Application of neural network approaches in creating a monitoring system of technological parameters of wells operated by steam assisted gravity drainage

Oleg V. Denisov “TatASU” LLC Almetyevsk, Russian Federation
Ruslan G. Girfanov “TatASU” LLC Almetyevsk, Russian Federation
Bulat F. Zakiyev Field Office “Almetievneft” Almetyevsk, Russian Federation
Arslan V. Nasybullin TatNIPIneft” institute Bugulma, Russian Federation
In this article were presented results of the application of neural network approach in the creation of a monitoring system of technological parameters of wells operated by steamassisted gravity drainage. Showcased visual representation of design criteria deviations.
Materials and methods
On the basis of the using of bi-directional Kohonen self-organizing maps is built models of mutual behavior of process parameters. Derived models are used to detect deviations from the regular operation of process control and / or oil facilities.
It was implemented the system ofprocess parameters monitoring based on the using of neural network approaches that can detect unintended operationof wells operated by cyclic steam stimulation.
The proposed method of analysis of the information was tested and shown its effectiveness for the detection of wells that require immediate attention and adjustment of operating parameters. Timely responseto such situations will prevent accidentsand the cost of downtime, as well as to optimize the processes for more effective oil production. The proposed approachto the analysis of telemetry data also can significantly reduce the cost of the working time of technological personnel to monitor the current oilfield objects.
Download article
control of technological parameters Kohonen sel