Hasil untuk "Petroleum refining. Petroleum products"

J. Speight

Wang Xiaoming, Song Chenglin, Sun Wei et al.

Quantitative assessment was carried out on the sticking risk in the drilling and completion stage to fill the gap in this field， and provide scientific basis and technical support for the risk control of sticking in drilling operations. Based on the analysis of causes， inducing factors and common types of sticking， a fault tree model was constructed and mapped to a Bayesian network model. Using expert knowledge and real data from an oil field in Southwest Sichuan Basin， fuzzy analytic hierarchy process （FAHP） was used to quantify the risk of sticking， and the main risk factors affecting sticking were identified by sensitivity analysis. The research results show that human error is the main cause of sticking， with its contribution rate significantly exceeding other factors. Sensitivity analysis identifies key causal factors such as improper drilling parameters， inadequate connections and long rig downtime. The constructed model can dynamically update the risk probability， thereby effectively supporting risk warning and decision-making. The research conclusions not only provide theoretical support for the oil and gas industry， but also provide practical guidance for risk management in drilling operations， and have important engineering application value.

YE HONGYING, CAO CHENG, ZHAO YULONG et al.

Carbon capture, utilization and storage (CCUS) is a key technology for achieving carbon neutrality, providing the dual benefits of enhanced energy production and reduced CO₂ emissions through CO₂-enhanced oil and gas recovery (EOR/EGR) and geological storage. However, the large-scale application of CCUS technology faces technical challenges such as engineering design and risk assessment. Traditional approaches, which rely on empirical formulas, experimental verification, and physical models, suffer from low computational efficiency, limited model accuracy, and difficulties in handling multi-dimensional coupling when addressing complex systems. Machine learning (ML), with its powerful data-driven analytical capabilities and adaptive optimization features, can establish high-precision prediction models, optimize operating parameters, predict reservoir fluid behavior, and assess leakage risks. This enables real-time monitoring and intelligent decision-making for complex systems, enhancing the safety and economic efficiency of CCUS technology. This study systematically reviews the applications of ML in CO₂-enhanced oil and gas recovery and geological storage. In terms of CO₂-enhanced oil and gas recovery, the applications cover percolation mechanism modeling, well pattern design optimization, production prediction and evaluation, multi-objective optimization, minimum miscibility pressure prediction, gas adsorption curve prediction, and CO₂-CH₄ diffusion assessment. For CO₂ geological storage, the applications include reservoir selection, research on CO₂ dissolution and diffusion mechanisms, geological storage performance prediction, and risk assessment. ML demonstrates significant advantages in improving prediction accuracy, optimizing operating parameters, and enhancing computational efficiency. It has made important progress in key fields such as reservoir selection, gas adsorption prediction, and storage performance prediction. However, challenges remain in terms of adaptability to complex geological scenarios, model universality, dynamic data processing capabilities, and physical interpretability.

Sander Tonkens, Sosuke Kojima, Chenhao Liu et al.

Control Barrier Functions (CBFs) are a powerful tool for ensuring robotic safety, but designing or learning valid CBFs for complex systems is a significant challenge. While Hamilton-Jacobi Reachability provides a formal method for synthesizing safe value functions, it scales poorly and is typically performed offline, limiting its applicability in dynamic environments. This paper bridges the gap between offline synthesis and online adaptation. We introduce refineCBF for refining an approximate CBF - whether analytically derived, learned, or even unsafe - via warm-started HJ reachability. We then present its computationally efficient successor, HJ-Patch, which accelerates this process through localized updates. Both methods guarantee the recovery of a safe value function and can ensure monotonic safety improvements during adaptation. Our experiments validate our framework's primary contribution: in-the-loop, real-time adaptation, in simulation (with detailed value function analysis) and on physical hardware. Our experiments on ground vehicles and quadcopters show that our framework can successfully adapt to sudden environmental changes, such as new obstacles and unmodeled wind disturbances, providing a practical path toward deploying formally guaranteed safety in real-world settings.

Jiwon Kim, Violeta J. Rodriguez, Dong Whi Yoo et al.

