Geoenergy Science and Engineering最新文献

{"title":"Novel poroperm characteristics design method for cement sheath to improve the long-term sealing integrity in gas storage well","authors":"Sheng Huang ,&nbsp;Tao Mao ,&nbsp;Donghua Su ,&nbsp;Zaoyuan Li ,&nbsp;Weitao Song ,&nbsp;Jinfei Sun","doi":"10.1016/j.geoen.2025.213832","DOIUrl":"10.1016/j.geoen.2025.213832","url":null,"abstract":"<div><div>Poroperm characteristics of the cement sheath are the crucial factors affecting the long-term safe and stable storage of underground gas in gas storage wells. In this paper, the permeability, connected porosity, and average connected pore radius of the cement sheath were characterized using a permeability test system, mercury intrusion, and computed tomography (CT). A dynamic permeability calculation model, which considered the pore compaction effect, was established and validated by test results. On this basis, a gas seepage length analysis model was further derived. The risk and influencing factors of gas seepage were explored and the critical poroperm characteristics indexes of the cement sheath were proposed. The results found that the connected pores inside the cement sheath will cause the risk of gas seepage. Decreasing the gas storage pressure, average connected pore radius, connected porosity, and permeability of the cement sheath can shorten the gas seepage length. The poroperm characteristics requirements charts were designed, and the average connected pore radius (<em>r</em>) and permeability (<em>K</em>_<em>a</em>) of the cement sheath under various well depths (<em>H</em>) and gas storage pressures (<em>P</em>_e) were quantified, that when the <em>r</em> and <em>K</em>_<em>a</em> meet <em>K</em>_<em>a</em> (10³ <em>P</em>_e – 9.8 <em>H</em>–121034 <em>r</em>⁻¹)/(13400 + 1.52 <em>H</em>) &lt; 0, the gas sealing of the gas storage wells' cement sheath can be ensured under different <em>H</em> and <em>P</em>_e. The research presents a novel approach for predicting the cement sheath sealing and proposes a set of specific performance indicators to effectively guide the design of the cement sheath's poroperm characteristics in gas storage wells, which is significant for guaranteeing the long-term safe operation of gas storage wells.</div></div>","PeriodicalId":100578,"journal":{"name":"Geoenergy Science and Engineering","volume":"250 ","pages":"Article 213832"},"PeriodicalIF":0.0,"publicationDate":"2025-03-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143591395","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Forward modeling and data inversion of cased-hole logging parameters for four detectors based on X-ray source","authors":"Haoyu Zhang ,&nbsp;Wensheng Wu ,&nbsp;Zhangxin Chen ,&nbsp;Xiaoyu Song","doi":"10.1016/j.geoen.2025.213830","DOIUrl":"10.1016/j.geoen.2025.213830","url":null,"abstract":"<div><div>In complex well conditions, the evaluation of cased-hole logging parameters such as formation density <em>ρ</em>_<em>b</em>, casing thickness <em>h</em>_<em>s</em>, cement sheath thickness <em>h</em>_<em>c</em> and cement sheath density <em>ρ</em>_<em>c</em> can effectively reduce the risk of downhole operations when compared to open-hole logging. Moreover, the use of X-ray sources to replace isotopic radioactive sources in nuclear logging has emerged as a new trend, but there is relatively little research on the calculation of logging parameters for X-ray cased-hole logging. As a result, further research on forward modeling of detector response and data inversion is required. Because the cement sheath density response was not included in the prior forward modeling formula, we intend to include a geological model constraint whose objective is precisely the cement sheath density in the case of unclear cement sheath circumstances. To achieve this, the energy window related to the cement sheath density was identified first based on the X-ray source. Subsequently, a four-detector formation model was developed under relevant constraints. The model was then extended into a cement sheath density correction function and integrated into the original response formula of the four detectors concerning the counting rate. Finally, a new forward equation for X-ray cased-hole logging parameters of the four detectors was established. For the inversion problem of cased-hole logging parameters, Differential Evolution (DE) algorithm was employed to find the optimal values. The proposed forward and inverse methods for X-ray cased-hole logging parameters were applied to simulation examples and a practical case, which yielded satisfactory results. This method can effectively capture the geological characteristics and monitor the changes in casing and cement sheath, which offers significant implications for subsequent geological exploration.