| Popis: |
The rise of automation technologies has enabled storage and sensing of operational data in real-time for key variables in oil production wells. These data are a key component in decision-making in the operation and control of the production of the field, if they can be converted to timely and reliable information. This paper focuses on the early detection and diagnosis in real time of operating events mainly by the presence of gas and solids that negatively impact the well production. Expert systems techniques for the characterization of reference operational events are used. They are based on known rules of behavior of electrical and mechanical operating variables, constructed by combining knowledge of the phenomenon and empirical data. The matching of behavior of real-time operational variables processed by techniques of statistical process control - SPC together with reference events determine the presence of operational problems by generating real-time alerts that allow it to act in a timely way and mitigate the negative impact on the production of hydrocarbons from the oil fields. Early detection algorithms and real-time diagnostic operational events were programmed in a software tool for real-time processing, which was applied to the monitoring of two case studies in wells Castilla and Guatiquia operated by Ecopetrol SA in Colombia. Finally the main findings, conclusions and recommendations from this study are submitted. |
