Hasil untuk "Petroleum refining. Petroleum products"

Farshid Zabihian

Anh-Tuan Mai, Cam-Van Thi Nguyen, Duc-Trong Le

Multimodal emotion recognition in conversation (MERC) requires representations that effectively integrate signals from multiple modalities. These signals include modality-specific cues, information shared across modalities, and interactions that emerge only when modalities are combined. In information-theoretic terms, these correspond to \emph{unique}, \emph{redundant}, and \emph{synergistic} contributions. An ideal representation should leverage all three, yet achieving such balance remains challenging. Recent advances in contrastive learning and augmentation-based methods have made progress, but they often overlook the role of data preparation in preserving these components. In particular, applying augmentations directly to raw inputs or fused embeddings can blur the boundaries between modality-unique and cross-modal signals. To address this challenge, we propose a two-phase framework \emph{\textbf{D}ivide and \textbf{R}efine} (\textbf{DnR}). In the \textbf{Divide} phase, each modality is explicitly decomposed into uniqueness, pairwise redundancy, and synergy. In the \textbf{Refine} phase, tailored objectives enhance the informativeness of these components while maintaining their distinct roles. The refined representations are plug-and-play compatible with diverse multimodal pipelines. Extensive experiments on IEMOCAP and MELD demonstrate consistent improvements across multiple MERC backbones. These results highlight the effectiveness of explicitly dividing, refining, and recombining multimodal representations as a principled strategy for advancing emotion recognition. Our implementation is available at https://github.com/mattam301/DnR-WACV2026

Kelvin A. Sanoja-López, A. Balu, H. D. P. Carmona et al.

Abstract This contribution offers a comprehensive guide to understanding and applying heterogeneous catalysis for transforming biomass‐derived hydrocarbons and oxygenates into high‐value added products at the laboratory scale. Analysis of molecular foundations, catalyst selection criteria, experimental setups, and characterization techniques establishes a methodological basis for designing catalytic processes relevant to basic research and pre‐industrial applications. Key catalytic mechanisms, including cracking, isomerization, C─C coupling, hydrogenation, and oxidation, are addressed, emphasizing the specific roles of active sites and how their acidic, basic, metallic, or redox nature influences system activity and selectivity. The comparison between batch and continuous flow reactors underscores the importance of their suitability to control the needed operating variables, including reproducibility and extrapolating kinetic data to industrial conditions. Likewise, rigorous mass balances and advanced analytical techniques are essential for comprehensively evaluating catalytic performance. The strategic relevance of heterogeneous catalysis is demonstrated in petroleum refining, biofuel production, plastic waste valorization, and renewable fuel synthesis. In these contexts, using bifunctional catalysts, including modified mesoporous materials and hybrid systems, has enabled significant improvements in conversion, selectivity, and operational process stability. This guide is a practical starting point for researchers entering the field, supporting the development of experimental projects and in‐depth studies in heterogeneous catalysis.

WANG Jianwei, PENG Lyu, WANG Zequn, YAN Shumei, PAN Lu, LIN Lixin, WANG Rui, XU Chen, LIU Shu, HUANG Xiaojuan

