Publications

In recent development of digital transformation in the oil and gas industry, the need to implement intelligent closed-loop control at the field is increasing. Particularly in the upstream production operation where most of the activities still require manual intervention and regular field visits. With the deployment of an autonomous system, it is expected that field visit activity can be reduced, while oil production can be optimized. However, the adoption of such technology is still low due to operators’ resistance in trusting such solutions to run autonomously in the field. To have a high confidence that a particular Edge-AI solution will perform as expected in the field, a Hardware-In-the-Loop (HIL) testing methodology is proposed. In the experiment, an IoT edge gateway running a gas lift optimization algorithm is used as the device under test. All the required equipment is modeled and simulated using Python. In the final setup, the simulation computer and the device under test is connected physically as if it is running in the field. In the modeling phase, two scenarios are developed to simulate decreasing reservoir pressure and increasing water-cut over time. Then the model is run as the environment input for the algorithm running on the edge controller. The test result showed that the algorithm yields optimum gas injection rate over time, countering the decrease trend of oil rate. It also showed that the generated injection setpoint is always within the configured safe threshold (300 – 1000 Mcf/d). At the end, it can be concluded that the proposed testing methodology is flexible and can easily be adopted to test any other use cases that involves intelligent closed-loop control without the cost and risk involved if the testing is done directly in the field. By showing the test result, it is expected to increase the operator’s confidence in deploying intelligent closed-loop control solutions in their field.