Large language models (LLMs) are increasingly used for mental health support, yet they can produce responses that are overly directive, inconsistent, or clinically misaligned, particularly in sensitive or high-risk contexts. Existing approaches to mitigating these risks largely rely on implicit alignment through training or prompting, offering limited transparency and runtime accountability. We introduce PAIR-SAFE, a paired-agent framework for auditing and refining AI-generated mental health support that integrates a Responder agent with a supervisory Judge agent grounded in the clinically validated Motivational Interviewing Treatment Integrity (MITI-4) framework. The Judgeaudits each response and provides structuredALLOW or REVISE decisions that guide runtime response refinement. We simulate counseling interactions using a support-seeker simulator derived from human-annotated motivational interviewing data. We find that Judge-supervised interactions show significant improvements in key MITI dimensions, including Partnership, Seek Collaboration, and overall Relational quality. Our quantitative findings are supported by qualitative expert evaluation, which further highlights the nuances of runtime supervision. Together, our results reveal that such pairedagent approach can provide clinically grounded auditing and refinement for AI-assisted conversational mental health support.

JIANG Beibei, LIU Jiabo, ZHANG Guoqiang, WANG Dong, LI Ying, LUO Hongwen, ZHOU Lang

In the context of the global carbon neutrality strategy, greenhouse gas emissions —mainly CO₂— are continuously rising, exerting adverse effects on the global climate, ecosystems, and human life. Geological storage of CO2 is an important technological approach to achieving carbon neutrality targets. As sealing barriers within potential storage formations, the sealing property of caprocks is crucial for the long-term or even permanent CO2 storage. Salt-gypsum caprocks exhibit favorable properties such as low porosity, low permeability, high structural stability, and high breakthrough pressure, making them promising candidates for long-term and secure CO2 storage. However, their physicochemical characteristics differ significantly from those of other lithological caprocks, posing challenges to evaluating their sealing performance for CO2 storage. Therefore, there is an urgent need to establish an evaluation method tailored to salt-gypsum caprocks. Firstly, based on the Analytic Hierarchy Process (AHP), a comprehensive evaluation index system was developed by considering key influencing factors, such as macro indicators, micro indicators, and breakthrough pressure, affecting the sealing performance of caprocks. Four grading levels were defined for each index, and the influence weight of each index on the sealing performance of salt-gypsum caprocks was determined. Secondly, by integrating the Fuzzy Comprehensive Evaluation Method, the total weight for evaluating the sealing performance of salt-gypsum caprocks for CO2 storage was calculated. This resulted in the development of a comprehensive evaluation method of CO2 storage sealing tailored to these types of caprocks. Finally, the method was applied to the Gaoshiti-Moxi block in the Sichuan Basin as a case study, where the CO2 storage sealing performance of its potential salt-gypsum caprock in a depleted gas reservoir was systematically evaluated. The results revealed that the total weight of the sealing performance evaluation of Gaoshiti-Moxi structural gas reservoir caprocks ranged from [2.5,3.0), corresponding to a grade of “relatively good”, indicating a relatively strong capacity for CO2 storage. This suggested the site was suitable for the future application of Carbon Capture and Storage (CCS) technology. The research results can provide technical guidance for site selection and storage safety evaluation of CO2 storage in depleted gas reservoirs with salt-gypsum caprocks.

Mi Kaifu, Shen Pengyu, Qiao Shihang et al.

The current staged deep shale gas testing and production mode cannot meet the requirement for interactions between surface processes and production data under complex conditions. Thus， it is necessary to develop a technology for integrated deep shale gas testing and production （hereinafter referred to as “integrated technology”） and associated equipment. Considering the complexities of deep shale gas such as high temperature and high pressure， multi-phase flowback and temperature and pressure fluctuation， an integrated technology featuring digital wellhead fine control and multi-phase separation intelligent metering was proposed. Through constructing a gas well choke production control and multi-phase two-dimensional separation control model， taking modular equipment as the carrier， the integrated technology was systematically analyzed. The field applications in deep shale gas fields of southern Sichuan Basin show that the number of process modules of the integrated technology is reduced by 60.6%， and the field operation time is shortened by 68.1%. For production capacity analysis， the dynamic control method of “fixed choke coarse adjustment + electric throttling fine adjustment” considering wellbore critical sand production can stably control the pressure drop rate of shale gas wells less than 0.1 MPa/d. For dynamic sand fluid discharge， the liquid level-time two-dimensional collaborative separation control method can effectively reduce the adjustment frequency of control units. The research results effectively improve the overall production efficiency of deep shale gas wells， and provide reference for the efficient development and optimal management of deep shale gas fields.

Yatai Ji, Teng Wang, Yuying Ge et al.

