Hasil untuk "Petroleum refining. Petroleum products"

Limin Wang, Shijie Zhang, Zi Long et al.

Nonylphenols (NPs), widely used as emulsifiers in petroleum production and refining, are compounds of environmental concern, with endocrine-disrupting effects. They can be released during oil extraction and processing, carried into petroleum products, and subsequently emitted during downstream applications such as combustion. Despite these potential pathways, information on their occurrence in petroleum streams remains limited, partly due to the lack of reliable methods for measuring NPs in complex petroleum matrices. In this study, we developed an analytical method combining normal-phase chromatography (NPC), solid-phase extraction (SPE), and liquid chromatography–Orbitrap high-resolution mass spectrometry (LC–Orbitrap-HRMS) for NP determination in crude oils and petroleum products. NPC was performed using alumina (5% water deactivation) as the stationary phase. The column was eluted sequentially with n-hexane, n-hexane/dichloromethane (4:1 and 1:1, v/v), dichloromethane, and dichloromethane/methanol (2:1, v/v). The first three fractions were discarded, and the remaining two fractions were combined and further purified using a C18 SPE cartridge to analysis. The method showed high recovery (82.8 ± 2.6%) and a low detection limit (1.0 ng/g) in crude oil. Application revealed widespread occurrence of NPs, with concentrations up to 784.4 ng/g in crude oils and up to 439.1 ng/g in refined fuels, indicating that these compounds can persist through refining and may be released during downstream use. These results demonstrate that the method is suitable for the routine monitoring of NPs in petroleum-related samples and provide a practical tool for supporting sustainable refining practices and improved environmental management in the upstream oil and gas sector.

LÜ Wei, XUE Fangfang, HE Sixian et al.

Winsor Ⅲ microemulsion flooding is an effective means to solve the low production percentage of high-temperature， high-salinity and low-permeability reservoirs， but the research on the construction and mechanism understanding of Winsor Ⅲ microemulsion flooding system is insufficient. In this paper， the Winsor Ⅲ microemulsion system with temperature resistance and salt resistance was constructed through the solubility test， oil-water interfacial tension test， and phase behavior test of two types of surfactants and their composite systems under the conditions of simulated reservoirs. Then， the feasibility of the system in enhancing oil recovery of low-permeability reservoirs was explored through core displacement experiments， contact angle testing experiments， and visual core displacement experiments. The mechanisms of enhancing oil recovery were analyzed from two aspects： changes in wettability and solubilization and emulsification. The results show that alcohol ether carboxylate surfactant ST 982-B can construct a Winsor Ⅲ microemulsion system at the mass fraction of 0.3%-0.5%； the microemulsion system can effectively enhance oil recovery of low-permeability reservoirs by 7.8%-9.7%. The microemulsion system can effectively improve the rock wettability， and the “denudation” of oil film after solubilization and emulsification of the system further enhances the oil displacement efficiency.

Hou Min, Zhou Tianming, Chen Degang et al.

To meet the demands of exploration and development of 10 000-meter ultradeep wells， research on key technologies for the design of kilo-ton heavy-duty ultra-high derrick substructure was conducted. After analyzing the functional requirements and key technical problems to be solved of the heavy-duty ultra-high derrick substructure， the main technical parameters and structural scheme of the heavy-duty ultra-high derrick substructure were determined. The derrick adopts a front-opening K-shaped structure， with an effective height of 61 m and a maximum rated static hook load of 11 250 kN， and the overall rotational hoisting is carried out using wire ropes. The substructure is divided into front and rear floors. The front floor is a parallelogram structure with a height of 15 m， and the equipment on the floor can be installed when the substructure is located at a low position， and then raised to the working position via hydraulic cylinders for overall rotational hoisting. The rear floor is used to place equipment such as drawworks. Moreover， the key technologies for the design of heavy-duty ultra-high derrick substructure were systematically summarized， including the design technology of derrick substructure hoisting system， analysis technology of derrick substructure stability and structural optimization technology of derrick substructure. The research results provide reference for the development of the supporting derrick substructure of 10 000-m drilling rigs in China.

