Instantaneous and reliable monitoring of the stream of gas, oil and water from a petroleum production wells requires accurate and robust multiphase flow meters that can survive the harsh conditions deep down in the well.
Fibre optical sound sensors are good candidates to be used as components in such flow meters, but their full potential is yet to be realized.
In this project we will address some of the challenges, such as:
- The development of physical models that describe how sound and acoustic signals are generated and transmitted in the well
- Efficient signal processing and removal of background noise
- Utilizing machine learning and artifical intelligence for interpretation and usage of sensor data in flow meter applications
