Baker Hughes has developed a multiphase flowmeter that combines downhole sensor technology with neural network capabilities to provide real-time, affordable flow rate estimates from wells produced with electrical submersible pumping (ESP) systems. When properly calibrated and maintained, the Neuraflow™ multiphase flowmeter delivers flow rate estimates with a greater than 90 percent accuracy rate when compared to actual flow rate measurements.

Flow measurements are critical for optimizing reservoir performance and ESP system efficiency as well as diagnosing well and ESP system issues. However, direct measurement of multiphase flow rates is expensive and, therefore, typically performed only intermittently. Neuraflow, which is installed in the ESP surface variable speed drive (VSD), offers an affordable alternative to conventional direct flow monitoring techniques via real-time flow rate estimates. The Neuraflow device uses existing standard downhole measurements coupled with additional system measurements and learning algorithms to infer a flow rate based on known reservoir and fluid properties. This cost-effective solution measures flow at four separate downhole locations. The four flow measurements are compared and then used to calibrate the model for a high degree of accuracy.

"This is an important addition to our existing production measurement portfolio. Neuraflow meets a long-held customer need for accurate, affordable flow measurements," notes Neil Harrop, president of Baker Hughes' completions and production business segment. "This complements our expanding digital oilfield capability with real-time measurements that can make a significant difference in optimizing production."

Tags: Baker Hughes Incorporated