Copyright and instructions for use: the copyright of this translation belongs to SPWLA Southwest Chapter and is only for academic education purpose. All rights reserved to SPWLA. Please contact us if you find any mistakes in the translation.
A Machine-Learning-Based Approach to Assistive Well-Log Correlation Seth Brazell, Alex Bayeh, Michael Ashby, Darrin Burton. Petrophysics, 60(4): 469 – 479, DOI: SPWLA-2019-v60n4a1. https://www.onepetro.org/journal-paper/SPWLA-2019-v60n4a1
The process of well-log correlation requires significant time and expertise from the interpreter, is often subjective and can be a bottleneck to many subsurface characterization workflows. Algorithmic approaches to well-to-well correlation suffer from the inherent heterogeneity of geophysical measurements in the wellbore, both from a geologic and data-quality perspective.We demonstrate a rigorous and repeatable method for well-log correlation by deploying a correlation tool that leverages a machine learning model for pattern matching between well logs and programmed stratigraphic correlation techniques. A supervised-learning approach was used to train a novel deep convolutional neural network (CNN) architecture using over five million data samples, which were derived from thousands of well logs and expert interpreted correlations. To ensure that a robust pattern-matching model was trained, well logs from several US onshore basins with various tectonic regimes and environments of deposition were used to construct training and validation datasets. The result is a universal model for pattern matching of wireline measurements that can incorporate multiple geophysical-log signals as input data and can be deployed at scale without the need for retraining. Overall, the pattern-matching model was able to achieve a level of accuracy of 96.6% and classification area-under-the curve (AUC) of 0.954 on a separate validation dataset.
The universal deep CNN is one component of the correlation tool. Algorithmic three-dimensional search logic was constructed around the deep CNN model which determines the optimal correlation and marker propagation pathway. Rules-based criteria have also been applied to the model output ensuring conformance to stratigraphic principles including preserving stratigraphic order and honoring present-day structural trends. We present several examples to highlight the strengths and weaknesses of this machine-learning-based approach to well-log correlation which can be used to efficiently generate high-density datasets for regional exploration, development mapping and reservoir characterization exercises.
测井相关性分析需要大量的时间和解释人员的专业知识，且通常具有主观性，成为了井筒地球物理表征工作流程的瓶颈。从地质和数据质量的角度，井间相关性分析算法受井筒地球物理测量固有非均质性的影响。我们通过使用一种机器学习的模型，将配套的测井解释成果和地层相互对比，证明了一种严格且可重复的相关性分析方法。使用监督——学习方法训练这种新的深度卷积神经网络（CNN）结构，该结构使用超过 500 万个数据样本。这些样本均来自于数千口测井解释专家的解释结果。为了确保能够训练出一个稳定可靠的匹配模型，构建训练和验证数据集的样本来自美国多个不同构造形态和沉积环境的陆上盆地。其训练结果为一个普适性的电测数据匹配模式，该模式可以将多个地球物理测井信号作为输入数据，并在不重复训练的情况下将该数据加载至模型中。总体而言，该匹配模型能够在独立的验证数据集上达到 96.6%的准确度和 0.954 的曲线下分类面积。
Total Organic Carbon Characterization Using Neural-Network Analysis of XRF Data Lateef Owolabi Lawal, Mohamed Mahmoud, Olalekan Saheed Alade, and Abdulazeez Abdulraheem. Petrophysics, 60(4): 480 – 493, DOI:SPWLA-2019-v60n4a2. www.onepetro.org/journal-paper/SPWLA-2019-v60n4a
The co-occurrence of kerogen and nonkerogen minerals in shale poses a great challenge; most importantly, different scales of measurements and ranges of analytical instruments become the prerequisite for characterization. The traditional shale-characterization technique adopts mineralogical analysis for the inorganic constituent and the total organic carbon (TOC) for the organic matter (kerogen). However, despite modern laboratory analytical techniques, the direct and simultaneous determination of the organic and inorganic constituents of a shale formation may be costly and unrealistic.
Hence, the use of the cost effective and more efficient X-ray fluorescence (XRF) tools for the elemental characterization of shale. The missing TOC problem, typical of well logging analysis, albeit, is a major challenge of this method. The objective of this work is to carry out quantitative analysis and interpretation of geochemical and mineralogical composition for the evaluation of organic rich shale formations using a neural network (NN) with the primary interest to optimize the performance of the XRF technique. For this purpose, a machine-learning artificial neural network (ANN) method has been devised to map easy-to-measure nondestructive XRF data of organic-rich shale to total organic carbon. Subsequently, the developed model, based on the existing dataset, was used to predict missing TOC. In the data-driven ANN model, 70% of the dataset was used for training, 15% for validation and 15% for testing. The accuracy and improvement of the NN model was established based on statistical parameters as performance metric. Furthermore, a quantitative calibration function to map the XRF data to TOC is developed based on the extraction of weights and biases of the NN. The implementation of the proposed calibration function for the calculation of TOC in comparison to the measured TOC was obtained with the coefficient of determination (R2)and mean absolute percentage error (MAPE) of 0.974 and 14.54%, respectively. These results confirm the applicability of the proposed calibration function in facilitating the simultaneous measurement of TOC and elemental concentration of shale for both laboratory and field-scale applications.
