PlumX Metrics
Embed PlumX Metrics

Enhancing Rate of Penetration Prediction in Drilling Operations: A Data Stream Framework Approach

SSRN, ISSN: 1556-5068
2024
  • 0
    Citations
  • 109
    Usage
  • 0
    Captures
  • 0
    Mentions
  • 0
    Social Media
Metric Options:   Counts1 Year3 Year

Metrics Details

  • Usage
    109

Article Description

In the oil and gas industry, drilling operations serve as a cornerstone in the quest for sustainable hydrocarbon production. Among the myriad factors shaping drilling success, the Rate of Penetration (ROP) through subsurface rock formations emerges as a pivotal parameter. Precise prediction and optimization of ROP assume heightened significance, given their profound impact on cost efficiency, safety protocols, and the overarching aim of minimizing environmental footprints. Predictive machine learning algorithms utilize data transformations to capture intricate patterns, offering flexibility in optimizing drilling operations. While machine learning algorithms consistently yield favorable results, optimizing ROP prediction warrants careful consideration of ROP data dynamics, including concept drift—a phenomenon wherein the underlying distribution of ROP data changes over time. We advocate addressing ROP prediction within the data stream framework to accommodate high-speed data, frequent model updating, and effective handling of concept drift. Our comparative analysis between traditional machine learning methods and online learning algorithms demonstrates a notable improvement. Employing suitable data stream algorithms led to a significant 63% reduction in absolute error rate and a substantial 77% decrease in processing time, on average.

Bibliographic Details

João Roberto Bertini Junior; Bahram Lavi

Elsevier BV

Multidisciplinary; Rate of penetration prediction; Data-driven modeling; Data streams; Drilling Optimization; Concept drift

Provide Feedback

Have ideas for a new metric? Would you like to see something else here?Let us know