WANG SHULI, WANG JINGUO, ZHANG CHENGSEN et al.

Accurate interpretation of well logging data is crucial for the evaluation of reservoir fluid properties in oil and gas exploration. Conventional well logging methods rely on petrophysical models that correlate parameters such as porosity, permeability, and oil and gas saturation with reservoir fluid properties to achieve reservoir classification. However, complex geological conditions often lead to issues such as anomalies, multi-factor coupling, and ambiguous fluid boundaries in well logging data. These challenges limit the adaptability of conventional methods and bring uncertainties in interpretation results. To improve the accuracy of reservoir fluid evaluation, this study incorporated cloud model theory into conventional well logging evaluation and proposed an evaluation method for reservoir fluid based on a 2D cloud model. The method selected porosity and gas saturation as key logging parameters and utilized cloud models to process the fuzziness and randomness in well logging data, thereby establishing a mathematical model for reservoir fluid classification. First, a 2D cloud model for well logging evaluation was derived based on cloud model theory, with clarified geophysical significance assigned to its mathematical parameters (expectation, entropy, and hyper-entropy). 2D cloud diagrams of the reservoir were generated using a cloud generator. Subsequently, similarity analysis was applied to quantitatively classify reservoir types, enhancing interpretation accuracy. To validate the effectiveness of this method, well logging data from the Kuqa Depression in the Tarim Basin were used for application analysis, with results compared with those obtained from conventional methods, cloud model evaluation, and well testing. The results showed that the proposed method accurately characterized reservoir fluid properties in complex reservoirs. Compared with conventional methods, the 2D cloud model not only provided qualitative classification of reservoir types but also quantified uncertainties in fluid properties, thus improving the stability and reliability of evaluation results. The findings indicate that the reservoir fluid evaluation method based on 2D cloud model effectively reflects reservoir fluid characteristics and exhibits strong adaptability in complex reservoir environments. The final evaluation results demonstrate strong consistency with well testing results, verifying the method’s feasibility and effectiveness. As a valuable supplement to conventional well logging interpretation, this method provides a new approach for improving the accuracy of well logging data interpretation and optimizing fluid property identification in complex reservoirs.

Li Yong, Xu Zhendong, Wang Chenmin et al.

To solve the problems of low cutting efficiency and unstable cutting fracture quality caused by severe vibration and eccentric rotation of the cutter device in downhole hydraulic cutting operation of tubing and casing， a cutter down-attached anchor tool was developed， which could also realize conventional power up-attachment under special working conditions. The tool uses a rotary seal structure to achieve the “rotation + anchoring” function of the anchor， and solves the problems of small anchoring range and unstable anchoring through blade-type anchor blocks and a dual-piston structure. The field test results show that the down-attached anchor and the hydraulic cutter exhibit excellent operational compatibility. Compared with up-attached anchor cutting operation， the tool string is not eccentric after cutting operation of down-attached anchor， and the overall cutting time is reduced from 4 min to 2 min（ø73.0 mm tubing）， with cutting time shortened by 50%. The down-attached anchor structure can simultaneously anchor at least two sizes of tubing， indicating that it maintains anchoring stability even at larger opening angles. The tool features a simplified and reliable structure， with universal applicability across multiple sizes and working conditions. The successful development of the tool provides guidance for further improving the efficiency of field cutting operations.

Mohammad Abdul Hafeez Khan, Yash Jain, Siddhartha Bhattacharyya et al.

Text-to-image (T2I) generation models have made significant strides but still struggle with prompt sensitivity: even minor changes in prompt wording can yield inconsistent or inaccurate outputs. To address this challenge, we introduce a closed-loop, test-time prompt refinement framework that requires no additional training of the underlying T2I model, termed TIR. In our approach, each generation step is followed by a refinement step, where a pretrained multimodal large language model (MLLM) analyzes the output image and the user's prompt. The MLLM detects misalignments (e.g., missing objects, incorrect attributes) and produces a refined and physically grounded prompt for the next round of image generation. By iteratively refining the prompt and verifying alignment between the prompt and the image, TIR corrects errors, mirroring the iterative refinement process of human artists. We demonstrate that this closed-loop strategy improves alignment and visual coherence across multiple benchmark datasets, all while maintaining plug-and-play integration with black-box T2I models.

