Intelligent Methods for Analyzing High-Frequency Production Data to Optimize Well Operation Modes

Abstract

The digital transformation implies a new approach to petroleum engineering. It based on the analysing of high-frequency data, automatization of business processes and the spread of artificial intelligence. Now modern wells are highly equipped, their operation is monitored by more than 50 sensor types (pressures and temperatures in different system units, electric telemetry data from submersible equipment and etc.). More than 10,000 measurements are accumulated daily for the well. Manual processing of such a large amount of information is impossible. A petroleum engineer usually analyses average values and only when there are problems does it necessary to check high-frequency data. Cause of high frequency measurements contain valuable information, the task of developing algorithms for the automatic analysis of large amounts of field data is relevant. In addition, oil reserves are decreasing, the physics of processes in the reservoir, well and surface facilities is becoming more complex. Highly productive oil reservoirs are being replaced with hard-to-recover reserves, oil and gas condensates and fringes. In those conditions it is highly relevant to apply advanced methods of information analysis and mathematical models. Methods of automatic analysis of high-frequency telemetry data are at the stage of active development and introduction into technological processes in petroleum industry [1]. The article presents the solutions of various problems of petroleum engineering by using advanced methods of data analysis and shows the tools that have allowed to achieve economic effects. The most important intraday data reviewing duties are quickly identification of down time, well mode regime optimization, preventing frow rate deviations from the planned one. The prompt decision allows to identify decrease of oil rate and cut back non-production expense.