Decision Support System for Drilling Process Optimization

Wolfgang Mathis

Research output: ThesisDoctoral Thesis


This work introduces a decision support system for oil and gas-well drilling process optimization. The time spent to drill a well is typically longer than what is technically possible due to down times. Identified lost time consists of incidents which are documented in daily drilling reports. The invisible lost time is defined as time losses, when operations are not running at the technical limit. Unfortunately, human-reported operations are not exact enough, to identify the invisible lost time. To overcome these limitations, an expert system is applied to derive a highly detailed process description, which is used to replace the coarse human-reported operation breakdown. The necessary input parameters are sensor measurements regularly recorded for more than a decade. This provides a huge repository to derive benchmarks. To enable a timely and successful decision making process, a 5-step data management strategy is introduced. The system is highly automated, to reduce human interference as much as possible. Prior to the operation detection, special attention is given to data acquisition and data quality control. An expert systems is applied to replace human observations and Online Analytical Processing (OLAP) supports the data interpretation. Furthermore, data visualization techniques are discussed, to support the human analyst in his decision making process. Finally, practical examples are presented, which prove the efficiency of the introduced system.
Translated title of the contributionEntscheidungsunterstützungssystem für die Prozess-Optimierung von Kohlenwasserstoff-Bohrungen
Original languageEnglish
  • O'Leary, Paul, Assessor B (internal)
  • Thonhauser, Gerhard, Assessor A (internal)
Publication statusPublished - 2007

Bibliographical note

embargoed until null


  • Drilling Process Optimization Event Operation Recognition Expert System Artificial Intelligence Quality control Visualization

Cite this