Machine learning–based method for automated well-log processing and interpretation

Wu, Po-Yen; Jain, Vikas; Kulkarni, Mandar; Abubakar, Aria

doi:10.1190/segam2018-2996973.1

Cited by 29 publications

(11 citation statements)

References 8 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Por último, Khandelwal y Singh (2010) se dedicaron a predecir los valores de porosidad neutrón y densidad empleando los parámetros de entrada de los registros de rayos gamma, resistividad y sónico. Esto sin mencionar las investigaciones realizadas para automatizar la interpretación estratigráfica (Igbokwe, 2011;Wu y Nyland, 1987) y el procesamiento de registros de pozo (Wu et al, 2018).…”

Section: Discussionunclassified

Inferencias del factor fotoeléctrico (PEF) en registros de pozo con machine learning

Pardo-Díaz¹,

Vargas²

2021

bol. geol.

View full text Add to dashboard Cite

Los registros de pozo convencionales son importantes para la realización de análisis petrofísicos, amarres sísmicos y correlación estratigráfica. El presente estudio propone una metodología para realizar predicciones en estos registros haciendo uso de machine learning (ML), una herramienta altamente aplicada en múltiples disciplinas. El software de entrenamiento utilizado fue WEKA (Waikato Environment for Knowledge Analysis), en el que se generó un modelo para la predicción del registro de Absorción Fotoeléctrica (PDPE o PEF), a partir de tres atributos, los registros de Rayos Gamma (GRGC), Densidad (DEN) y Corrección de Densidad (DCOR). Esta metodología fue aplicada a registros de pozo de la Formación San Fernando, cuya unidad equivalente sería la Formación Mirador, en el sector sur-occidental de los Llanos Orientales de Colombia. Fueron usados los registros de trece pozos para hacer el entrenamiento del modelo y otros seis pozos fueron usados para evaluar el desempeño de este. Los resultados confirman la posibilidad de correlacionar registros que miden características diferentes en las rocas y evidencian que las inferencias en registros de pozo con ML requieren un filtrado minucioso para tomar la tendencia de los datos, y una optimización clara para prevenir el sobreentrenamiento en el modelo.

show abstract

Section: Discussionunclassified

Inferencias del factor fotoeléctrico (PEF) en registros de pozo con machine learning

Pardo-Díaz¹,

Vargas²

2021

bol. geol.

View full text Add to dashboard Cite

show abstract

“…这就导致类似方法对专家工作辅助的有限性, 未能完全将专家时间从繁杂的基础性工作中释放, 对专家的人工依赖过高, 所需要的时间成本过于高昂. 因此, 自动化的油气储集层识别技术一直以来就是学术界以及工业界探索的重要课题 [9,10] . 近些年来, 随着计算机技术的不断发展, 尤其是以机器学习、深度学习为代表的人工智能技术, 使得越来越多的研究者开始尝试使用智能技术来对海量的地质测井数据进行分析解释.…”

Section: 地质与油气智能勘测研究unclassified

“…事实上, 随着计算机技术的不断发展, 尤其是人工智能技术的突破, 近些年来越来越多的研究者开始关注采用计算机技术辅助测井解释以及智能油气储集层识别任务 [9,10] . 真实工业场景下的智能油其中, 基于相似性建模主要是依赖时间序列数据间的相似度度量方法, 对序列类别按照相似关系进行预测.…”

unclassified

Method of oil and gas reservoir detection based on geological knowledge distillation learning

李徵¹,

刘淇²,

王喆锋³

et al. 2020

Sci. Sin.-Inf.

View full text Add to dashboard Cite

“…It has been widely explored in all different applications, including those in geophysics, due to its great potential in pattern recognition and nonlinear regression. In geophysics, specifically for well log measurements, unsupervised learning algorithms such as self-organizing map, cross-entropy clustering, and Gaussian mixture model have been applied to logs for automatic zoning and lithofacies recognition (Fung et al, 1995;Wu et al, 2018). Supervised learning methods, mainly classification or regression neural networks, have been employed for lithological classification, and permeability, porosity, or shear wave velocity estimation from well log data (Ahmadi & Chen, 2019;An et al, 2018;Anemangely et al, 2019;Bestagini et al, 2017).…”

Section: Introductionmentioning

confidence: 99%

Shale Anisotropy Model Building Based on Deep Neural Networks

You

Cheng

2020

JGR Solid Earth

View full text Add to dashboard Cite

Seismic anisotropy parameters are essential in the processing and interpretation of modern array data with multicomponent, long offsets and wide azimuth acquisitions. Traditional well logs do not measure anisotropy in a vertical well and thus cannot provide the needed information. Conventional calibration‐based as well as recent inversion‐based rock physics modeling methods involve tuning parameters and subjective choices that are largely empirical and difficult to generalize. Here we present a machine learning approach to alleviate these problems. Since it is impossible to collect massive labeled field well log data, we generate paired synthetic data of features (porosity, density, vertical P and S wave velocities, P wave and shear moduli) and labels (bulk and shear moduli of rock matrices and aspect ratio of ellipsoidal cracks). By tuning hyperparameters we obtain an optimal fully connected neural network with four hidden layers that fits well with the synthetic data. The neural network is applied to published laboratory measurements and field well log data from a Chinese well and a U.S. well without any modification. We show that anisotropy models estimated by the deep neural network agree well with the inversion results and with the laboratory measurements. The neural network optimized by extensive training based on massive synthetic data removes the subjectivity in parameter selection, generalizes to different geological environments, and has the potential to provide real‐time anisotropy estimation while logging.

show abstract

Machine learning–based method for automated well-log processing and interpretation

Cited by 29 publications

References 8 publications

Inferencias del factor fotoeléctrico (PEF) en registros de pozo con machine learning

Inferencias del factor fotoeléctrico (PEF) en registros de pozo con machine learning

Method of oil and gas reservoir detection based on geological knowledge distillation learning

Shale Anisotropy Model Building Based on Deep Neural Networks

Contact Info

Product

Resources

About