BR 
EN Petroleum engineers usually create hundreds of models of a reservoir to deal with its uncertainties. Since running flow simulations in all models is time-expensive, selecting a subset of “representative models” (or RMs) for simulations can reduce total simulation time without compromising analysis quality. However, judging the representativeness of the RMs and choosing the best model are hard tasks that visualization techniques can help to improve. This paper explores visualization techniques to aid engineers in evaluating and comparing the representativeness of a “solution” (i.e., a set of RMs) and of multiple solutions. We propose an interactive dashboard featuring: (a) a representativeness heatmap of multiple solutions, and (b) a set of crossplots of production variables of a solution, enhanced with convex hulls of representative and represented models. Experienced petroleum analysts evaluated the proposed visualizations positively, indicating the potential of these visualizations to enhance the process of choosing representative models.
Mês: outubro 2024
The influence of major faults and fractures on the development of non-matrix porosity system in a pre-salt carbonate reservoir, Santos Basin – Brazil
Faults and fractures are central for characterizing the permeability distribution in carbonate reservoirs since they act as pathways for diagenetic fluids that often favor intense rock dissolution and permeability. Usually, high permeability zones and fractures are not easily recognized in seismic data due to limited resolution and they are often associated with higher concentrations of hydrocarbons or even significant fluid losses during drilling, thus creating a challenge for hydrocarbon exploration. In the Santos Basin, southeast Brazil, the pre-salt carbonate reservoirs from the Barra Velha Formation (BVE) are the main hydrocarbon producers in Brazilian Atlantic margin and well-known for being extremely heterogeneous, exhibiting complex dual-porosity systems. In this study, we built a conceptual model of these fracture zones and non-matrix porosity formation that helped narrowing the understanding of these complex systems. The characterization of faults and fractures was carried out using seismic, well-logs, and borehole image data to understand the influence of these structures in the porosity formation along the Barra Velha Formation. In the study area, three fault sets were defined (F1, F2, and F3) from seismic data. F1 represents to the larger faults, while the F3 fault set represents the smaller faults related to the reactivation of F1; both sets being oriented NE-SW. The F2 fault set is associated with the rift formation and is oriented to NNE-SSW. These three fault sets compartmentalized the studied area into different domains, each exhibiting distinct fracture sets. The natural open fractures were formed during the reactivation of rift faults and are oriented mainly NW, NNE-NNW, NE, and ENE and were identified across the entire study area, but with different intensity values. The fracture intensity closely relates to the distance from major faults where the wells with the highest fracture intensity occurs located 150–590 m from the larger F1 fault set. Scan-lines were conducted throughout the area to determine the fault width, and an average value of 1.2 km was established. Seven non-matrix porosity classes were characterized and classified into stratigraphically concordant and discordant non-matrix pore types at well scale through borehole image interpretation. The Barra Velha Formation exhibit higher occurrence of stratigraphically discordant non-matrix porosity related to fractured zones while stratigraphically concordant non-matrix porosity is mainly controlled by the paleotopography of the study area. Overall, non-matrix porosity formation tends to follow an orientation that suggests a preferential dissolution flow towards NE and ENE directions. Intervals with higher silica content shows a positive correlation with both fracture intensity and stratigraphically discordant non-matrix porosities. This work provides a conceptual model about the fractures and non-matrix porosity distribution in pre-salt carbonate rocks that can help address some of associated structural and stratigraphic uncertainties during field appraisal and development.
