An Unsupervised Machine-Learning Workflow for Outlier Detection and Log Editing With Prediction Uncertainty

Abstract

Recent advances in data science and machine learning (ML) have brought the benefits of these technologies closer to the mainstream of petrophysics. ML systems, where decisions and self-checks are made by carefully designed algorithms, in addition to executing typical tasks such as classification and regression, offer efficient and liberating solutions to the modern petrophysicist. The outline of such a system and its application in the form of a multilevel workflow to a 59-well multifield study are presented in this paper. The main objective of the workflow is to identify outliers in bulk density and compressional slowness logs and to reconstruct them using data-driven predictive models. A secondary objective of the project is to predict shear slowness in zones where such data do not exist. The system is fully automated, designed to optimize the use of all available data, and provide uncertainty estimates. It integrates modern concepts for outlier detection, predictive classification, and regression, as well as multidimensional scaling based on inter-well similarity. Benchmarking of ML results against those created by experienced petrophysicists shows that the ML workflow can provide high-quality answers that compare favorably to those produced by human experts. A second validation exercise, that compares acoustic impedance logs computed from ML answers to actual seismic data, provides further evidence for the accuracy of the ML-generated results. The ML system supports the petrophysicist by easing the burden on repetitive and burdensome quality control tasks. The efficiency gains and time savings created can be used for enhanced effective cross-discipline integration, collaboration, and further innovation.

Harnessing Advanced Machine-Learning Algorithms for Optimized Data Conditioning and Petrophysical Analysis of Heterogeneous, Thin Reservoirs

“…Petrophysics is a crucial discipline for determining the characteristics of reservoirs and developing new fields. Improved drilling efficiency, data repair, reservoir property prediction, reservoir rock type, and other petrophysics-related tasks have all benefited from the use of ML. , …”

“…Within the geoscience domain, the adoption of ML algorithms such as the tree-based approach dates to the early 1960s . Machine-learning applications within petrophysics include the prediction of reservoir properties, , outlier detection, prediction of missing acoustic slowness, and well log repair. ,, Within the petrophysical workflow, data quality assurance is an important step and can take up a significant amount of time within a project. Friedman and Smith identified that poor data quality leads up to a 20% reduction in labor productivity.…”

“…The impact of poor quality on the precision and accuracy of the prediction can be significant. Consequently, this can influence key decisions about the future of a well or a field. , However, several ML algorithms are capable of mitigating such issues , designed an unsupervised outlier detection algorithm that could identify the outliers in density and sonic logs and determine the footmark of a well from an arbitrary number of logs. A technique to predict missing logs by correlating a similar petrophysical log response from surrounding wells is proposed …”

Data-driven lithofacies prediction in complex tight sandstone reservoirs: a supervised workflow integrating clustering and classification models

“…Random forest is a popular machine-learning algorithm developed by Breiman (2001) and has been successfully used in various classification and regression problems (Akkurt et al 2018). The algorithm works by creating a large number of decision trees from bootstrap samples of the training data.…”