Hasil untuk "Petroleum refining. Petroleum products"

Jakob Piribauer, Vinzent Zschuppe

Iterative abstraction refinement techniques are one of the most prominent paradigms for the analysis and verification of systems with large or infinite state spaces. This paper investigates the changes of truth values of system properties expressible in computation tree logic (CTL) when abstractions of transition systems are refined. To this end, the paper utilizes modal logic by defining alethic modalities expressing possibility and necessity on top of CTL: The modal operator $\lozenge$ is interpreted as "there is a refinement, in which ..." and $\Box$ is interpreted as "in all refinements, ...". Upper and lower bounds for the resulting modal logics of abstraction refinement are provided for three scenarios: 1) when considering all finite abstractions of a transition system, 2) when considering all abstractions of a transition system, and 3) when considering the class of all transition systems. Furthermore, to prove these results, generic techniques to obtain upper bounds of modal logics using novel types of so-called control statements are developed.

Jianli Qiang, YanChi Yang, Mingjin Cai et al.

Abstract Due to the fluid leak-off effect in the reservoir, the bottomhole pressure decreases after the pumps are shut down. Analyzing the shut-in pressure decline curve of a fractured well is a common method for determining fracturing parameters. Although the G-function pressure decline analysis method briefly explains the pressure decline process, it is not accurate for calculating fracture parameters in fractured low-porosity gas reservoirs. This paper considers the influence of natural fractures on the leak-off coefficient and proposes an approach to evaluate fracture complexity by using the fluctuation characteristics of the construction pressure curve and the G-function characteristics during fracture closure. In this study, the pressure decline curve was segmented to determine fracture parameters, and a shut-in pressure decline analysis model for fractured low-porosity gas reservoirs was established. The fracture complexity is characterized by the fluctuation of the superposition derivative curve, and the approximate series is calculated to quantitatively evaluate fracture complexity. Field data from multiple wells were used to calculate the approximate series, thereby verifying its practicality. Results from actual case data show a positive correlation between fracture complexity and the approximate series. Additionally, this paper adopts a comprehensive filtering model to remove the noise caused by water hammer effects during the shut-in process, improving data quality and analysis accuracy. The feasibility and reliability of the model are validated using actual data from fractured wells in the Dabei Oilfield.

LI Chunlei, YANG Heshan, ZHANG Hongxia et al.

Reservoir’s static pressure is an essential basic data in the development and research of oil and gas fields. Its acquisition conditions are strict， and the sample number is extremely small. Static pressures are estimated with empirical methods based on dynamic pressure data during the production process； however， data errors are significant. To address the above issues， a static pressure prediction method for CO2 flooding oil reservoirs based on the time series partitioning Transformer model was proposed， utilizing deep learning theory. Model parameters were selected based on correlation analysis， and iterative interpolation was used to fill in samples to construct a static pressure prediction sample set. According to the principle of channel independence， the multivariate time series was divided into univariate time series， and a time series partitioning mechanism was introduced to divide the time series into subsequential blocks to capture local features. Based on the Transformer model architecture， a multi-head self-attention mechanism was utilized to extract features， and a self-supervised learning mechanism was employed to enhance the ability to capture complex dynamic characteristics， achieving the prediction of reservoirs’ static pressure. The research results indicate that the proposed model can accurately predict the static pressures at the middle of the oil reservoir of each well in the active well group， significantly improving prediction accuracy.

Yili REN, Changmin ZENG, Xin LI et al.

Existing sandstone rock structure evaluation methods rely on visual inspection, with low efficiency, semi-quantitative analysis of roundness, and inability to perform classified statistics in particle size analysis. This study presents an intelligent evaluation method for sandstone rock structure based on the Segment Anything Model (SAM). By developing a lightweight SAM fine-tuning method with rank-decomposition matrix adapters, a multispectral rock particle segmentation model named CoreSAM is constructed, which achieves rock particle edge extraction and type identification. Building upon this, we propose a comprehensive quantitative evaluation system for rock structure, assessing parameters including particle size, sorting, roundness, particle contact and cementation types. The experimental results demonstrate that CoreSAM outperforms existing methods in rock particle segmentation accuracy while showing excellent generalization across different image types such as CT scans and core photographs. The proposed method enables full-sample, classified particle size analysis and quantitative characterization of parameters like roundness, advancing reservoir evaluation towards more precise, quantitative, intuitive, and comprehensive development.

Xu Yang, Weijun Shen, Xizhe Li et al.