The lower member of the Pinghu Formation (hereinafter referred to as the lower Pinghu member) in the well block W of the Xihu Sag is an important oil and gas-bearing system. An accurate characterization of its sedimentary evolution patterns and reservoir distribution is critical for guiding future exploration and development. Based on core, drilling, and geophysical data, this study analyzed the sedimentary microfacies, evolution processes, and dominant controlling factors of the lower Pinghu member. The results showed that the lower Pinghu member (sand groups P12~P9) could be divided into third-order sequences, mainly comprising deltaic and tidal flat deposits influenced by tidal processes. The P12 sand group, deposited during a lowstand system tract with relatively low sea level, was primarily composed of deltaic deposits, issueed by tidal deposits. During deposition of the P11 and P10 sand groups in the transgressive system tract, sediment supply weakened and delta development was curtailed. Thus, tidal flat environments became dominant in the study area. The P9 sand group, deposited during the highstand system tract, experienced increased sediment supply, tidal flat deposition reduction, and delta progradation towards the basin. Analysis of the sedimentary evolution process clarified that sediment supply, sea level fluctuations, and the paleogeomorphology controlled the microfacies migration and evolution in the well block W. Firstly, paleogeomorphology directly controlled the depositional accommodation and determined the spatial distribution of sedimentation. Secondly, abundant sediment supply and relatively lower sea level promoted deltaic development, leading to the formation of distributary channel and mouth bar sand bodies. On the contrary, the reduction of sediment supply and rising relative sea level restricted deltaic propagation while enhancing tidal power, resulting in the development of tidal flats, tidal channels, and tidal sand bars. In the study area, the relative intensity of deltaic and tidal processes was controlled by changes in relative sea level and sediment supply. During deposition of the P12 and P9 sand groups, sufficient sediment supply and relatively low sea levels favored delta development. On the contrary, during marine transgression stage corresponding to the P11-P10 sand groups, the sediment supply weakened and the relative sea levels rose. Under such conditions, deltaic deposits were vulnerable to damage, which favored the development of tidal sediments. However, the development of deltaic and tidal flat deposits in response to changes in relative sea level and sediment supply was also controlled by paleogeomorphology. During deposition of the P12 sand group, the presence of a nose-shaped paleo-uplift in the central part of the study area limited eastward progradation of the western delta. This resulted in differences in sedimentary facies types between the east and west sides of the nose-shaped paleo-uplift during deposition of the P12 sand group. The western fault trough zone was dominated by deltaic deposits, while the eastern fault step zone was dominated by tidal deposits. During deposition of the P11-P9 sand groups, the influence of the nose-shaped paleo-uplift weakened, and the sedimentary facies types in the study area were relatively uniform (P11-P10 was mainly dominated by tidal flat deposits; P9 was mainly dominated by deltaic deposits). This study offers insights into the spatiotemporal distribution characteristics of favorable reservoirs in the study area and adjacent zones. In the western fault trough zone of the P12 sand group and in the P9 sand group, deltaic sand bodies such as channels, mouth bars, and sheet sands are the favorable sand body types, and their exploration and development should be primarily guided by the deltaic depositional model. In eastern fault step zone of the P12 sand group and in the P11-P10 sand groups, the dominant sand bodies are tidal sand bars or tidal channels extending seaward and parallel to the shoreline, and their exploration and development should follow the tidal depositional model.

Seamus Somerstep, Ya'acov Ritov, Mikhail Yurochkin et al.

Standard techniques for aligning large language models (LLMs) utilize human-produced data, which could limit the capability of any aligned LLM to human level. Label refinement and weak training have emerged as promising strategies to address this superalignment problem. In this work, we adopt probabilistic assumptions commonly used to study label refinement and analyze whether refinement can be outperformed by alternative approaches, including computationally intractable oracle methods. We show that both weak training and label refinement suffer from irreducible error, leaving a performance gap between label refinement and the oracle. These results motivate future research into developing alternative methods for weak to strong generalization that synthesize the practicality of label refinement or weak training and the optimality of the oracle procedure.

Yao Shu, Qixin Zhang, Kun He et al.

