Artificial Intelligence-based Approach for Predicting Mud Pump Failures

Faraz Feizi

Research output: ThesisMaster's Thesis

198 Downloads (Pure)


A significant component in the drilling operation is the circulation system. Drilling rigs have a crucial dependency on mud pumps, and a failure in the mud pumps will impose the drilling operation to stop completely; consequently, the drilling cost will increase due to the associated nonproductive time. Therefore, companies try to detect failures before occurring by implementing different techniques and strategies for improving pump operation time and efficient maintenance management to reduce or eliminate non-productive time, health, and environmental safety risks. Different tools and techniques that support the real-time monitoring of mud pumps have been proposed in the last decade; one of them is Artificial Intelligence (AI), which has shown promising results. Therefore, the ultimate goal of this thesis is to investigate the possibility of using artificial intelligence techniques to detect specific mud pump failures by utilizing only the pump pressure and flow rate as input features. This thesis is divided into three main parts. The first part of the thesis presents and discusses the general failure detection techniques and maintenance strategies. The second part of this work presents the common drilling mud pump failures and the impact of failures on drilling operation efficiency and HSE, and what are the state-of-the-art non-intrusive sensors that can be used to detect the pump failure signatures. The last part of the thesis elaborates on the steps of developing a conceptual approach based on artificial intelligence techniques to detect failures in drilling mud pumps. In order to validate and determine the limits of the developed tool, a case study was conducted using real historical data.
Translated title of the contributionKünstliche Intelligenz für die Prognose von Ausfällen von Schlammpumpen
Original languageEnglish
Awarding Institution
  • Montanuniversität
  • Elmgerbi, Asad, Co-Supervisor (internal)
  • Thonhauser, Gerhard, Supervisor (internal)
Award date20 Oct 2023
Publication statusPublished - 2023

Bibliographical note

no embargo


  • Mud Pump
  • Predictive Maintenance
  • Artificial Intelligence
  • Machine Learning
  • Failure and Fault Detection
  • Mud Pump Failures

Cite this