Cookie Notice

We use cookies to make parts of our website work, and to improve your visitor experience.
If you only allow necessary cookies, some features of our website may not work.

Technical Talk - AI in Hydraulic Fracturing: The journey from Guessing to Certainty

      Add to your calendar Last updated - 29/03/2023 08:53

Technical lecture
12 April 2023 20:00 - 21:00
This event has finished


How AI is used in the real world

Unplanned downtime caused by unexpected equipment failure on a hydraulic fracturing operation is one of the main reasons of Non-Productive Time (NPT), which has an effect on CAPEX and overall job efficiency. In order to increase overall work efficiency and decrease NPT, an AI-enabled digital twin was created for the fracking process that includes frac pumps, hydraulic operated valves, manifolds, and pressure control equipment.

In order to identify early warning indicators of failure patterns in pump cavitation, bearings, transmissions, suction and discharge valves, valve leaks, gate erosion, and incomplete movement, the AI-enabled digital twin combines field data with predictive AI models. Critical signature patterns are found using physics-based criteria and machine-learning models, which give an early warning of subsystems anomalies.

The real-time streaming equipment data, which includes valve positions, vibration, temperatures, process maximum and average pressures, is fed into these ML models. Operators receive real-time model results, enabling prompt intervention to avoid major failures.

The AI models were created to identify failure signatures based on training data using known failure events and tagging them on a time series data.The models have proven to be quite effective in assisting the operator in identifying process irregularities through real-time alerts on potential failures and consequently aid in early intervention to prolong the operational life of assets and prevent NPT.

Thiago G. Machado is chief product developer for surface digital, controls and automation product management with TechnipFMC in Houston. He has 15 years experience in the oil and gas industry, having worked in automation and digital with TechnipFMC in their subsea and surface business units. Machado holds a BS degree in electrical engineering from Severino Sombra University in Brazil.

  • The application of AI to complex processes in hostile environments
  • How the development of a digital twin is performed
  • How and why a digital twin can assist with failure prediction
  • How to prolong equipment life (and save considerable cost) using an AI digital twin

PEB- PBU applied for.

Supported by 


Thiago G. Machado Chief Product Developer for surface digital, controls and automation product management with TechnipFMC in Houston, USA.


Singapore Branch
Microsoft TEAMS

Contact Details

Andy Bell
Email: Send a message

Alternative contact LeeVeen Ng
Email: Send a message

Cart Shopping basket (0)

© 2023 Institution of Mechanical Engineers. IMechE is a registered charity in England and Wales number 206882