across resolutions. The results show the importance of scale-consistent modeling approaches for accurate CO2 storage assessment and highlight the risks of relying on overly simplified coarse models in the design and optimization of carbon storage projects in giant saline aquifers.

Accurate prediction of temperature-dependent liquid viscosity is essential for process design and molecular screening but remains challenging for novel hydrocarbons. Purely data-driven graph neural networks (GNNs) can achieve good in-range accuracy yet often lack thermodynamic consistency and become unstable under temperature shifts or sparse calibration data. This work introduces ExPUFFIN (Extended Path-Unifying Feed-Forward Interfaced Network), a hybrid GNN framework that embeds established viscosity–temperature correlations directly into the output layer by learning molecule-specific parameters of Andrade-type and empirical formulations. The approach is evaluated on a curated dataset of 305 pure hydrocarbons using standard test metrics, interpolation analysis, Leave-One-Temperature-Out evaluation, and extrapolation beyond the experimental range. While the baseline model attains comparable test errors, it exhibits oscillatory and inconsistent behaviour outside the training domain. In contrast, the ExPUFFIN variants preserve smooth, monotonic viscosity–temperature trends and substantially reduce extrapolation errors. The results show that incorporating inductive bias into the network output improves robustness and reliability without compromising predictive accuracy. • ExPUFFIN integrates GNNs with inductive viscosity–temperature correlations to improve prediction reliability. • The hybrid architecture yields smoother, more consistent viscosity–temperature curves. • ExPUFFIN enhances interpolation stability and extrapolation performance compared to a purely data-driven baseline.

Ongoing research on adsorbent deactivation in dryingprocesseshas unveiled crucial insights into the loss of performance with prolongeduse. Our study builds on previous work by examining the effects ofhigh temperatures, hydrocarbon exposure, and cycle duration on chabaziteadsorbents with a Si/Al ratio of approximately 2 and in the sodiumform. By aging the material under controlled conditions, we assessedthe impact on its structural and textural properties. Our findingsreveal that maintaining high temperatures, coupled with exposure to n-heptane vapor, induces a mild degradation of the crystallinestructure, with more pronounced effects during longer aging periods.Notably, hydrocarbons play a critical role in adsorbent deactivation,adsorbing in both the inner pores and the outer surface of the zeolite.This leads to a deterioration of the textural characteristics, whichdirectly correlates with an increase in the carbon content of thebulk material. Additionally, samples exposed to n-heptane vapor exhibited a more homogeneous composition comparedto those subjected to manual liquid n-heptane addition.Overall, the degree of degradation varies among aged samples, indicatingthe need for tailored applications based on specific aging conditions.

Biodiversity monitoring programs need to deliver accurate, timely, and actionable predictions. To establish a predictive monitoring program for deep-sea benthos of the Santos Basin, Brazil, we developed a two-stage structured model that allowed comparison of biodiversity predictions obtained from environmental simulations (2M-Sim). We also modeled the environmental variables as a function of spatial and temporal variables and compared this model's predictions with predictions obtained from real environmental data (2M). We built unstructured models (1M) as references to evaluate whether the proposed structured approach was reliable. We expected no significant differences between 1M and 2M or between 2M and 2M-Sim. Data were obtained from 100 stations at depths of 25-2400m during two surveys (2019 and 2021). In our model framework, we used 12 benthic macro- and meiofaunal variables, 44 sediment and water column environmental variables, and four spatial and temporal variables. We applied a random forest algorithm to the structured and unstructured models. All comparisons were performed with 20% of the dataset set aside for validation. The average accuracy was 72%, 69%, and 68% for the 1M, 2M, and 2M-Sim models, respectively. Accuracies of 2M ranged from 38% to 84% and were generally higher for macrofauna. The observed accuracy loss from 1M to 2M (3%) and from 2M to 2M-Sim (1%) was not significant for any biodiversity variable. The 2M model identified 30 significant environmental variables; bottom water parameters and sedimentary phytopigment and carbonate concentrations were the best predictors. Our approach supports biodiversity conservation by optimizing data needs and future sampling and by guiding data-driven management decisions for benthic biodiversity.