Abstract For reservoirs exhibiting severe heterogeneity due to prolonged water flooding, conventional polymer flooding demonstrates limited capacity to alter reservoir properties under such conditions. Crosslinked polymer flooding, characterized by its three-dimensional network structure, significantly enhances reservoir conformance, sweep efficiency, and oil displacement efficiency, making it a key technique for strongly heterogeneous reservoirs in the middle to late stages of development. In this study, a geological model based on a one-injection–one-production well pattern was constructed to simulate the crosslinked polymer flooding process. The model incorporates critical factors including formation heterogeneity, wall adsorption, and the inaccessible pore volume associated with polymer injection. Using the Computer Modelling Group’s Steam, Thermal, and Advanced Processes Reservoir Simulator (CMG-STARS), model was validated through grid independence analysis and calibrated via history matching. A systematic investigation was conducted to assess the influence of key parameters—namely, crosslinked polymer concentration, injection rate, and injection strategy—on oil displacement performance. The results indicate that, under the studied conditions, a well spacing of 200 m, a mass concentration of 0.55%, and an injection rate of 400 m³/d effectively reduce polymer adsorption and shear degradation while promoting the formation of a stable gel, thereby enhancing plugging efficiency. An injection strategy initiated at a 35% water cut, combined with 2–3 alternating cycles, was found to mitigate water channeling through early-stage reservoir modification. This study identifies the dominant factors controlling the oil displacement performance of crosslinked polymer systems, clarifies their relative advantages and limitations compared to conventional polymers, and offers a reliable reference framework for parameter selection in field-scale crosslinked polymer flooding operations.

Xuan Yu, Dayan Guan, Yanfeng Gu

Multimodal Large Language Models (MLLM) often struggle to interpret high-resolution images accurately, where fine-grained details are crucial for complex visual understanding. We introduce Zoom-Refine, a novel training-free method that enhances MLLM capabilities to address this issue. Zoom-Refine operates through a synergistic process of \textit{Localized Zoom} and \textit{Self-Refinement}. In the \textit{Localized Zoom} step, Zoom-Refine leverages the MLLM to provide a preliminary response to an input query and identifies the most task-relevant image region by predicting its bounding box coordinates. During the \textit{Self-Refinement} step, Zoom-Refine then integrates fine-grained details from the high-resolution crop (identified by \textit{Localized Zoom}) with its initial reasoning to re-evaluate and refine its preliminary response. Our method harnesses the MLLM's inherent capabilities for spatial localization, contextual reasoning and comparative analysis without requiring additional training or external experts. Comprehensive experiments demonstrate the efficacy of Zoom-Refine on two challenging high-resolution multimodal benchmarks. Code is available at \href{https://github.com/xavier-yu114/Zoom-Refine}{\color{magenta}github.com/xavier-yu114/Zoom-Refine}

XUE Yifan, WEN Zhigang, HUANG Yahao et al.

The study of the filling veins in deep reservoirs within the strike-slip fault zone in the north of Fuman Oilfield utilizes a range of methods including petrographic characteristics, analysis of rare earth elements andSr（strontium） isotopes, fluorescence spectra of oil inclusions, microscopic thermodynamics, and U-Pb isotopic dating of carbonate rocks. The findings reveal two stages of calcite vein formation in this area. These veins originate from the formation water of the middle and Lower Ordovician sources, with no evidence of oxidizing fluid intrusion, suggesting that the deep to ultra-deep oil and gas reserves have maintained good sealing properties in later stages. Furthermore, based on the burial history deduced from inclusions and low U-Pb isotope dates from carbonate rocks, it has been determined that there are three distinct stages of hydrocarbon charging in the deep Ordovician strata of the northern strike-slip fault zone in the Tarim Basin. These stages correspond to （459±7.2） Ma（middle Caledonian）, （348±18） Ma（early Hercynian）, and 268 Ma（late Hercynian）. It is noted that the early Hercynian period was the key phase for hydrocarbon accumulation in the deep and ultra-deep carbonate rocks in the north of Fuman Oilfield, with a significant correlation observed between oil and gas charging and fault activity.

Abdul Rehman Baig, Sulaiman A. Alarifi, Mobeen Murtaza et al.

