How Cognite Data Fusion enabled effective slug prediction

Aker BP’s oil fields, Alvheim and Volund, have experienced this typical trouble with slugging. Several factors can influence slugging and some of these are transient such as, opening wells or re-routing wells. This can lead to the restriction of production potential for third-party tie-backs.

One major challenge was to reproduce these issues with physical models in real time, as two seemingly similar conditions can result in different outcomes. Aker BP needed an effective way to monitor and prevent potential slugging.

Integrating with Solution Seeker, Cognite’s Asset Data Insight enables the complex analytical process required for slug prediction. Through Cognite Data Fusion, Aker BP could access thousands of live and historical time series, continuously analyzing them for pattern recognition and statistics.

This aggregated data is fed into machine learning models with self-learning algorithms that identify field behavior and correlation for an automatic predictive model generation. Finally, Aker BP was able to develop optimization models that can be auto-updated with data. They deliver real-time, actionable insight to production engineers, supporting production maximization.

Operators now have access to user-friendly decision support and algorithms that deliver early warning signs of imminent slugging. The Alvheim oil field saw a 1% increase in production, thanks to improved slug handling and prediction capabilities.