Discrete diffusion models have emerged as a promising direction for vision-language tasks, offering bidirectional context modeling and theoretical parallelization. However, their practical application is severely hindered by a train-inference discrepancy, which leads to catastrophic error cascades: initial token errors during parallel decoding pollute the generation context, triggering a chain reaction of compounding errors and leading to syntactic errors and semantic hallucinations. To address this fundamental challenge, we reframe the generation process from passive denoising to active refining. We introduce ReDiff, a refining-enhanced diffusion framework that teaches the model to identify and correct its own errors. Our approach features a two-stage training process: first, we instill a foundational revision capability by training the model to revise synthetic errors; second, we implement a novel online self-correction loop where the model is explicitly trained to revise its own flawed drafts by learning from an expert's corrections. This mistake-driven learning endows the model with the crucial ability to revisit and refine its already generated output, effectively breaking the error cascade. Extensive experiments demonstrate that ReDiff significantly improves the coherence and factual accuracy of generated content, enabling stable and efficient parallel generation far superior to traditional denoising methods. Our codes and models are available at https://rediff-hku.github.io/.

Zhennan Jiang, Kai Liu, Yuxin Qin et al.

Robotic manipulation policies are commonly initialized through imitation learning, but their performance is limited by the scarcity and narrow coverage of expert data. Reinforcement learning can refine polices to alleviate this limitation, yet real-robot training is costly and unsafe, while training in simulators suffers from the sim-to-real gap. Recent advances in generative models have demonstrated remarkable capabilities in real-world simulation, with diffusion models in particular excelling at generation. This raises the question of how diffusion model-based world models can be combined to enhance pre-trained policies in robotic manipulation. In this work, we propose World4RL, a framework that employs diffusion-based world models as high-fidelity simulators to refine pre-trained policies entirely in imagined environments for robotic manipulation. Unlike prior works that primarily employ world models for planning, our framework enables direct end-to-end policy optimization. World4RL is designed around two principles: pre-training a diffusion world model that captures diverse dynamics on multi-task datasets and refining policies entirely within a frozen world model to avoid online real-world interactions. We further design a two-hot action encoding scheme tailored for robotic manipulation and adopt diffusion backbones to improve modeling fidelity. Extensive simulation and real-world experiments demonstrate that World4RL provides high-fidelity environment modeling and enables consistent policy refinement, yielding significantly higher success rates compared to imitation learning and other baselines.

P. Kasar, D. K. Sharma, M. Ahmaruzzaman

Abstract The thermal and catalytic decomposition of waste plastics through pyrolysis is one of the best approaches of handling plastics waste and most prospective alternative of converting waste to wealth by transforming the waste plastics into lighter fuel oils which has potential to at least replenish petroleum resources if not replace the fossil fuels through this process of recycling. Further advancement in this area of research was to co-process waste plastics along with petroleum residues and other used oil with a similar intention, opening up a new prospect of reclaiming and upgrading altogether, two relatively low graded materials to a superior quality product which may be further refined for reprocessing in the petroleum refinery. In this paper, an attempt has been made to review the literature on cracking of plastics waste and it’s co-processing with petroleum residues and other heavy oil, the types of reactors and the catalyst employed in the process. The resulting product especially the liquid product from the co-processing of waste oil with polyolefin waste material has been found to possess good calorific values in the range of 44–47 MJ kg−1 while the heating value of the gaseous product was found to be in the range of 27–32 MJ Nm−3 and other characteristics similar to those of conventional fuel like diesel and thereby have a very good potential to be used as transportation fuel and other chemical feedstocks on further refining, while co-processing with other heavy oils residues have been found to have similar potential and prospect as an alternative to the conventional fuels and energy.

QU Jinghui, LIU Xingbin, LIU Dongmei et al.

The capacitance method is widely used to measure the water content of oil-water two-phase flow, whether it is the surface oil-water measurement or the stratification test of the production profile in the well. To measure the water holdup (water phase volume fraction), it is necessary to establish the relationship between the water holdup and the sensor capacitance, so it is necessary to establish the theoretical or empirical model of the miscible equivalent dielectric constant and the water holdup. Since the formation water produced from wells and tap water used in surface laboratory calibration have good electrical conductivity at low and medium frequencies, considering only the dielectric property without considering the electrical conductivity will have a significant impact on the accuracy of water retention measurement. Based on the effective electric field theory of dielectric polarization, the equivalent dielectric constant model of oil-water two-phase flow is established, considering water as a good conductor under the condition of low water content and fine bubble flow. According to this model, the equivalent permittivity of oil-water two-phase flow is determined only by the water holdup and the permittivity of oil phase, and has nothing to do with the permittivity of water phase and the conductivity of water phase. The correctness of the model is verified by preliminary experiments. The research is of great significance for the design and field application of the water content sensor for oil-water two-phase flow when the oil is continuous phase.

