Monitoring a vast oil field without digital support presents a significant challenge, crucial for optimising Artificial Lift Systems (ALS). In the context of sucker rod pumping wells, real-time collection and analysis of dynamometer cards (dyna cards) are vital for understanding pump behaviour and overall system health.
Traditional manual collection of dyna cards twice a week across 47 horizontal wells is insufficient. Real-time dyna cards require at least 256 data points per minute for meaningful insights. Analysis of these cards optimises well production, enhancing pump and rod run life, preventing well downtime, and reducing production losses.
The integration of IoT, Cloud Computing, and Machine Learning shifted from reactive to proactive ALS optimisation. SeSuite Central, a cloud-based Decision Support System, facilitates data transmission. An algorithm automates dyna card classification, utilising computer-driven pattern recognition.
Machine learning libraries identify pump signatures, creating dashboards for quick ALS performance analysis. The system generates smart alarms based on statistical and machine learning settings, reducing well downtime by addressing issues proactively.
This approach, backed by digitalisation and domain knowledge, enables informed decision-making, efficiently managing over 47 wells remotely, overcoming resource limitations. The initiative significantly saves power consumption for low Productivity Index wells and prevents pump and rod failures, translating into substantial cost.
Teams involved in this project include:
- Digital Team: Atul Patni, Abhisek Roy, Himshella Sharma
- Business Team: Manish Kumar, Nakul Varma, Ravi Chandak, Sujit Jadhav, Amit Ranjan, Shailesh Chauhan, Joy Singhal, Avinash Bora.
- Partner Team: Manjunath Rao (Uthunga Technologies).
Impact of the Project:
~0.8 Mn $ Annual Savings
Well Downtime savings for 47 wells.
~100 BOPD per well per month saved.
Prevention of pump failures @ one per per month saves ~0.1 Mn USD annually).