</div></div>","PeriodicalId":100578,"journal":{"name":"Geoenergy Science and Engineering","volume":"250 ","pages":"Article 213830"},"PeriodicalIF":0.0,"publicationDate":"2025-03-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143601553","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Effect of nano-surface on water flow in methane hydrate montmorillonite reservoir: A molecular dynamics study","authors":"Zhaoran Wu ,&nbsp;Kai Zhang ,&nbsp;Shihui Ma ,&nbsp;Guijing Li ,&nbsp;Lei Wang ,&nbsp;Zaixing Liu ,&nbsp;Hongpeng Xu ,&nbsp;Yanghui Li","doi":"10.1016/j.geoen.2025.213826","DOIUrl":"10.1016/j.geoen.2025.213826","url":null,"abstract":"<div><div>The seepage characteristics of methane hydrate reservoirs in the South China Sea are important for hydrate exploitation projects. The properties of reservoir surface will affect the flow state and thus affect gas and water production. Therefore, study the flow characteristics of the fluid in the reservoir is essential. The flow of water in montmorillonite and methane hydrate nanopores was studied by molecular dynamics simulation. The results indicate that the electronegativity of montmorillonite impacts the distribution of water molecules in the nanopore. Additionally, the water film near the wall separated the fluid in the middle of the pore and the wall. Methane hydrates trap water molecules as part of a cage structure. For the condition of Poiseuille flow, the first density peak decreases with the increase of driving force. The Navier-Stokes equation yields a fluid with an apparent viscosity two orders of magnitude smaller than the theoretical viscosity. During the flow process, the hydrogen bond network density remains constant in the montmorillonite nanopore, whereas it decreases in the methane hydrate nanopore. Adding cations in the nanopore encrypts the hydrogen bond network, impeding the flow, and the increase in the number of hydrogen bonds depends on the total charge of the cations.</div></div>","PeriodicalId":100578,"journal":{"name":"Geoenergy Science and Engineering","volume":"250 ","pages":"Article 213826"},"PeriodicalIF":0.0,"publicationDate":"2025-03-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143601617","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Numerical simulation of hydraulic fracturing optimization in multi-well and multi-layer shale gas development: Insights from inter-well interference analysis","authors":"Botong Du ,&nbsp;Fengshou Zhang ,&nbsp;Egor Dontsov ,&nbsp;Keyu Meng","doi":"10.1016/j.geoen.2025.213825","DOIUrl":"10.1016/j.geoen.2025.213825","url":null,"abstract":"<div><div>The development of unconventional shale gas is gradually evolving towards multi-layer and multi-well developments with small well spacing to improve resource recovery. However, the inter-well interference mechanism and the procedures for determining an optimal well spacing in multi-layer developments remain unclear. This study establishes a field-scale distinct element numerical modeling based on the data from the Luzhou deep shale gas reservoir, in Southwest China. The hydraulic fracturing process of three staggered wells is hydromechanically simulated, followed by an analysis of the interaction mechanism between multiple wells and the effect of well spacing on hydraulic fracturing. The results show that the effect of interference between wells on fracturing effectiveness varies non-linearly with increasing well spacing. Fracture growth is limited when well spacing is too small, resulting in reduced stimulated reservoir volume (SRV). Excessive well spacing is detrimental to the formation of complex fracture networks and results in a waste of reservoir resources between wells. Furthermore, a multi-factor scoring method was proposed to evaluate the effectiveness of hydraulic fracturing. This method considers reservoir stimulation, fracture network quality, and resource utilization to generate comparable scores. The assessment result clearly shows that an optimal well spacing exists for multi-well and multi-layer development and numerical modeling can be used to identify this value. This study describes a possible mechanism for multi-well interactions and provides a methodology for the design of multi-well and multi-layer developments in Luzhou and related shale gas development areas.