Abstract Sand production in oil and gas wells is a significant concern, leading to equipment erosion, reduced well productivity, and safety hazards. Researchers have developed an eco-friendly solution to consolidate sand via an Enzyme-induced Carbonate Precipitation (EICP) process. It fortifies loose sand in wells, preventing it from resurfacing. This study addresses this challenge by developing a novel EICP solution effective at high temperatures (120 °C). This advancement goes beyond previous formulations, which often exhibited low strength at elevated temperatures. In this study, we developed six different solutions to consolidate sand at different temperatures with various bio- and synthetic polymers, the resulted sand consolidation has been tested by obtaining the precipitation composition after consolidation, visualizing consolidated sand structures, assessing strength and measuring permeability of the consolidated sand. AN 125, a synthetic copolymer based on Acrylamide and 2-Acrylamido-2-Methylpropane Sulfonic Acid (AM-AMPS), emerged as the most effective additive. It fostered the strongest consolidated sand at both temperatures (2,175 psi at 70 °C and 2,155 psi at 120 °C). It also exhibited superior thermal stability compared to bio-polymers like xanthan gum, which degraded at 120 °C. The EICP solution with AN 125 led to a moderate permeability decrease of around 30% during simulated sand pack flooding, indicating minimal impact on well flow. The developed formulation offers a robust and environmentally friendly approach to sand consolidation in oil and gas wells, enhancing well integrity and production efficiency. Furthermore, this work emphasizes the significance of a proper methodology towards evaluating the suitability of bio-polymers and synthetic copolymers for sand consolidation using EICP formulations.

Tongwen JIANG, Huan QI, Zhengmao WANG et al.

Based on the microfluidic technology, a microscopic visualization model was used to simulate the gas injection process in the initial construction stage and the bottom water invasion/gas injection process in the cyclical injection-production stage of the underground gas storage (UGS) rebuilt from water-invaded gas reservoirs. Through analysis of the gas-liquid contact stabilization mechanism, flow and occurrence, the optimal control method for lifecycle efficient operation of UGS was explored. The results show that in the initial construction stage of UGS, the action of gravity should be fully utilized by regulating the gas injection rate, so as to ensure the macroscopically stable migration of the gas-liquid contact, and greatly improve the gas sweeping capacity, providing a large pore space for gas storage in the subsequent cyclical injection-production stage. In the cyclical injection-production stage of UGS, a constant gas storage and production rate leads to a low pore space utilization. Gradually increasing the gas storage and production rate, that is, transitioning from small volume to large volume, can continuously break the hydraulic equilibrium of the remaining fluid in the porous media, which then expands the pore space and flow channels. This is conducive to the expansion of UGS capacity and efficiency for purpose of peak shaving and supply guarantee.

Hugh Medal, Samuel Affar

We study a class of two-stage stochastic programs in which the second stage includes a set of components with uncertain capacity, and the expression for the distribution function of the uncertain capacity includes first-stage variables. Thus, this class of problems has the characteristics of a stochastic program with decision-dependent uncertainty. A natural way to formulate this class of problems is to enumerate the scenarios and express the probability of each scenario as a product of the first-stage decision variables; unfortunately, this formulation results in an intractable model with a large number of variable products with high-degree. After identifying structural results related to upper and lower bounds and how to improve these bounds, we present a successive refinement algorithm that successively and dynamically tightens these bounds. Implementing the algorithm within a branch-and-cut method, we report the results of computational experiments that indicate that the successive refinement algorithm significantly outperforms a benchmark approach. Specifically, results show that the algorithm finds an optimal solution before the refined state space become too large.

Aodi Li, Liansheng Zhuang, Xiao Long et al.

Domain generalization aims to learn a model from multiple training domains and generalize it to unseen test domains. Recent theory has shown that seeking the deep models, whose parameters lie in the flat minima of the loss landscape, can significantly reduce the out-of-domain generalization error. However, existing methods often neglect the consistency of loss landscapes in different domains, resulting in models that are not simultaneously in the optimal flat minima in all domains, which limits their generalization ability. To address this issue, this paper proposes an iterative Self-Feedback Training (SFT) framework to seek consistent flat minima that are shared across different domains by progressively refining loss landscapes during training. It alternatively generates a feedback signal by measuring the inconsistency of loss landscapes in different domains and refines these loss landscapes for greater consistency using this feedback signal. Benefiting from the consistency of the flat minima within these refined loss landscapes, our SFT helps achieve better out-of-domain generalization. Extensive experiments on DomainBed demonstrate superior performances of SFT when compared to state-of-the-art sharpness-aware methods and other prevalent DG baselines. On average across five DG benchmarks, SFT surpasses the sharpness-aware minimization by 2.6% with ResNet-50 and 1.5% with ViT-B/16, respectively. The code will be available soon.