Recently, zeroth-order (ZO) optimization plays an essential role in scenarios where gradient information is inaccessible or unaffordable, such as black-box systems and resource-constrained environments. While existing adaptive methods such as ZO-AdaMM have shown promise, they are fundamentally limited by their underutilization of moment information during optimization, usually resulting in underperforming convergence. To overcome these limitations, this paper introduces Refined Adaptive Zeroth-Order Optimization (R-AdaZO). Specifically, we first show the untapped variance reduction effect of first moment estimate on ZO gradient estimation, which improves the accuracy and stability of ZO updates. We then refine the second moment estimate based on these variance-reduced gradient estimates to better capture the geometry of the optimization landscape, enabling a more effective scaling of ZO updates. We present rigorous theoretical analysis to show (a) the first analysis to the variance reduction of first moment estimate in ZO optimization, (b) the improved second moment estimates with a more accurate approximation of its variance-free ideal, (c) the first variance-aware convergence framework for adaptive ZO methods, which may be of independent interest, and (d) the faster convergence of R-AdaZO than existing baselines like ZO-AdaMM. Our extensive experiments, including synthetic problems, black-box adversarial attack, and memory-efficient fine-tuning of large language models (LLMs), further verify the superior convergence of R-AdaZO, indicating that R-AdaZO offers an improved solution for real-world ZO optimization challenges.

Dandan Yang, Victoria Gould

Graph products of monoids provide a common framework for free products and direct products. Trace monoids are graph products of finitely generated free monoids. We investigate the interaction of certain finitary conditions with graph products. Specifically, we examine the conditions of being weakly left noetherian (that is, every left ideal is finitely generated) and weakly left coherent (that is, every finitely generated left ideal has a finite presentation) and the related conditions of the ascending chain condition on principal left ideals, being left ideal Howson, and being finitely left equated. All of these conditions, and others, are preserved under retract; as a consequence, if a graph product has such a property, then so do all the constituent monoids. We show that the converse is also true for all the conditions listed except that of being weakly left noetherian. In the latter case we precisely determine the graph products of monoids which are weakly left noetherian.

Congpei Qiu, Yanhao Wu, Wei Ke et al.

Contrastive Language-Image Pre-training (CLIP) excels in global alignment with language but exhibits limited sensitivity to spatial information, leading to strong performance in zero-shot classification tasks but underperformance in tasks requiring precise spatial understanding. Recent approaches have introduced Region-Language Alignment (RLA) to enhance CLIP's performance in dense multimodal tasks by aligning regional visual representations with corresponding text inputs. However, we find that CLIP ViTs fine-tuned with RLA suffer from notable loss in spatial awareness, which is crucial for dense prediction tasks. To address this, we propose the Spatial Correlation Distillation (SCD) framework, which preserves CLIP's inherent spatial structure and mitigates the above degradation. To further enhance spatial correlations, we introduce a lightweight Refiner that extracts refined correlations directly from CLIP before feeding them into SCD, based on an intriguing finding that CLIP naturally captures high-quality dense features. Together, these components form a robust distillation framework that enables CLIP ViTs to integrate both visual-language and visual-centric improvements, achieving state-of-the-art results across various open-vocabulary dense prediction benchmarks.

D. de la Rosa-Urbalejo, Brandon Reyes-Ramírez, J. M. Riesco-Avila et al.

Pyrolysis of plastic waste generates alternative fuels that are useful for thermal devices. However, it is the subsequent fractionation process that truly distinguishes itself, refining these fuels to a quality comparable to petroleum derivatives such as diesel and gasoline, thereby making them suitable for direct use in internal combustion engines. This study evaluated the effect on emissions of an internal combustion engine fueled with waste plastic oil diesel-like fuel and its blend with biodiesel. The tests were conducted on-road using a typical light-duty diesel vehicle for passenger or courier transportation. Three tests were carried out, the first one with petroleum diesel (D100) as the reference fuel, the second one with waste plastic oil diesel-like fuel (WPO-DL), and the last with a blend of 80% WPO-DL and 20% biodiesel obtained from waste cooking oil (WPO-DL80WCO20). The main findings were a reduction in the generation of NOx when using WPO-DL. However, using WPO-DL increased the production of CO and HC, attributed to the presence of more heavy hydrocarbons compared to that of D100. Additionally, for the WPO-DL80WCO20 blend under acceleration conditions HC decreased under all acceleration conditions. These findings demonstrate that the distilled fuel obtained from the pyrolytic oil product of non-recyclable plastic waste by conventional methods would contribute to the global energy transition and reduce the presence of tons of non-recyclable plastic waste by traditional methods.

