Autonomous ESP gas handling solution expected to increase production by 400,000 bbl a year | SLB

Autonomous ESP gas handling solution expected to increase production by 400,000 bbl a year

minus
Ecuador

Combining edge computing with automation enabled an operator to optimize their ESP performance and production in brownfields in hours instead of days or weeks, with a digital ESP gas-handling workflow.

The gas coming from ESP wells that produce from high-GOR reservoirs has devastating effects on well productivity and ESP run life resulting from issues such as gas locking of the pump. Previously, operators managed these wells with regular site visits and manual operations. A field technician would visit the field to manually bleed off the casing head pressure and leave the annular casing valve cracked open until the next gas event occurs. Wells that needed intervention could take days to weeks to get back to the standard producing behavior.

Solution
An automated ESP gas-handling system, including wireless sensors, and a solar-powered skid with a flow computer and an automated choke valve to control the annular gas flow rate, was managed by the Agora™ edge AI and IoT solutions to optimize the well and ESP performance.

AI at the edge
Agora edge automatically captured tubing head pressure and casing head pressure from the wireless sensors, gas flow rate from the skid, and ESP data, and then synchronized the parameters prior to transmitting the data to the cloud. All this data is also consumed by a well model installed in the edge computing gateway, which accounts for the effect of the produced gas on ESP performance and autonomously sends commands to control the fully automated skid supplied by our partner, Sensia.

The edge application captures the gas rate measured by the flow computer and identifies the correct opening of the casing annular valve based on well-flow simulation models, with the ability to incorporate a physics-based model using the Eclipse™ industry-reference reservoir simulator and data-driven estimation using machine learning. The edge application sends commands to the automated skid to adjust the casing valve, preventing production halting gas-lock conditions.

This iterative process results in proper ESP optimization in hours instead of days or weeks, with no manual intervention required. By deploying the automated ESP gas-handling system to over 40 high-GOR wells, the operator expects to gain 400,000 bbl/y (1,095 bbl/d more per well) and reduce 14,000 km of driving in harsh environments within the next year. It also eliminates the intervention expense of sending a field technician to the field to investigate and remedy the persistent gas-lock issues.