Patrick J. Park, Sebastian Herzele, Timothy W. Koeth

We re-examine a common narrative that experimental errors by Walther Bothe in 1941 led Germany to abandon graphite as a reactor moderator during World War II. Using document-based nuclear archaeology, we first show that both American and German scientists used an incorrect carbon scattering cross section, thereby undermining the accuracy of all wartime data, including their conclusions on carbon's absorption. Moreover, we argue that the availability of exceptionally pure petroleum coke in the United States, rather than any academic breakthrough, decisively enabled their production of nuclear-grade graphite. In contrast, Bothe's Siemens electrographite had more boron contamination than any graphites considered in Fermi's experiments, rendering it genuinely impractical as a moderator. By reframing the decision to eschew graphite as a deliberate decision rather than a mere experimental oversight, we believe the German decision was a rational consequence of a complex interplay between material constraints and wartime priorities.

Wangtao Sun, Shizhu He, Jun Zhao et al.

With good explanatory power and controllability, rule-based methods play an important role in many tasks such as knowledge reasoning and decision support. However, existing studies primarily focused on learning chain-like rules, which limit their semantic expressions and accurate prediction abilities. As a result, chain-like rules usually fire on the incorrect grounding values, producing inaccurate or even erroneous reasoning results. In this paper, we propose the concept of tree-like rules on knowledge graphs to expand the application scope and improve the reasoning ability of rule-based methods. Meanwhile, we propose an effective framework for refining chain-like rules into tree-like rules. Experimental comparisons on four public datasets show that the proposed framework can easily adapt to other chain-like rule induction methods and the refined tree-like rules consistently achieve better performances than chain-like rules on link prediction. The data and code of this paper can be available at https://anonymous.4open.science/r/tree-rule-E3CD/.

Masataka Satoh

In this paper, we introduce the notion of generalized quasi-shuffle products and give a criterion for their associativity. These extend the quasi-shuffle products introduced by Hoffman, which are often used to describe the stuffle and shuffle product for multiple zeta values. For $q$-analogues of multiple zeta values, the description of an analogue for the shuffle product can often not be described with the classical notion of quasi-shuffle products. We show that our generalization gives a natural extension to also include these types of products and we prove a generalization of a duality between the $q$-shuffle product and the $q$-stuffle product.

Boštjan Brešar, Jaka Hedžet

Given a graph $G$ and assuming that some vertices of $G$ are infected, the $r$-neighbor bootstrap percolation rule makes an uninfected vertex $v$ infected if $v$ has at least $r$ infected neighbors. The $r$-percolation number, $m(G,r)$, of $G$ is the minimum cardinality of a set of initially infected vertices in $G$ such that after continuously performing the $r$-neighbor bootstrap percolation rule each vertex of $G$ eventually becomes infected. In this paper, we consider percolation numbers of strong products of graphs. If $G$ is the strong product $G_1\boxtimes \cdots \boxtimes G_k$ of $k$ connected graphs, we prove that $m(G,r)=r$ as soon as $r\le 2^{k-1}$ and $|V(G)|\ge r$. As a dichotomy, we present a family of strong products of $k$ connected graphs with the $(2^{k-1}+1)$-percolation number arbitrarily large. We refine these results for strong products of graphs in which at least two factors have at least three vertices. In addition, when all factors $G_i$ have at least three vertices we prove that $m(G_1 \boxtimes \dots \boxtimes G_k,r)\leq 3^{k-1} -k$ for all $r\leq 2^k-1$, and we again get a dichotomy, since there exist families of strong products of $k$ graphs such that their $2^{k}$-percolation numbers are arbitrarily large. While $m(G\boxtimes H,3)=3$ if both $G$ and $H$ have at least three vertices, we also characterize the strong prisms $G\boxtimes K_2$ for which this equality holds. Some of the results naturally extend to infinite graphs, and we briefly consider percolation numbers of strong products of two-way infinite paths.

Shuai Jian, Zhang Yi

The analysis on the operating pressure cycle characteristics of gas pipeline is an important research basis for predicting the fatigue life of the pipeline. Based on the rain flow counting method, the pressure cycle in the random load spectrum was counted, the outlet pressure change history of two stations in a pipeline was counted, and the pressure cycle amplitude, mean value, frequency proportion and damage proportion of the pipeline in the stations were analyzed. The analysis results show that the Nmber of pressure cycles with pressure ratio greater than 0.8 in the gas pipeline accounts for more than 90% of the total cycles; the acquisition cycle has little effect on the counting of large pressure cycles, but has great effect on the counting of small pressure cycles; the acquisition cycle is selected for pipeline pressure cycle based on the principle of no loss of large cycle; the higher the frequency proportion of large cycle, the greater the proportion of damage to the pipeline is, but when the Nmber of cycles is large and the cycle amplitude is small, a large Nmber of small cycles are the main causes of pipeline damage. The analysis conclusions provide a theoretical support for the fatigue life prediction of oil and gas pipelines.