Shalbolova Urpash, Bissenov Kylyshbay, Makhanov Sagat

Diversification of the oil and gas complex of Kazakhstan is aimed not only at the development and development of oil and gas fields, but also at the further development of the manufacturing industry, in particular, at the construction of new and modernization of existing oil refining capacities. The article presents the results of analytical and research work to determine the effects of diversification on the national economy of Kazakhstan in the case of the construction of a new oil refinery and the modernization of existing production. At the stage of construction of oil refineries, indirect effects for the country's economy mainly appear, as a share of Kazakhstani content. In addition, the activities of the oil refining sector have a multiplier effect on the inter-industry balance in the structure of the national economy. The results of previously conducted studies are presented with an emphasis on multiplier effects, in which economic development takes place in other industries that are most interconnected with the petroleum products production sector. The purpose of the study is to reveal the economic effects for the economy of Kazakhstan when expanding production of oil refining products with added value.

Seyed Reza Nabavi, Mohammad Javad Jafari, Zhiyuan Wang

Background: Multilayer perceptron (MLP) aided multi-objective particle swarm optimization algorithm (MOPSO) is employed in the present article to optimize the liquefied petroleum gas (LPG) thermal cracking process. This new approach significantly accelerated the multi-objective optimization (MOO), which can now be completed within one minute compared to the average of two days required by the conventional approach. Methods: MOO generates a set of equally good Pareto-optimal solutions, which are then ranked using a combination of a weighting method and five multi-criteria decision making (MCDM) methods. The final selection of a single solution for implementation is based on majority voting and the similarity of the recommended solutions from the MCDM methods. Significant Findings: The deep learning (DL) aided MOO and MCDM approach provides valuable insights into the trade-offs between conflicting objectives and a more comprehensive understanding of the relationships between them. Furthermore, this approach also allows for a deeper understanding of the impact of decision variables on the objectives, enabling practitioners to make more informed, data-driven decisions in the thermal cracking process.

Peng Zhang, Ting Wu, Jinsheng Sun et al.

Existing interactive point cloud segmentation approaches primarily focus on the object segmentation, which aim to determine which points belong to the object of interest guided by user interactions. This paper concentrates on an unexplored yet meaningful task, i.e., interactive point cloud semantic segmentation, which assigns high-quality semantic labels to all points in a scene with user corrective clicks. Concretely, we presents the first interactive framework for point cloud semantic segmentation, named InterPCSeg, which seamlessly integrates with off-the-shelf semantic segmentation networks without offline re-training, enabling it to run in an on-the-fly manner. To achieve online refinement, we treat user interactions as sparse training examples during the test-time. To address the instability caused by the sparse supervision, we design a stabilization energy to regulate the test-time training process. For objective and reproducible evaluation, we develop an interaction simulation scheme tailored for the interactive point cloud semantic segmentation task. We evaluate our framework on the S3DIS and ScanNet datasets with off-the-shelf segmentation networks, incorporating interactions from both the proposed interaction simulator and real users. Quantitative and qualitative experimental results demonstrate the efficacy of our framework in refining the semantic segmentation results with user interactions. The source code will be publicly available.

Giovane Avancini, Nathan Shauer, Francisco T. Orlandini et al.

This contribution introduces the idea of refinement patterns for the generation of optimal meshes in the context of the Finite Element Method. The main idea is to generate a library of possible patterns on which elements can be refined and use this library to inform an h adaptive code on how to handle complex refinements in regions of interest. There are no restrictions on the type of elements that can be refined, and the patterns can be generated for any element type. The main advantage of this approach is that it allows for the generation of optimal meshes in a systematic way where, even if a certain pattern is not available, it can easily be included through a simple text file with nodes and sub-elements. The contribution presents a detailed methodology for incorporating refinement patterns into h adaptive Finite Element Method codes and demonstrates the effectiveness of the approach through mesh refinement of problems with complex geometries.