Tanmay Ambadkar, Đorđe Žikelić, Abhinav Verma

Logical specifications have been shown to help reinforcement learning algorithms in achieving complex tasks. However, when a task is under-specified, agents might fail to learn useful policies. In this work, we explore the possibility of improving coarse-grained logical specifications via an exploration-guided strategy. We propose AutoSpec, a framework that searches for a logical specification refinement whose satisfaction implies satisfaction of the original specification, but which provides additional guidance therefore making it easier for reinforcement learning algorithms to learn useful policies. AutoSpec is applicable to reinforcement learning tasks specified via the SpectRL specification logic. We exploit the compositional nature of specifications written in SpectRL, and design four refinement procedures that modify the abstract graph of the specification by either refining its existing edge specifications or by introducing new edge specifications. We prove that all four procedures maintain specification soundness, i.e. any trajectory satisfying the refined specification also satisfies the original. We then show how AutoSpec can be integrated with existing reinforcement learning algorithms for learning policies from logical specifications. Our experiments demonstrate that AutoSpec yields promising improvements in terms of the complexity of control tasks that can be solved, when refined logical specifications produced by AutoSpec are utilized.

Junyang Zhang

A graph $Γ$ is said to be stable if $\mathrm{Aut}(Γ\times K_2)\cong\mathrm{Aut}(Γ)\times \mathbb{Z}_{2}$ and unstable otherwise. If an unstable graph is connected, non-bipartite and any two of its distinct vertices have different neighbourhoods, then it is called nontrivially unstable. We establish conditions guaranteeing the instability of various graph products, including direct products, direct product bundles, Cartesian products, strong products, semi-strong products, and lexicographic products. Inspired by a condition for the instability of direct product bundles, we propose a new sufficient condition for circulant graphs to be unstable and refine existing instability conditions from the literature. Based on these results, we categorize unstable circulant graphs into two distinct types and further propose a classification framework.

GAO Yongjin, CHEN Yi, SUN Xiangcan et al.

Volcanic rocks， such as andesite and tuff， are widely developed in the Carboniferous system in Karamali area， northern Xinjiang. The tectonic background of their formation is still controversial. In this paper， two typical Carboniferous profiles from Karamali Mountain were selected. Andesite and tuff samples from the lower Carboniferous Jiangbasitao Formation， the upper Carboniferous Bashan Formation， and Huxingliang Formation were observed under a microscope， and their main trace elements were analyzed. The results show that the contents of MgO （0.5%-2.35%） and Mg# （15.7%-42.5%） are low to medium in samples， indicating that the crystallization differentiation of rock samples occurs during the formation process. The distribution curve of trace elements in the samples is obviously different from that of N-MORB， E-MORB， and OIB. The samples are almost all in the calc-alkaline series in Ta/Yb-Th/Yb and La-Y-Nb diagrams， while the samples fall subduction components and approach the mantle series in Nb/Yb-La/Yb diagrams. The Nb/Y-Rb/Y and Ba/La-Ba/Nb diagrams also show that the samples are affected by both fluid enrichment and melt enrichment. The Nb/U and Ce/Pb ratios of the samples are 1.6-11.1 and 0.47-12.2， respectively， which are close to the range of continental crust and indicate that the source of the material may be mixed with continental crust materials. According to the Th-Hf/3-Nb/16 diagram， all the samples are located in the island arc basalt region. The comprehensive analysis shows that Karamali area was in the continental island arc environment of active continental margin from the early Carboniferous to the late Carboniferous， and this environment lasted at least from the early Carboniferous to the late Carboniferous.

Siwei Xia, Xueqi Hu, Li Sun et al.