Haoran Yu, Changmin Liu, Shuo Liu et al.

Spent petroleum refining catalyst is regarded as the important secondary resource for valuable metals. However, common recycling strategies, including soda roasting, acid and alkaline solutions leaching and chemically precipitation, produced large quantities of high salinity wastewater. This study proposed an efficient method to recovery of Mo and Ni from the spent hydrodesulfurization (HDS) catalyst via O2-rich roasting and organic acid leaching with the advantage of less salinity wastewater production. The transformation of Mo(IV) into soluble Mo(VI) was enhanced by O2-rich atmosphere roasting, and 98.64% of Mo(IV) was oxidized at 650 ℃ for 2 h in atmosphere containing 30% of O2. The oxidation process of Mo(IV) was agreed with the shrinkage pore model, and regulated by surface reaction and internal diffusion. 97.97% of Mo(VI) was leached from roasted product by oxalic acid, separated with complexation extraction agent of Ala-TBP and recovered as (NH4)8Mo10O34 and (NH4)2Mo3O10 by evaporative crystallization. Ni was leached out from spent catalyst with 1 mol/L acetic acid, and precipitated as NiC2O4 with oxalic acid. 95.92% of Mo and 96.77% of Ni were recovered from spent HDS catalyst with this recycling route. This study provided a high-efficient and eco-friendly method to recovery of valuable metals from spent catalyst.

S. Kumari, Sonali Sengupta

Abstract The petroleum fraction of crude oil is laden with notorious nitrogen compounds, whose removal facilitates fuel oil transport and enhances the longevity of hydrodesulfurization catalysts. Petroleum refining operation thus demands the adoption of unconventional techniques, such as coupled oxidative–extractive denitrogenation. Kinetic models can provide insights into the reaction mechanism and the design and performance of reactors. In this work, nine different kinetic models were developed by applying the Langmuir–Hinshelwood and Eley–Rideal approaches to various reaction pathways. The screening of the kinetic models was done based on the values of the rate law parameters and corresponding R2 and adjusted R2 values calculated through a multi-parametric, non-linear regression based on the Levenberg–Marquardt algorithm. Out of all the models, the Eley–Rideal model corresponding to the rate expression rs9 was found to be the best fit, reflecting that hydrogen peroxide remained in bulk while pyridine alone was adsorbed on the catalyst. The selection of suitable extractants for the extractive removal of oxidation products was made based on their extraction efficiency and partition coefficients. The ternary liquid–liquid equilibrium (LLE) data for the (acidulated water-pyridine-isooctane) and (acidulated water-pyridine N-oxide-isooctane) systems were plotted, and Othmer–Tobias, Hand, and Bachman correlations explained the consistency of the experimental LLE data.

Arno R. De Klerk, G. Chauhan, C. Halmenschlager et al.

Fully formulated synthetic jet fuel is an aviation turbine fuel that does not contain petroleum‐derived kerosene and comprises the hydrocarbon compound classes n‐alkanes, isoalkanes, cycloalkanes, and aromatics. When the aim is to produce sustainable aviation fuel, one potential process pathway is by indirect liquefaction via Fischer–Tropsch synthesis. Fischer–Tropsch synthesised paraffinic kerosene plus aromatics (FT SPK/A) is a product that is fully formulated and can in principle be qualified as Jet A‐1. The synthetic jet fuel must ultimately meet all of the Jet A‐1 specifications. However, there are still hurdles on the path toward global approval of fully formulated synthetic jet fuel. In this study, several different refining pathways are shown that can be employed to produce FT SPK/A. The refining pathways have the desirable attribute of being generally useful and not limited to a specific refining technology. A case study is also presented in which FT SPK/A was produced, characterised and compared to Jet A‐1 specification requirements. It illustrated that it was practical to produce a fully formulated jet fuel via Fischer–Tropsch refining.