Sun Jiaxiang, Zhao Hongshan, Ma Li

The strata below the Qingshuihe Formation in the Junggar Basin are found with high uncertainty,low drillability and complicated pressure system,which severely restrains the progress of exploration and development by deep exploration wells in the lower-assemblage reservoirs.Therefore,the challenges in drilling deep layers in Well Zheng 10 were analyzed,and the key technologies were discussed,including optimization of casing program,development of special-shaped-cutter PDC bit,high-efficiency wellbore-stabilizing drilling fluid system and constant-bottomhole-pressure drilling.The results show that,the package of technologies for deep exploration wells in the lower-assemblage reservoir enabled Well Zheng 10 to record the largest well depth(7 802 m)in the Junggar Basin.The combination of the ridged-cutter PDC bit and the durable positive displacement motor(PDM)with uniform wall thickness has allowed an enhanced rate of penetration(ROP)in deep layers of the Junggar Basin hinterland.The synthetic-based drilling fluid can effectively suppress wellbore chipping due to hydration swelling of water-sensitive strata induced by drilling fluid filtrate invasion,and it is preferred for avoiding wellbore instability and ensuring wellbore quality in target layers in central Junggar Basin.The research findings provide guidance for the enhancement of ROP and efficiency in complex ultra-deep well drilling.

Wanpeng Zhang, Zongqing Lu

Large Language Models (LLMs) have demonstrated significant success across various domains. However, their application in complex decision-making tasks frequently necessitates intricate prompt engineering or fine-tuning, leading to challenges in unseen downstream tasks and heavy demands on computational resources. Meanwhile, Reinforcement Learning (RL) has been recognized as effective in decision-making problems but struggles in environments with sparse rewards, such as open-world games. To overcome these challenges, we introduce AdaRefiner, a novel framework designed to enhance the synergy between LLMs and RL feedback. The key component of AdaRefiner is a lightweight Adapter Language Model (LM), which automatically refines task comprehension based on feedback from RL agents. This method mitigates the need for intricate prompt engineering and intensive LLM fine-tuning while maintaining the LLMs' generalization abilities and enhancing their decision-making capabilities in downstream tasks. Empirical evaluations of AdaRefiner on 22 diverse tasks within the open-world game Crafter have demonstrated its superior effectiveness, especially in guiding agents towards higher-level and common-sense skills. Our work makes contributions to the automatic self-refinement of LLMs with RL feedback, offering a more adaptable and efficient solution for complex decision-making problems.

A. Kayode Coker

Sun Qiaolei, Jin Zuwen, Wang Jiangang et al.

In order to analyze the operation strength of HXJ180 offshore workover rig under various complex working conditions,the command stream method was used to build a workover rig model.By means of numerical simulation on the operation strength of the workover rig in 32 sub-working conditions corresponding to 4 working conditions and 8 wind directions,the cloud chart of displacement and equivalent stress distribution of the workover rig was obtained.Based on the <i>UC</i> value determination method in AISC 360—16,the operation strength of workover rig under various working conditions was checked and verified.The study results show that under all working conditions,the maximum equivalent stress is concentrated at the front end of the pipe setback region of upper movable base,and the wind direction has little effect on the position of the maximum stress point.Under two working conditions,the wind direction has little effect on the displacement distribution of the workover rig,and the maximum displacement point is located in the region of derrick crown base and bearing beam of gas-solid pipeline respectively.Under the unexpected and expectable working conditions,the wind direction has a significant impact on the displacement distribution of the upper part and monkey board of the derrick,and the maximum displacement point is also mainly concentrated in the region.Under all working conditions,the maximum <i>UC</i> value of the component occurs at the front end of the pipe setback region of upper movable base,and the maximum <i>UC</i> value is less than 1,indicating that the comprehensive strength of the workover rig is enough to meet the operating requirements of the API standard.The analysis method and related results in this paper provide reference for strength check and design improvement of drilling and workover rigs.

Mukhutdinov N.U., Khozhiev B.I., Karshiev O.A. et al.

The article discusses the features of the geological structure, the identification of petroleum bearing promising zones and the scientific substantiation of the optimal areas of prospecting and exploration for the Khorezm monocline and adjacent territories through a comprehensive analysis of cosmo-geological study and geological and geophysical data. A comprehensive interpretation made it possible to minimize the existing multivariance in solving certain problems related to the tectonic features of the study area, determining possible zones of oil and gas accumulation, identifying fracture zones, predicting structural traps of the sedimentary cover for setting prospecting seismic surveys in a poorly studied, but potentially petroleum bearing promising territory of Uzbekistan.