Non-Intrusive Monitoring of Subsea Boosting Systems
- Jens Abrahamsen (BHGE) | Frank Tore Sæther (BHGE) | Svein Arild Haugen (BHGE) | Stian Bentsen (BHGE) | Yngve Villanger (BHGE) | Ole Økland (Equinor)
- Document ID
- Offshore Technology Conference
- Offshore Technology Conference, 6-9 May, Houston, Texas
- Publication Date
- Document Type
- Conference Paper
- 2019. Offshore Technology Conference
- 7 Management and Information, 7.2.1 Risk, Uncertainty and Risk Assessment, 4.5 Offshore Facilities and Subsea Systems, 4.5 Offshore Facilities and Subsea Systems, 7.2 Risk Management and Decision-Making
- Condition Monitoring, Subsea Processing, Subsea Boosting, Sensor technology, Machine Learning
- 1 in the last 30 days
- 194 since 2007
- Show more detail
- View rights & permissions
Critical rotating machinery topside is monitored by sensors providing detailed information on the mechanical condition. Early signs of abnormalities can be detected and corrected. This is contrary to subsea pumps where only a very small percentage of the installed base is monitored. This is due to the lack of availability of qualified proximity probes for subsea pumps, although this has started to be addressed for recent applications. Furthermore, sensors inside the machine add complexity, failure modes and cost, while benefits are limited as repair requires the machine to be recovered to the surface. Consequently, a run to failure philosophy is applied.
With sensors being omitted because of cost and risk, the operator often operates in an uncomfortable "blind" mode. With no information on remaining life time and the risk of a creeping failure underway, the only mitigation is to accept the cost for a spare pump and IMR vessel ready for a replacement at minimum notice.
By applying an array of sensing elements, the presented condition monitoring system measures and analyses the acoustic and electric field generated by a subsea pump. The sensor system operates completely independently, can be installed and replaced without interfering with the pump and will also provide monitoring of components adjacent to the pump. The level of precision provided is not designed for detailing a repair plan (e.g. defining the position of balance weights), however it is fully adequate for early detection of a fault under development and will help guide the operator to maximize remaining lifetime by tuning the process parameters.
Using data acquired over a 10 year period, we will demonstrate how the system has successfully been used to monitor the health status of two different subsea pumps. Both pumps were eventually damaged, and we will investigate how the system provided the operator with a warning 6-8 weeks before the pump had to be stopped.
|File Size||2 MB||Number of Pages||18|