Jianan Zhao, Qianlong Wen, Mingxuan Ju et al.

Graph structure learning (GSL), which aims to learn the adjacency matrix for graph neural networks (GNNs), has shown great potential in boosting the performance of GNNs. Most existing GSL works apply a joint learning framework where the estimated adjacency matrix and GNN parameters are optimized for downstream tasks. However, as GSL is essentially a link prediction task, whose goal may largely differ from the goal of the downstream task. The inconsistency of these two goals limits the GSL methods to learn the potential optimal graph structure. Moreover, the joint learning framework suffers from scalability issues in terms of time and space during the process of estimation and optimization of the adjacency matrix. To mitigate these issues, we propose a graph structure refinement (GSR) framework with a pretrain-finetune pipeline. Specifically, The pre-training phase aims to comprehensively estimate the underlying graph structure by a multi-view contrastive learning framework with both intra- and inter-view link prediction tasks. Then, the graph structure is refined by adding and removing edges according to the edge probabilities estimated by the pre-trained model. Finally, the fine-tuning GNN is initialized by the pre-trained model and optimized toward downstream tasks. With the refined graph structure remaining static in the fine-tuning space, GSR avoids estimating and optimizing graph structure in the fine-tuning phase which enjoys great scalability and efficiency. Moreover, the fine-tuning GNN is boosted by both migrating knowledge and refining graphs. Extensive experiments are conducted to evaluate the effectiveness (best performance on six benchmark datasets), efficiency, and scalability (13.8x faster using 32.8% GPU memory compared to the best GSL baseline on Cora) of the proposed model.

Chris Kottke, Frédéric Rochon

Manifolds with fibered corners arise as resolutions of stratified spaces, in many body compactifications of vector spaces, moduli spaces, and other settings. We define a category of fibered corners manifolds which has products and transverse fiber products, generalizing both the resolutions of products of stratified spaces and many body products, which are special cases. The product in the fibered corners category is a resolution of the cartesian product by blow-up which we call the 'ordered product'. This ordered product is a natural product for wedge (aka incomplete edge) metrics and quasi-fibered boundary metrics, a class which includes QAC and QALE metrics.

Sadegh Dahim, Behrouz Harimi, Mohammad Hossein Ghazanfari et al.

The liquid bridge is considered a good means to maintain capillary continuity between overlying matrix blocks if its stability in fractures is preserved. Despite several studies focusing on the liquid bridge in different environments, little attention is paid to dig through a single liquid bridge between thin sections of minerals found in fractured reservoirs. In this study, a set of experiments was conducted to investigate liquid bridge stability and surface profile for different values of liquid volume and surface wettability conditions. It is found that critical fracture aperture is linearly proportional to the contact angle and to the third root of liquid volume, which is depicted by a newly developed expression. An accurate method for computation of capillary pressure of liquid bridge (known as fracture capillary pressure) from the experimentally determined interface profiles, based on the numerical solution of the Young-Laplace equation, is proposed. Following the Plateau sequence, both nodoid and unduloid shape bridges are observed with an increase in fracture aperture, corresponding to positive and negative fracture capillary pressure, respectively. It is interesting to note that instability of liquid bridges occurs at small negative values of capillary force where some attraction force exists between fracture faces. By applying a 1D mathematical model of liquid dripping, a modified expression for the prediction of critical fracture aperture is proposed, including fluid and flow-related parameters. The findings of this study help to better incorporate the role of liquid bridge and corresponding fracture capillary pressure in capillary continuity in fractured porous media.

Iosifidi A.G., Popov V.V.

New paleomagnetic determinations are presented for a collection of rocks of the Middle-Upper Devonian (Givetian and Frasnian stages), sampled from an outcrop on the southwestern slope of the Pavlov granite quarry. Three characteristic components of natural remanent magnetization have been identified. The bipolar component C corresponds to the Late Paleozoic remagnetization in the Carboniferous and two components of Givetian - Frasnian age. The bipolar component of D2NR passes the geomagnetic field polarity reversal test. The obtained position of the paleomagnetic pole in terms of the bipolar component of the natural remanent magnetization is consistent with the available data on the Middle Devonian strata of the Main Devonian Field (Russian platform).