Mahsa PARHIZGAR KERADEH, Seyyed Alireza TABATABAEI-NEZHAD

This study used the diethylene triamine pentaacetic acid (DTPA)-seawater (SW) system to modify the sandstone rock wettability and enhance oil recovery. The investigation involved conducting wettability measurement, Zeta potential measurement, and spontaneous imbibition experiment. The introduction of 5% DTPA-SW solution resulted in a significant decrease in the rock-oil contact angle from 143° to 23°, along with a reduction in the Zeta potential from −2.29 mV to −13.06 mV, thereby altering the rock surface charge and shifting its wettability from an oil-wet state to a strongly water-wet state. The presence or absence of potential determining ions (Ca2+, Mg2+, SO42−) in the solution did not impact the effectiveness of DTPA in changing the rock wettability. However, by tripling the concentration of these ions in the solution, the performance of 5% DTPA-SW solution in changing wettability was impaired. Additionally, spontaneous imbibition tests demonstrated that the 5% DTPA-SW solution led to an increase in oil recovery up to 39.6%. Thus, the optimum mass fraction of DTPA for changing sandstone wettability was determined to be 5%.

Xiaohua Ke, Jinsheng Sun, Xiongwen Yang et al.

Abstract The Weiyuan shale gas field faces problems of long drilling cycles and high development costs. Improving the drilling efficiency of polycrystalline diamond compact bits in shale formations will significantly reduce the overall well cost and duration. Previous applications have demonstrated that conventional PDC bits on the market cannot meet the demand for drilling acceleration. In this work, a new three-dimensional concave-shaped PDC cutter was proposed to improve drilling efficiency. The special 3D concave-shaped cutter has two symmetrical curved ridges on the concave surface and a circular plane at the center. The cutting mechanism of the new 3D concave-shaped cutter has been studied by laboratory experiments and numerical simulations. The research data revealed that, compared with a flat cutter, the tangential force of the original 3D concave-shaped cutter was reduced by 1.4%–35.0%, the axial force was reduced by 6.7%–37.6%, and the mechanical specific energy (MSE) was reduced by 1.6%–35.59%. Simulations showed that the shear action of the 3D concave-shaped cutter was divided into two continuous parts, with the sides and the center surface being stressed successively, which is helpful for extending shear cracks, forming trilobal cuttings, and improving cutting efficiency. With the special 3D concave-shaped cutter, an 8½-inch drill bit was designed and manufactured and tested on the Longmaxi shale in the Weiyuan block. Through field tests, we further compared the performance of the 3D concave-shaped cutter PDC bit with that of the flat cutter PDC bit. The 3D concave-shaped PDC bit had a 41.8% better footage and 22.6% better rate of penetration (ROP) in field test.

Liao Shengjun, Yu Zenghui, Hou Hongwei et al.

Due to the restriction of thrust and space of the test bench,it is often difficult to carry out dynamic test on large drill pipes,so the combination of numerical simulation and scaled test piece test is a good solution to obtain the dynamic characteristics of drill pipes.In order to improve the dynamic performance parameter accuracy of the scaled drill pipe,a large drill pipe was taken as an example to study the modal parameters of the drill pipe,discuss the relationship between the finite element unit division scale and the convergence of the results,and determine the appropriate finite element unit division scale that meets the convergence requirements.The study results show that with comprehensive consideration to the calculation accuracy,computing efficiency and processability of scaled drill pipe,the optimal scaling ratio of the drill pipe is determined to be 25%.The comparison of modal frequency of 25% scaled drill pipe test and numerical simulation shows that the relative error is less than 3.8%.The conclusions provide theoretical and technical reference for the division of drill pipe numerical simulation units and the design of large drill pipe scaled test piece.