利用神经网络分析 X 射线荧光数据的有机碳含量表征
页岩中干酪根和非干酪根矿物的共存给页岩的表征带来巨大的挑战；其关键在于，测量尺度与分析仪器的范围是有效表征的前提。传统的页岩表征技术采用无机矿物分析测量非有机质组分、采用有机碳含量表征有机质（干酪根）。然而，尽管现代实验室分析技术能够测定页岩组分中的有机和无机组分，但费用昂贵、不易大规模推广。因此，使用成本较低和效率更高的 X 射线荧光数据对页岩进行元素表征。然而，常规测井分析中 TOC 的缺失问题是该方法面临的主要挑战。本研究的目的在于利用神经网络（NN）对富含有机质页岩地层进行地球化学和矿物学成分的定量分析和解释，以优化XRF 技术的性能。基于这样的目的，设计了一种机器学习人工神经网络（ANN）方法，将富有机页岩中易测、无损的 XRF 数据映射到总有机碳。随后，基于现有数据集开发的模型被用于预测损失的 TOC 含量。在数据驱动的人工神经网络模型中，70%的数据集用于培训，15%用于验证，15%用于测试。以统计参数作为性能指标，建立了神经网络模型的精度和改进方法。此外，在提取神经网络权值和偏差的基础上，建立了将 xrf 数据映射到 TOC 的定量标定函数。计算总有机碳的校准函数与实际测量总有机碳含量相交会，相关系数（r 2 ）和平均绝对百分比误差（mape）分别为 0.974%和14.54%。该结果验证了校准模型实际应用的可行性，能够在实验室和现场实现页岩总有机碳和各类元素的测量。
The Compressibility Factor (Z) of Shale Gas at the Core Scale Huy Tran, A. Sakhaee-Pour. Petrophysics, 60(4): 494 – 506, DOI: SPWLA-2019-v60n4a3. www.onepetro.org/journal-paper/SPWLA-2019-v60n4a3
The compressibility factor (Z) of a gas inside a nanosize conduit depends on the conduit‘s characteristic size, in contrast to wide conduits whose dimensions have no effect on the gas compressibility. Nanofluidics, which is a field of study concerned with the fluid flow in nanosize conduits, can quantify the gas compressibility factor in a simple topology, such as a uniform tube with a circular cross section, but it is not apparent how those results are relevant to a complex pore space in the matrix of a shale at the core scale. This study determines the compressibility factor of a shale gas by accounting for the effective connectivity of the pore space at the core scale. We use effective pore-throat and pore-body sizes, which are interpreted using an acyclic pore model applied to the core-scale measurements and not high-resolution images. Eleven shale formations whose data are available in the literature are investigated (Bakken, Barnett, Eagle Ford, Haynesville, Marcellus, Monterey, New Albany, Niobrara, Utica, Wolfcamp, and Woodford). The results, which have applications in developing realistic models based on petrophysical measurements, show the compressibility factor (Z) of the shale formation at the core scale as a function of gas pressure.
纳米尺度空间中页岩气压缩因子（Z）取决于通道的尺寸，而尺寸较大孔隙空间对气体的压缩性无影响。纳米流体学专指针对纳米级空间中流体的流动进行研究，可以在简单的拓扑结构中量化气体压缩因子，例如具有圆形横截面的均匀孔隙。但是，在岩心尺度下，上述研究结果与页岩基质中复杂孔隙空间之间的相关性并不明显。该研究通过计算岩心尺度下孔隙空间的有效连通性确定页岩气的压缩因子。我们使用有效的孔喉、孔隙体尺寸，该结果来源于岩心尺度测量不可循环模型，而非高分辨率的图像。11 个页岩地层资料（Bakken、Barnett、Eagle Ford、Haynesville、Marcellus、Monterey、New Albany、Niobrara、Utica、Wolfcamp 和 Woodford）参与研究。研究结果表明，岩心尺度下页岩地层的压缩因子（Z）与气体压力呈函数关系，可用于建立基于岩石物理实验的实际模型。
Practical Approach to Derive Wettability Index by NMR in Core Analysis Experiments Wim Looyestijn. Petrophysics, 60(4): 507 – 513, DOI: SPWLA-2019-v60n4a4. www.onepetro.org/journal-paper/SPWLA-2019-v60n4a4
Wettability is a crucial factor for the dynamic properties of oil reservoirs. Early recognition of the wettability condition of a recovery may have a significant impact on the development options and of the expected recovery factor. NMR relaxation times of pore fluids are dependent on the wetting through surface relaxation, and are thus known to contain this valuable information. This paper describes an easy-to-implement and reliable procedure to calculate a quantitative wettability index from standard NMR measurements, such as can be made in conjunction with SCAL experiments. The interpretation is fully auditable, and therefore suitable to be part of a standard protocol.
Please refer to the document for more details.