</div></div>","PeriodicalId":100578,"journal":{"name":"Geoenergy Science and Engineering","volume":"250 ","pages":"Article 213825"},"PeriodicalIF":0.0,"publicationDate":"2025-03-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143611070","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Synergistic enhancement of mechanical properties in oil-well cement reinforced with KH560 and Aluminosilicate fibers: A multi-scale perspective","authors":"Tianle Liu ,&nbsp;Weijian Wang ,&nbsp;Mingsheng Chen ,&nbsp;Xiaoyang Ni","doi":"10.1016/j.geoen.2025.213828","DOIUrl":"10.1016/j.geoen.2025.213828","url":null,"abstract":"<div><div>Oil well cementitious materials are brittle materials and are prone to fracture under stress in complex wells, reducing seal integrity. To address the above problems, this study uses KH560 coupling agent and Alumino-silicate fibers (ASF) to synergistically modify oil well cementitious materials, and proposes a new method to enhance the overall performance of oil well cement. Mechanical tests, CT, SEM, FTIR, EDS, and molecular dynamics simulations were utilized to comprehensively evaluate the mechanical strength, pore structure, micro-morphology, chemical reactions, elemental distribution, and modification mechanism of the materials using multi-scale analytical methods. Experimental temperature and pressure are 80 °C and normal pressure. The macro-mechanical tests results demonstrated that the compressive and flexural strengths of the modified cement test block of 7 day increased by 4.17 % and 22.79 %, respectively. Additionally, the optimal dosage of pretreated ASF was 1.0 % of the oil well cement. The mesoscopic and microscopic CT and SEM tests results indicated that the modified ASF can optimize the pore structure and exhibited a robust bond with cement. The FTIR and EDS chemical tests results show that the characteristic peaks of Al-O, Si-O and O-H chemical bond vibrations are extraordinarily sharp, and the Al, O and Si atoms are mainly distributed near the ASF. The nanomolecular simulation results show that the main reason for enhanced interfacial bonding after modification is the stable O_KH560-Si_C-S-H, O_KH560-Si_ASF and O_KH560-Al_Al2O3 bonding, and the KH560 molecules play the “bridge” role between C-S-H and ASF. This study provides useful insights and feasible solutions to control high temperature cracking of cementing structures.</div></div>","PeriodicalId":100578,"journal":{"name":"Geoenergy Science and Engineering","volume":"250 ","pages":"Article 213828"},"PeriodicalIF":0.0,"publicationDate":"2025-03-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143591373","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Rate transient and pressure distribution of a multiply fractured vertical well in the shale gas reservoir with randomly distributed natural fractures based on embedded discrete fracture model","authors":"Guoqiang Xing ,&nbsp;Wuyang Yang ,&nbsp;Dongdong Shao ,&nbsp;Mingxian Wang ,&nbsp;Xiangji Dou ,&nbsp;Xiang Wang","doi":"10.1016/j.geoen.2025.213833","DOIUrl":"10.1016/j.geoen.2025.213833","url":null,"abstract":"<div><div>Due to the different sizes of hydraulic fractures (<em>HFs</em>) and natural fractures (<em>NFs</em>), complex fracture network with random fracture properties easily occurs in the shale gas reservoir. By employing the finite difference method and embedded discrete fracture model (<em>EDFM</em>), the numerical solution was obtained to investigate the rate behavior and pressure distribution of the multiply fractured vertical well in the shale gas reservoir with randomly distributed <em>NFs</em>. The results of the numerical model were validated with those of the commercial software. With the newly-developed numerical solution, effects of reservoir properties and fracture properties on the rate transient and pressure distribution of the shale gas reservoir are discussed in detail and the simulation of randomly distributed <em>NFs</em> with random fracture properties is further conducted. The results show that larger Langmuir volume and pressure comprehensively contribute to increasing the well's daily production. Meanwhile, the number of intersected <em>NFs</em> near the wellbore and <em>HFs</em> are conducive to the well's daily production while the permeability of <em>HFs</em> is only effective in increasing production during the early production times. Furthermore, pressure distribution results reveal that the shape of distributed volumes mainly depends on fracture properties, such as the number of <em>NFs</em>, while the size of distributed volumes mainly relates to the permeability of the shale-matrix and the angle between adjacent <em>HFs</em>. In addition, the random fracture simulation results imply that the reservoir pressure distribution near the wellbore decreases rapidly when <em>HFs</em> intersect with <em>NFs</em> in the early production times and pressure throughout the shale gas reservoir is almost equal to the bottom-hole pressure except the corner of the shale gas reservoir after 5–10 years production.