Real-Time Data-Driven Detection of the Rock-Type Alteration During a Directional Drilling

Abstract

During the directional drilling, a bit may sometimes go to a nonproductive rock layer due to the gap about 20m between the bit and high-fidelity rock type sensors. The only way to detect the lithotype changes in time is the usage of Measurements While Drilling (MWD) data. However, there are no general mathematical modeling approaches that both well reconstruct the rock type based on MWD data and correspond to specifics of the oil and gas industry. In this article, we present a data-driven procedure that utilizes MWD data for quick detection of changes in rock type. We propose the approach that combines traditional machine learning based on the solution of the rock type classification problem with change detection procedures rarely used before in Oil&Gas industry. The data come from a newly developed oilfield in the north of western Siberia. The results suggest that we can detect a significant part of changes in rock type reducing the change detection delay from 20 to 1.8 meters and the number of false positive alarms from 43 to 6 per well.

InDiD: Instant Disorder Detection via Representation Learning

“…In [18] the method constructs anomaly score using expert knowledge and basic machine learning approaches. Then it aggregates obtained low-dimensional representations to get anomaly statistics similar to that in [22].…”

“…Anomaly detection in sequential data actively develops in a framework of machine learning and deep learning. Heuristic approaches include expert-based solutions, adaptations of CNN and GAN methodologies [12,21,8,18]. The majority of methods concentrate on detecting an anomaly in a sequence instead of an accurate anomaly moment detection [21,12].…”

Similarity learning for wells based on logging data

“…A typical solution for the problem is aggregation features on intervals using, e.g. mean and std Christ and Braun (2018); Gurina et al (2020); Romanenkova et al (2019). So, each interval of size (100, ) corresponds to the vector of features of size 2 , where is the number of utilized features.…”

“…Its combination with Gradient Boosting, a specific method for the training, allows overcoming many technical difficulties. For example, such models can easily handle data of various sample sizes and quality, automatically process missing data, learn quickly with a large number of features Kozlovskaia and Zaytsev (2017); Romanenkova et al (2019) and are suitable in different settings Ke et al (2017); Dorogush et al (2018). We use a gradient boosting realization XGBoost with default hyperparameters Chen et al (2015).…”

Usage of specific attention improves change point detection

“…The goal of Change Point Detection (CPD) is to find the moment of data distribution shift. Such tasks appear in different areas, from monitoring systems to video analysis [1] to oil&gas [11]. One of the recent methods for CPD [10] proves that a recurrent neural network model constructs meaningful representations and solve a problem better than non-principled approaches.…”