Liu Jun, Wang Pucheng, Guo Xiaoqiang

Aiming at the problem of vibration failure of deepwater riser under the condition of soft suspension emergency evacuation, combined with the Hamiltonian variational and virtual work principle, the finite element method was used to build a 3D nonlinear dynamic model of soft suspension drilling riser; meanwhile, the nonlinear factors such as longitudinal and transverse coupling effect, ocean vortex-excited effect and large deformation of riser were considered. With the help of test parameters and results presented in the literatures, the effectiveness of the nonlinear dynamic model was verified. The field parameters of riser in HYSY981 drilling platform were used to explore the influence of buoyancy modules layout and typhoon return period on the nonlinear vibration response characteristics of soft suspension riser under emergency evacuation conditions. The research results show that the reasonable arrangement of buoyancy modules can reduce the stress and dynamic axial force, and the optimal arrangement method of buoyancy modules can be determined by the built theoretical model; in the high flow rate area, the buoyancy modules layout of the riser should be reduced, the maximum stress and displacement of the riser occur near the top or bottom area, and the safety of the riser at this position should be paid attention to; the currentwise displacement of the riser increases with the increase of the typhoon return period, and the displacement near the upper part changes greatly, so in the evacuation process, the collision between the upper part of the riser and the moon pool should be avoided. The research results provide reference for understanding the vibration response characteristics of soft suspension riser under evacuation conditions.

YE Shen, QIAO Jiangmei, LI Tongchun

The cave is the main reserve space in fracture-cavity carbonate reservoir. The basic method of fracture-cavity reservoirs exploitation is to generate hydraulic fractures connecting the boreholes and caves through hydraulic fracturing. Based on the TOUGH-AiFrac coupling solver, the influence of different water injection pressures and internal pressure of caves on hydraulic fracture propagation is studied respectively. The results show that with the increase of water injection pressure, the influence of in-situ stress on hydraulic fractures is gradually weakened, and the fractures tend to initiate and propagate along the initial direction. When the water injection pressure is 1.6 times greater than the horizontal maximum principal stress, the rangeability of the fracture initiation directions weaken obviously, and the fracture direction tends to be stable. When the water injection pressure is 2.4 times greater than the horizontal maximum principal stress, the fracture propagation path tends to be stable. The greater internal pressure of the caves, the stronger the cave’s effect on hydraulic fractures. The caves with internal pressure reaching two times of the maximum principal stress exert a kind of “attraction” effect on the hydraulic fractures, which will be gradually strengthened with the increase of cave’s internal pressure while be weakened with the increase of injection pressure. The research results can be used as a guideline to optimize the fracturing work parameters and thereby enhance the oil/gas production rate according to the geological conditions.

Lin Lijun, He Mingge, Wang Qingyuan et al.

In view of the shortcomings of traditional downhole throttling production adjustment mode(replacing downhole flow nipple by wireline operation)such as long operation cycle, high cost and high risk, with the realization of intelligent natural gas production control by remote wireless control downhole production adjustment as the objective, the mechanical engineering, microcomputer, automatic control and wireless communication technologies were comprehensively applied to study the key technologies such as the design of mechanical system structure of CAD/CAE based wireless control downhole choke, the design of gas pressure wave based coding and decoding method and wellhead pressure wave based coding technology, and the design of control algorithm of wireless control downhole choke. In addition, based on the remote wireless control downhole intelligent throttling method and the control principle, a wireless control downhole choke was developed and applied in the field. The application results show that the overall performance of the wireless control downhole choke is stable, and the pressure wave wireless control technology is reliable. It has been successfully applied in the downhole wireless intelligent production adjustment in gas wells, and has the characteristics of simple operation, low cost and low risk. This technology provides a new solution for the automatic control of wellbore in smart oil and gas fields, and the proposed methods and ideas also have certain reference significance for the development of other downhole intelligent tools.