Pattern-Based Drilling Downhole Vibration Diagnostics

Physical Sciences : Petroleum

Available for licensing


  • Eric van Oort , Petroleum and Geosystems Engineering
  • Theresa Baumgartner , University of Texas at Austin
  • Pradeepkumar Ashok, Ph.D. , Petroleum and Geosystems Engineering

Background/unmet need

Drill string vibrations are frequent and persistent drilling performance limiters. Severe levels of vibrations might lead to damage of downhole tools, while even mild levels of vibrations can significantly slow down drilling progress. Pre-drilling vibration prevention methods can be altered using knowledge about vibration types and severity in similar drilling operations. Vibration mitigation methods during drilling inherently depend on the awareness of the presence of one or more vibration types and severity downhole. The actions that need to be taken by drilling engineers depend on the type of vibration. In order to efficiently prevent escalating vibration levels in a well, an automated method in which the type of vibration(s) present can be identified and analyzed is necessary for drilling engineers to perform appropriate procedures to prevent potential well and tool damage.

Invention Description

Researchers at The University of Texas at Austin have proposed a method that allows for the automatic differentiation of vibration types from patterns of downhole sensor measurements. The method consists of high-frequency sensor data in time-based windows, an algorithm for classifying vibration types, and the determination of vibration severity based on average and maximum vibration levels relative for each sensor type. A kinetic model has been developed that is capable of accurately reproducing patterns of position, velocity, and acceleration of a sensor within a drill string downhole, resulting in simulated visibility of the type of drilling dysfunction in real-time.

Previous methods for detecting and identifying drill string vibrations are based on the idea that a particular sensor measures a particular type of vibration by orienting the sensor on the string in a way to "sense" a particular motion, velocity, or acceleration of the drill string. The proposed invention is based on the understanding that any type of vibration influences measurements in all sensors regardless of their orientation. The automatic classification of vibration type and severity allows for the detection of whirl vibration patterns, which was previously achieved through the establishment of absolute thresholds of vibration levels from a particular vibration sensor. With the proposed method, vibration patterns can be determined from data obtained by sensors in any orientation.


  • Real-time analysis and observation allows for improved tailored mitigation actions
  • Reduces data loads and tool memory limitations, which saves time and cost for post-drilling data processing
  • Faster and more efficient drilling
  • Lower drilling costs
  • Increased production revenue


  • Automated drill string vibration mitigation
  • Real-time automatic detection and identification of drill string vibrations
  • Capable of unambiguously distinguishing vibration types
  • Kinematic model that visually reproduces vibration patterns
  • Automated analysis of drilling data
  • Creates visibility of the type of drilling dysfunction in real-time
  • Reduces data loads and tool memory limitations to gain understanding of vibrations throughout entire well
  • Processes large amounts of data in a post-drilling process for further analysis

Market potential/applications

Drill string vibrations can cause significant damage to the tools necessary for drilling procedures. Unmitigated vibrations give rise to large costs, constituting up to 10% of total well costs due to hindered efficiency, reduced wellbore quality, and increased need for maintenance. This invention could be critical to companies such as oil and gas service providers, as its implementation could significantly enable vibration reduction and prolong the lifespan of drilling equipment.

Development Stage

Proof of concept