Learning characteristic parameters and dynamics of centrifugal pumps under multiphase flow using physics-informed neural networks
Electrical submersible pumps (ESPs) are prevalently utilized as artificial lift systems in the oil and gas industry. These pumps frequently encounter multiphase flows comprising a complex mixture of hydrocarbons, water, and sediments. Such mixtures lead to the formation of emulsions, characterized by an effective viscosity distinct from that of the individual phases. Traditional multiphase flow meters, employed to assess these conditions, are burdened by high operational costs and susceptibility to degradation. To this end, this study introduces a physics-informed neural network (PINN) model designed to indirectly estimate the fluid properties, dynamic states, and crucial parameters of an ESP system. A comprehensive structural and practical identifiability analysis was performed to delineate the subset of parameters that can be reliably estimated through the use of intake and discharge pressure measurements from the pump. The efficacy of the PINN model was validated by estimating the unknown states and parameters using these pressure measurements as input data. Furthermore, the performance of the PINN model was benchmarked against the particle filter method utilizing both simulated and experimental data across varying water content scenarios. The comparative analysis suggests that the PINN model holds significant potential as a viable alternative to conventional multiphase flow meters, offering a promising avenue for enhancing operational efficiency and reducing costs in ESP applications.
Investigação da Compatibilidade entre Fluidos da Linha de Injeção Química de Desemulsificante em Poço de Petróleo
The chemical injection (CI) lines in oil-producing wells have been plagued by issues related to clogging, hindering the injection of demulsifiers into the oil reservoir, which directly impacts oil productivity. This problem might require stoppages in the oil production for cleaning and/or pigging the line, and even replace downhole equipment (chemical injection valves). Given this scenario, the present study aims to investigate the compatibility of fluids present in the chemical injection line. If these chemicals are not compatible and result in solid formation, it could lead to pipe blockages. To assess this, monoethylene glycol (MEG), commonly used as a flushing fluid and hydrate inhibitor, and an ethoxylated polymeric surfactant, used as a demulsifier, were mixed to observe any physical change in the solution. The tests were conducted at room temperature (30 °C) and 60 °C, for up to 72 h, with visual monitoring during this period. In addition, rheological tests were carried out on pure fluids and their mixtures to evaluate if there were viscoelastic changes. These studies made it possible to detect a whitish gel at the bottom of the test tube formed through contact between the MEG and the demulsifier (phase separation) at both temperatures. This allows us to conclude that physical changes occurred in the mixture (MEG + demulsifier), forming a higher viscosity gel. Importantly, preventing this gel formation could possibly prevent clogging in the CI lines, as the gel could adhere to solid contaminants and contribute to blockages.
Better stopping through cross validation in an iterative ensemble smoother: A perspective from supervised machine learning
Iterative ensemble smoothers (IES) are among the popular reservoir data assimilation (RDA) algorithms for reservoir characterization. The actual deployment of an IES algorithm requires implementing certain stopping criteria, normally adopted for runtime control (e.g., by stopping the IES when it reaches the maximum number of iterations) and/or safeguarding the RDA performance (e.g., by preventing the simulated data from overfitting the actual observations). In practice, for various reasons, it is often challenging for existing stopping criteria to simultaneously achieve both purposes. One noticeable issue, as illustrated in this work, is that in many situations, the qualities of the estimated reservoir models may already start to deteriorate before a conventional stopping criterion activates to terminate the iteration process. Following this observation, one practically important question arises: Is it possible to further improve the efficacy of the IES algorithm by designing a different stopping criterion so that the IES can stop earlier, saving computational costs while achieving better RDA performance?
As one of the rare attempts in the community, this work aims to investigate the use of a new IES stopping criterion that has the potential to provide an affirmative answer to the above question. In this regard, our main idea is based on the concept of cross validation (CV), routinely adopted in supervised machine learning (SML) problems for early stopping to prevent SML models from overfitting the training data. Despite the noticed similarities between RDA and SML problems, some fundamental differences exist, making it fail to work well if one directly extends a vanilla CV procedure from SML to RDA. To tackle this identified challenge, we design an efficient CV procedure tailored for RDA problems, and inspect the performance of an IES algorithm equipped with this CV procedure (IES-CV) in both synthetic and real field case studies. Our numerical investigation indicates that the IES-CV algorithm achieves promising RDA performance in all case studies, confirming the possibility that with the aid of a proper stopping criterion, an IES algorithm can terminate at an appropriate iteration step with near-optimal RDA performance. Beyond these numerical findings, it is also our hope that the current work may help improve the best practices of applying IES to RDA problems, taking advantage of the effective, CV-based stopping criterion.