</div></div>","PeriodicalId":100578,"journal":{"name":"Geoenergy Science and Engineering","volume":"250 ","pages":"Article 213833"},"PeriodicalIF":0.0,"publicationDate":"2025-03-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143611069","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Mechanism of enhanced oil recovery by supercritical CO2 in nanopores, a molecular dynamics simulation","authors":"Jingru Wang ,&nbsp;Yuetian Liu ,&nbsp;Bo Zhang ,&nbsp;Bowei Zhang ,&nbsp;Yuting He ,&nbsp;Rukuan Chai","doi":"10.1016/j.geoen.2025.213804","DOIUrl":"10.1016/j.geoen.2025.213804","url":null,"abstract":"<div><div>Using CO₂ for enhanced oil recovery (EOR) in reservoirs is a Carbon Capture, Utilization, and Storage (CCUS) technique that achieves both economic and social benefits. Supercritical CO₂ demonstrates unique advantages in unconventional oil and gas extraction, but its mechanisms require further exploration and study. Current research has primarily focused on studying the miscibility and adsorption mechanisms of supercritical CO₂ with simple alkane mixtures. This paper uses molecular simulation methods to investigate the microscopic mechanisms of dynamic extraction of crude oil in rock pores by supercritical CO₂ during the oil EOR process. The study compares the distribution of crude oil in pores during the supercritical CO₂ flooding process under varying conditions of temperature, pressure, injection pressure, and alkane composition. CO₂ receives such cycle of stripping, driving, adsorbing, and transporting alkanes adsorbed on the surfaces. Additional research indicates that changing the composition ratio of crude oil in the pores affects the extraction efficiency of CO₂. With increased injection pressure, the extraction efficiency improves for higher-density crude oil, while for lower-density crude oil, the extraction efficiency decreases. Therefore, in practical development, considering that crude oil density gradually increases as extraction progresses, it is recommended to begin with a lower injection pressure and gradually increase it over time to maximize recovery efficiency. This study provides support and recommendations for the investigation of the mechanisms of EOR using supercritical CO₂.</div></div>","PeriodicalId":100578,"journal":{"name":"Geoenergy Science and Engineering","volume":"250 ","pages":"Article 213804"},"PeriodicalIF":0.0,"publicationDate":"2025-03-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143601531","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Geophysical monitoring of CO2 storage using rock physics template","authors":"Javad Sharifi","doi":"10.1016/j.geoen.2025.213808","DOIUrl":"10.1016/j.geoen.2025.213808","url":null,"abstract":"<div><div>Assessing and monitoring CO₂ storage through seismic signals is crucial for verifying and evaluating the safety of CO₂ injection. I integrated geological parameters and geophysical observations to develop an alternative methodology for monitoring geological CO₂ sequestration. In this context, I artificially reconstructed typical sedimentary rocks for CO₂ storage using a hybrid rock physics model based on solid matrix and pore types. Subsequently, various scenarios of CO₂ saturation were modeled, incorporating a range of lithologies, porosity, pore type, fluids, saturation degree, saturation type, and anisotropy. The seismic response was then theoretically extracted from each scenario. Analyzing the results indicated that each factor has a distinct effect and complex interaction on the seismic response of CO₂-saturated rock. Consequently, numerous factors were investigated simultaneously and incorporated into a single diagram, which I refer to as a template. Next, different rock physics templates were designed based on the most influential parameters affecting seismic response (e.g., V_P/V_S ratio, acoustic impedance, and lambda-mu-rho). As a result, the obtained diagrams were validated using ultrasonic measurements under simulated reservoir conditions. The validation data yielded promising results, demonstrating the applicability and reliability of the proposed templates for monitoring CO₂ storage. The findings of this research can be applied not only to the existing CO₂ but also to identifying the type of host rocks. Finally, the limitations and advantages of the modeling were discussed in detail, and it is recommended that the proposed templates be updated according to local geological settings.</div></div>","PeriodicalId":100578,"journal":{"name":"Geoenergy Science and Engineering","volume":"250 ","pages":"Article 213808"},"PeriodicalIF":0.0,"publicationDate":"2025-03-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143601554","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Integrated framework of Total Organic Carbon (TOC) content prediction and application in shale","authors":"Lu Qiao ,&nbsp;Shengyu Yang ,&nbsp;Qinhong Hu ,&nbsp;Huijun Wang ,&nbsp;Taohua He","doi":"10.1016/j.geoen.2025.213811","DOIUrl":"10.1016/j.geoen.2025.213811","url":null,"abstract":"<div><div>The accurate determination of Total Organic Carbon (TOC) content in source rocks through well logs is critical for reservoir identification and evaluation. While physical models, such as the ΔlogR method and density-based approaches, have traditionally been used for TOC prediction, they