In StyleGAN, convolution kernels are shaped by both static parameters shared across images and dynamic modulation factors $w^+\in\mathcal{W}^+$ specific to each image. Therefore, $\mathcal{W}^+$ space is often used for image inversion and editing. However, pre-trained model struggles with synthesizing out-of-domain images due to the limited capabilities of $\mathcal{W}^+$ and its resultant kernels, necessitating full fine-tuning or adaptation through a complex hypernetwork. This paper proposes an efficient refining strategy for dynamic kernels. The key idea is to modify kernels by low-rank residuals, learned from input image or domain guidance. These residuals are generated by matrix multiplication between two sets of tokens with the same number, which controls the complexity. We validate the refining scheme in image inversion and domain adaptation. In the former task, we design grouped transformer blocks to learn these token sets by one- or two-stage training. In the latter task, token sets are directly optimized to support synthesis in the target domain while preserving original content. Extensive experiments show that our method achieves low distortions for image inversion and high quality for out-of-domain editing.

SONG shuling, YANG Erlong, SHA Mingyu

The availability of shale oil directly affects the degree of effective exploration and development, and the mobility of shale oil is closely related to its occurrence state. Therefore, studying the occurrence state of shale oil plays an important role in its development. The pore model is established by graphene and quartz, and the occurrence state of n-octane and its mixture in nanopores is studied by molecular simulation method. The effects of pore size, temperature, pressure, shale oil composition, wall wettability and wall composition on the occurrence state are analyzed. The results show that: ①shale oil is multi-layer adsorbed in the pores and symmetrical about the pore center, and the thickness of the adsorption layer is 0.4~0.5 nm; ②The larger the pore size of the reservoir, the higher the temperature, the lower the pressure, the lighter the molecular component, the weaker the polarity, and the higher the wall wettability are, the more unfavorable the adsorption of oil molecules on the wall is;③ In the combined wall, due to the influence of graphene wall, the adsorption amount of shale oil molecules increases with the increase of quartz wall wetting humidity. In addition, the adsorption transfer phenomenon of n-hexanoic acid and cyclohexane also occurs.

Kyowoon Lee, Seongun Kim, Jaesik Choi

Diffusion-based planning has shown promising results in long-horizon, sparse-reward tasks by training trajectory diffusion models and conditioning the sampled trajectories using auxiliary guidance functions. However, due to their nature as generative models, diffusion models are not guaranteed to generate feasible plans, resulting in failed execution and precluding planners from being useful in safety-critical applications. In this work, we propose a novel approach to refine unreliable plans generated by diffusion models by providing refining guidance to error-prone plans. To this end, we suggest a new metric named restoration gap for evaluating the quality of individual plans generated by the diffusion model. A restoration gap is estimated by a gap predictor which produces restoration gap guidance to refine a diffusion planner. We additionally present an attribution map regularizer to prevent adversarial refining guidance that could be generated from the sub-optimal gap predictor, which enables further refinement of infeasible plans. We demonstrate the effectiveness of our approach on three different benchmarks in offline control settings that require long-horizon planning. We also illustrate that our approach presents explainability by presenting the attribution maps of the gap predictor and highlighting error-prone transitions, allowing for a deeper understanding of the generated plans.

Paavo Alku, Sudarsana Reddy Kadiri, Dhananjaya Gowda

In this study, formant tracking is investigated by refining the formants tracked by an existing data-driven tracker, DeepFormants, using the formants estimated in a model-driven manner by linear prediction (LP)-based methods. As LP-based formant estimation methods, conventional covariance analysis (LP-COV) and the recently proposed quasi-closed phase forward-backward (QCP-FB) analysis are used. In the proposed refinement approach, the contours of the three lowest formants are first predicted by the data-driven DeepFormants tracker, and the predicted formants are replaced frame-wise with local spectral peaks shown by the model-driven LP-based methods. The refinement procedure can be plugged into the DeepFormants tracker with no need for any new data learning. Two refined DeepFormants trackers were compared with the original DeepFormants and with five known traditional trackers using the popular vocal tract resonance (VTR) corpus. The results indicated that the data-driven DeepFormants trackers outperformed the conventional trackers and that the best performance was obtained by refining the formants predicted by DeepFormants using QCP-FB analysis. In addition, by tracking formants using VTR speech that was corrupted by additive noise, the study showed that the refined DeepFormants trackers were more resilient to noise than the reference trackers. In general, these results suggest that LP-based model-driven approaches, which have traditionally been used in formant estimation, can be combined with a modern data-driven tracker easily with no further training to improve the tracker's performance.

