A visual analytics approach to anomaly detection in hydrocarbon reservoir time series data

Soriano-Vargas, Aurea; Werneck, Rafael de Oliveira; Moura, Renato; Júnior, Pedro Mendes; Prates, Raphael; Castro, Manuel; Gonçalves, Maiara Moreira; Hossain, M.M.; Zampieri, Marcelo Ferreira; Ferreira, André; Davólio, Alessandra; Hamann, Bernd; Schiozer, Denis José; Rocha, Armando Freitas da

doi:10.1016/j.petrol.2021.108988

Cited by 17 publications

(1 citation statement)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…There is a vast literature on algorithms for analyzing time series data [1]. In particular, time series data plays a vital role in the petroleum industry in tasks such as oil forecasting [19] and anomaly detection in reservoir data [22], among others. The challenge that surfaces in DTs is coordinating access to an extensive collection of time series data corresponding to several types of physical properties placed at different locations and labeled with different tags.…”

Section: Visual Analytics Requirementsmentioning

confidence: 99%

Ontology Explorer: An Ontology-Based Visual Analytics System for Exploring Time Series Data in Oil and Gas

Santos,

Rivera,

Petry

et al. 2023

Frontiers in Artificial Intelligence and Applications

View full text Add to dashboard Cite

Data analytics is the best approach for extracting hidden patterns and trends from time series data. One strong limitation that restrains the use of the method is the difficulty in selecting the appropriate data when a large variety of records come from several providers. Domain ontologies can organize and offer a uniform view for data analysis without plastering the data in a rigid format. This work describes an innovative visualization platform for data analysis of petroleum production time series that allows the user to explore the semantics of the oil well data with the support of a well-founded domain ontology. The O3PO ontology represents a production plant’s installation assets, describing the equipment’s relevant properties, such as position, relationship with other assets, and sensor measurements. The visualization platform takes advantage of the ontology to assist the user in locating the installation component, equipment, and collected measurements, bringing the time series data to the analytic tool for analysis by various groups. The ontology provides a taxonomy for navigating between classes and subsequently selecting instances, components, and properties. We describe the use of the visualization tool in real-case data from an offshore oilfield in Brazilian Pre-salt. This work contributes to the conception of the next generation of digital twins for the oil and gas industry.

show abstract

Section: Visual Analytics Requirementsmentioning

confidence: 99%

Ontology Explorer: An Ontology-Based Visual Analytics System for Exploring Time Series Data in Oil and Gas

Santos,

Rivera,

Petry

et al. 2023

Frontiers in Artificial Intelligence and Applications

View full text Add to dashboard Cite

show abstract

Anomaly detection in oil-producing wells: a comparative study of one-class classifiers in a multivariate time series dataset

Fernandes,

Komati,

Assis de Souza Gazolli

2023

J Petrol Explor Prod Technol

View full text Add to dashboard Cite

Anomalies in oil-producing wells can have detrimental financial implications, leading to production disruptions and increased maintenance costs. Machine learning techniques offer a promising solution for detecting and preventing such anomalies, minimizing these disruptions and expenses. In this study, we focused on detecting faults in naturally flowing offshore oil and subsea gas-producing wells, utilizing the publicly available 3W dataset comprising multivariate time series data. We conducted a comparison of different anomaly detection methods, specifically one-class classifiers, including Isolation Forest, One-class Support Vector Machine (OCSVM), Local Outlier Factor (LOF), Elliptical Envelope, and Autoencoder with feedforward and LSTM architectures. Our evaluation encompassed two variations: one with feature extraction and the other without, each assessed in both simulated and real data scenarios. Across all scenarios, the LOF classifier consistently outperformed its counterparts. In real instances, the LOF classifier achieved an F1-measure of 87.0% with feature extraction and 85.9% without. In simulated instances, the LOF classifier demonstrated superior performance, attaining F1 measures of 91.5% with feature extraction and 92.0% without. These results show an improvement over the benchmark established by the 3W dataset. Considering the more challenging nature of real data, the inclusion of feature extraction is recommended to improve the effectiveness of anomaly detection in offshore wells. The superior performance of the LOF classifier suggests that the boundaries of normal cases as a single class may be ill-defined, with normal cases better represented by multiple clusters. The statistical analysis conducted further reinforces the reliability and robustness of these findings, instilling confidence in their generalizability to a larger population. The utilization of individual classifiers per instance allows for tailored hyperparameter configurations, accommodating the specific characteristics of each offshore well.

show abstract

Quantifying the Influence of Climate Variables on Vegetation Through Remote Sensing and Multi-dimensional Data Analysis

Dastour,

Hassan

2024

Earth Syst Environ

View full text Add to dashboard Cite

A visual analytics approach to anomaly detection in hydrocarbon reservoir time series data

Cited by 17 publications

References 13 publications

Ontology Explorer: An Ontology-Based Visual Analytics System for Exploring Time Series Data in Oil and Gas

Ontology Explorer: An Ontology-Based Visual Analytics System for Exploring Time Series Data in Oil and Gas

Anomaly detection in oil-producing wells: a comparative study of one-class classifiers in a multivariate time series dataset

Quantifying the Influence of Climate Variables on Vegetation Through Remote Sensing and Multi-dimensional Data Analysis

Contact Info

Product

Resources

About