are often limited by challenges related to parameter determination and accuracy. Additionally, predicting TOC in highly heterogeneous shale formations remains particularly challenging. Although machine learning has emerged as a powerful tool for data-driven TOC prediction, its performance is frequently constrained by manual hyperparameter tuning, which may result in suboptimal model outcomes. This study proposes an innovative framework that integrates Swarm Intelligence Optimization Algorithm (SIOA) with Gaussian Process Regression (GPR), termed SIOA-GPR. The proposed approach harnesses the optimization capabilities of SIOA to systematically determine GPR hyperparameters, thereby minimizing prediction errors and enhancing model performance. To validate the framework, TOC prediction was conducted using a dataset comprising 110 core samples and corresponding well logs from the Y1 well in the Subei Basin, China. The workflow involved splitting the data into two portions: 80% for training and hyperparameter optimization, and 20% for independent model validation. Experimental results demonstrate that the SIOA-GPR framework outperforms existing integrated machine learning models. The proposed method achieves superior performance metrics, such as higher correlation coefficients and lower mean absolute errors compared to its counterparts. Furthermore, the framework exhibits robust generalization capabilities across different shale intervals, highlighting its potential for broader applications in reservoir evaluation and engineering development. The SIOA-GPR framework addresses the limitations of both physical models and traditional machine learning approaches by automating hyperparameter tuning, thereby enhancing prediction accuracy and reliability. The proposed methodology offers a new perspective for TOC prediction in unconventional reservoirs, providing additive value to the existing body of literature on shale resource evaluation and development strategies.</div></div>","PeriodicalId":100578,"journal":{"name":"Geoenergy Science and Engineering","volume":"250 ","pages":"Article 213811"},"PeriodicalIF":0.0,"publicationDate":"2025-03-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143611068","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Intelligent prediction of rate of penetration through meta-learning and data augmentation synergy under limited sample","authors":"Zhengchao Ma ,&nbsp;Jintao Weng ,&nbsp;Junkai Zhang ,&nbsp;Yi Zhang ,&nbsp;Yanji Hao ,&nbsp;Shouceng Tian ,&nbsp;Gensheng Li ,&nbsp;Tianyu Wang","doi":"10.1016/j.geoen.2025.213818","DOIUrl":"10.1016/j.geoen.2025.213818","url":null,"abstract":"<div><div>The inherent data scarcity in oil and gas drilling operations frequently leads to insufficient accuracy and limited generalization capability of data-driven Rate of Penetration (ROP) prediction models. To address the critical challenge of poor model generalizability under small-sample conditions, this study proposes an innovative machine learning framework that synergistically integrates meta-learning strategies with synthetic data generation. The framework employs a meta-learning-optimized training protocol combined with a self-attention-enhanced multi-layer perceptron architecture, enabling rapid adaptation to new well conditions with limited samples. Furthermore, a Wasserstein generative adversarial network with gradient penalty is implemented to generate high-fidelity synthetic data, effectively mitigating sample scarcity. Experimental results demonstrate that the proposed method maintains robust performance under complex operational scenarios contaminated by gaussian noise of varying intensities, outperforming benchmark models including Transformer and Deep Neural Networks. Across different data augmentation ratios, the framework achieves consistent performance improvements, with mean absolute percentage error reduced by an average of 27%. Global and local interpretability analyses quantitatively uncover the driving mechanisms of ten parameters, such as weight on bit and rotary speed, on ROP predictions, thereby enhancing the model's interpretability and transparency. This study overcomes the tripartite challenges of “limited data, poor generalization, and interpretability constraints” in ROP prediction, while validating the model's scalability and the efficacy of data augmentation strategies. It addresses issues of poor robustness and limited generalization caused by small sample sizes, providing a reference for intelligent modeling in other data-scarce scenarios within the energy sector.</div></div>","PeriodicalId":100578,"journal":{"name":"Geoenergy Science and Engineering","volume":"250 ","pages":"Article 213818"},"PeriodicalIF":0.0,"publicationDate":"2025-03-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143578735","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}