Aman Madaan, Niket Tandon, Prakhar Gupta et al.

Like humans, large language models (LLMs) do not always generate the best output on their first try. Motivated by how humans refine their written text, we introduce Self-Refine, an approach for improving initial outputs from LLMs through iterative feedback and refinement. The main idea is to generate an initial output using an LLMs; then, the same LLMs provides feedback for its output and uses it to refine itself, iteratively. Self-Refine does not require any supervised training data, additional training, or reinforcement learning, and instead uses a single LLM as the generator, refiner, and feedback provider. We evaluate Self-Refine across 7 diverse tasks, ranging from dialog response generation to mathematical reasoning, using state-of-the-art (GPT-3.5, ChatGPT, and GPT-4) LLMs. Across all evaluated tasks, outputs generated with Self-Refine are preferred by humans and automatic metrics over those generated with the same LLM using conventional one-step generation, improving by ~20% absolute on average in task performance. Our work demonstrates that even state-of-the-art LLMs like GPT-4 can be further improved at test time using our simple, standalone approach.

Sascha Diefenbacher, Vinicius Mikuni, Benjamin Nachman

Machine learning-based simulations, especially calorimeter simulations, are promising tools for approximating the precision of classical high energy physics simulations with a fraction of the generation time. Nearly all methods proposed so far learn neural networks that map a random variable with a known probability density, like a Gaussian, to realistic-looking events. In many cases, physics events are not close to Gaussian and so these neural networks have to learn a highly complex function. We study an alternative approach: Schrödinger bridge Quality Improvement via Refinement of Existing Lightweight Simulations (SQuIRELS). SQuIRELS leverages the power of diffusion-based neural networks and Schrödinger bridges to map between samples where the probability density is not known explicitly. We apply SQuIRELS to the task of refining a classical fast simulation to approximate a full classical simulation. On simulated calorimeter events, we find that SQuIRELS is able to reproduce highly non-trivial features of the full simulation with a fraction of the generation time.

Ilfat Z. Rakhmatullin, Sergej V. Efimov, Ekaterina I. Kondratyeva et al.

Application of high resolution 13C nuclear magnetic resonance (NMR) spectroscopy to characterize Cuba oil and oil-containing rock samples from Cuban basin was demonstrated. The chemical shifts of 13C NMR functional groups for later determination the composition of the oil and rock samples were determined. The different contribution of the studied samples in the aliphatic and aromatic areas was determined. Molar fractions of primary, secondary, quaternary, tertiary, aromatic groups, aromaticity factor and the mean length of hydrocarbon chain length of aliphatic hydrocarbons were estimated. Comparative analysis on the quantitative level for all major hydrocarbon components, the aromaticity factor and the mean length of the hydrocarbon chain were carried out.

Mingyuan Fan, Cen Chen, Chengyu Wang et al.

