Fault diagnosis of downhole drilling incidents using adaptive observers and statistical change detection

Anders Willersrud, Mogens Blanke, Lars Imsland, Alexey K. Pavlov

Research output: Contribution to journalJournal articleResearchpeer-review

769 Downloads (Pure)

Abstract

Downhole abnormal incidents during oil and gas drilling causes costly delays, any may also potentially lead to dangerous scenarios. Dierent incidents willcause changes to dierent parts of the physics of the process. Estimating thechanges in physical parameters, and correlating these with changes expectedfrom various defects, can be used to diagnose faults while in development.This paper shows how estimated friction parameters and ow rates can de-tect and isolate the type of incident, as well as isolating the position of adefect. Estimates are shown to be subjected to non-Gaussian,t-distributednoise, and a dedicated multivariate statistical change detection approach isused that detects and isolates faults by detecting simultaneous changes inestimated parameters and ow rates. The properties of the multivariate di-agnosis method are analyzed, and it is shown how detection and false alarmprobabilities are assessed and optimized using data-based learning to obtainthresholds for hypothesis testing. Data from a 1400 m horizontal ow loop isused to test the method, and successful diagnosis of the incidents drillstringwashout (pipe leakage), lost circulation, gas in ux, and drill bit plugging aredemonstrated.
Original languageEnglish
JournalJournal of Process Control
Volume30
Pages (from-to)90-103
ISSN0959-1524
DOIs
Publication statusPublished - 2015

Keywords

  • Managed pressure drilling
  • Fault diagnosis
  • Statistical change detection
  • Adaptive observer
  • Multivariate t -distribution
  • Generalized likelihood ratio test

Fingerprint Dive into the research topics of 'Fault diagnosis of downhole drilling incidents using adaptive observers and statistical change detection'. Together they form a unique fingerprint.

Cite this