This paper presents a hybrid approach to predict theevolutionof technological maturity of R&D projects, using the context ofthe oil and gas (O&G) sector as an example. Integrating SystemDynamics (SD) and Agent-based Modeling (ABM) enables the proposedmultilevel framework to capture uncertainties inherent to R&Dprojects, including work effort, team size, and project duration,all of which influence technological progress. Although AB–SDhybrid models are well established in other fields, their applicationin R&D contexts remains limited. The AB–SD model combinessystem-level feedback structures governing work phases, rework cycles,and project duration with the explicit representation of decentralizedagents (e.g., team members, tasks, and controllers) whose interactionsdrive emergent project dynamics. A base-case scenario was developedto analyze the structural dynamics of early-stage innovation projects,simulating 15 parallel tasks over 156 weeks. In a comparative scenariowith sequential task execution, the model showed an 88% reductionin rework duration relative to the base case. The second scenarioevaluated mixed parallel–sequential task structures under varyingteam sizes. In parallel configuration, simulation results indicatedthat increasing team size reduced overall project duration and improvedtask completion rates, with optimal performance achieved for teamsof four to five members. These outcomes are consistent with empiricalobservations in R&D project management, where moderate team expansionenhances coordination efficiency without incurring communication overhead.However, as widely recognized in empirical studies, a substantialincrease in team size does not necessarily translate into higher completionrates, as excessive team growth often introduces communication complexityand management delays. Overall, the model outputs and the proposedmodeling framework are well aligned with expert understanding in thefield, confirming their validity as a quantitative tool for analyzingresource allocation, task scheduling efficiency, and technology maturityprogression.

Waxy oil gels exhibit complex rheological behavior, sharing characteristics with colloidal, polymeric, and fat gels. The phenomena of viscoplasticity, viscoelasticity, and thixotropy are manifested due to the dynamics of the fractal microstructure under cooling and shear conditions. Previous studies have developed models to probe these properties, including viscosity, storage modulus, yield stress, critical elastic strain and stress, and structure parameters. The models incorporate fractal characteristics within the theoretical framework. However, a unified approach to fractal rheological models for waxy oils is still lacking. This work aims to comprehensively review three types of fractal models for cluster suspensions and gels: viscosity models, elastoplastic models, and structure parameter models. The essential features of waxy oil rheology and the determination of yield stress are first outlined. Subsequently, models that describe the elastoplastic properties, such as storage modulus and yield stress, are described. Then, the viscosity models for fractal cluster suspensions are revisited, highlighting the occurrence of high correlation and multiple local minima in the regression procedure. Finally, the fractal-based structure parameters that connect the elastoplastic and viscosity models are examined. The review offers a comprehensive understanding of the complex rheological behavior of waxy oils, laying a foundation for future research in this field.

Abstract The formation of dense gases is driven by high pressures, critical temperatures, and high content of CO₂, conditions characteristic of ultra-deepwater production environments such as the Brazilian pre-salt. This study presents an experimental investigation of horizontal and slightly upward inclined dense-gas and liquid two-phase flow under hydrodynamic conditions representative of these scenarios. Due to the lack of experimental data in such regimes, the accuracy of existing predictive models remains limited, potentially leading to economic losses and environmental or safety risks. Experiments were conducted using a high-pressure inclinable loop equipped with a 30 m pipeline, employing pressurized sulfur hexafluoride (SF₆) as the gas phase and mineral oil as the liquid phase. The test section was configured at 0°, 5°, and 10°, and 125 experimental points were collected, including pressure, temperature, flow, holdup, and flow pattern data. Two-phase flow visualization was carried out using a high-speed camera, while liquid holdup and phase fraction distribution were determined with a collimated gamma-ray densitometer. The observed flow patterns included stratified smooth, stratified wavy, stratified wavy with mixing at the interface, slug, pseudo-slug, dispersed, and the rarely reported dual-continuous pattern. Notably, the dual-continuous flow pattern was identified for the first time under upward inclined conditions (5° and 10°). The experimental results demonstrate that pipe inclination and dense gas velocity are key factors in the transition between slug and pseudo-slug flow patterns. Additionally, interfacial instabilities and liquid splashing were observed at high dense-gas velocities. These findings address important knowledge gaps and support the development of more accurate predictive models for multiphase flow in complex production scenarios.