Recent works have brought attention to the vulnerability of Federated Learning (FL) systems to gradient leakage attacks. Such attacks exploit clients' uploaded gradients to reconstruct their sensitive data, thereby compromising the privacy protection capability of FL. In response, various defense mechanisms have been proposed to mitigate this threat by manipulating the uploaded gradients. Unfortunately, empirical evaluations have demonstrated limited resilience of these defenses against sophisticated attacks, indicating an urgent need for more effective defenses. In this paper, we explore a novel defensive paradigm that departs from conventional gradient perturbation approaches and instead focuses on the construction of robust data. Intuitively, if robust data exhibits low semantic similarity with clients' raw data, the gradients associated with robust data can effectively obfuscate attackers. To this end, we design Refiner that jointly optimizes two metrics for privacy protection and performance maintenance. The utility metric is designed to promote consistency between the gradients of key parameters associated with robust data and those derived from clients' data, thus maintaining model performance. Furthermore, the privacy metric guides the generation of robust data towards enlarging the semantic gap with clients' data. Theoretical analysis supports the effectiveness of Refiner, and empirical evaluations on multiple benchmark datasets demonstrate the superior defense effectiveness of Refiner at defending against state-of-the-art attacks.

Palash Panja, Wei Jia, Alec Nelson et al.

A machine learning architecture composed of convolutional long short-term memory (convLSTM) is developed to predict spatio-temporal parameters in the SACROC oil field, Texas, USA. The spatial parameters are recorded at the end of each month for 30 years (360 months), approximately 83% (300 months) of which is used for training and the rest 17% (60 months) is kept for testing. The samples for the convLSTM models are prepared by choosing ten consecutive frames as input and ten consecutive frames shifted forward by one frame as output. Individual models are trained for oil, gas, and water saturations, and pressure using the Nesterov accelerated adaptive moment estimation (Nadam) optimization algorithm. A workflow is provided to comprehend the entire process of data extraction, preprocessing, sample preparation, training, testing of machine learning models, and error analysis. Overall, the convLSTM for spatio-temporal prediction shows promising results in predicting spatio-temporal parameters in porous media.

Soo Ye Kim, Jianming Zhang, Simon Niklaus et al.

Depth maps are used in a wide range of applications from 3D rendering to 2D image effects such as Bokeh. However, those predicted by single image depth estimation (SIDE) models often fail to capture isolated holes in objects and/or have inaccurate boundary regions. Meanwhile, high-quality masks are much easier to obtain, using commercial auto-masking tools or off-the-shelf methods of segmentation and matting or even by manual editing. Hence, in this paper, we formulate a novel problem of mask-guided depth refinement that utilizes a generic mask to refine the depth prediction of SIDE models. Our framework performs layered refinement and inpainting/outpainting, decomposing the depth map into two separate layers signified by the mask and the inverse mask. As datasets with both depth and mask annotations are scarce, we propose a self-supervised learning scheme that uses arbitrary masks and RGB-D datasets. We empirically show that our method is robust to different types of masks and initial depth predictions, accurately refining depth values in inner and outer mask boundary regions. We further analyze our model with an ablation study and demonstrate results on real applications. More information can be found at https://sooyekim.github.io/MaskDepth/ .

Xinyu Mao, Nianyin Li, Fei Chen et al.

Abstract In situ heat systems are a technology that effectively solves paraffin deposition and improves oil recovery. Generally, the oxidation–reduction reaction of sodium nitrite and ammonium chloride generates a large amount of heat to promote the melting of paraffin. An in situ heat system combined with an acid-resistant fracturing fluid system can form an in situ heat fracturing fluid system, which solves the problem of the poor reformation effect caused by cold damage during the fracturing process of low-pressure and high-pour-point oil reservoirs. In this paper, with the goals of system heating up to 50 °C, a low H+ concentration, a high exotherm, and reduction of the toxic and harmful by-product NOX, the preferred in situ heat system was found to comprise 1.6 mol/L ammonium chloride, 1.0 mol/L sodium nitrite, and 0.8% hydrochloric acid. The effect of five factors on the heat production of the reaction was studied experimentally, and a reaction kinetic equation for the in situ heat system was proposed based on the results. The results showed that increasing the concentration of the reactants and lowering the ambient temperature produced more heat. The in situ heat system was used to conduct a crude oil cold damage elimination experiment, and the results of the removal experiments verified that the system could effectively but not completely reduce the cold damage. Overall, the in situ heat fracturing fluid system formed by the preferred in situ heat system combined with an acid-resistant fracturing fluid system could avoid cold damage in the formation during construction and increase the output.