Research Of Low Pow Production Solutions Based On Grey Correlation Analysis

In the past centry, the oil industry has promoted the development of global economy, andbecome a strategic issue restricting the economic development of countries of the worldbecause of its limited resources. Our country’s oil industry goes through nearly50years ofdevelopment, and most of the wells have entered the development of middle and late period.The reservoir effective thickness of the oil-bearing gets thin, and reserves abundance aregetting lower and lower, leading to the increase in number of low-yielding wells year by year.With the advance of enterprise information technology, the oil production companies havebuilt up a variety of database in the service of the well development, making the productionprogram of using data mining technology to research low-yielding wells possible.In order to reduce production energy consumption in this paper,the low-yielding wellsproduction program based on the data mining technology is researched. Combining the VC++,data mining technology and the oil well production plan,the three layer architecture of thewell data mining system design and development are put forward, realizing the constructionof the low-yielding well production management system platform.On the basis of the analysisof the data of53wells in a oil production area, for low-yielding wells with low crude oilfreezing point and low wax content, mathematical statistics methods are used to formulatewells’ open off work time namely adopting the intermittent pumping scheme. For wells withhigh wax content or not suitable for intermittent work, the continuous pumping scheme isdesigned, namely according to the impact for minimum standard, using the grey correlationanalysis-maximum entropy algorithm to analyze oil well production parameters, mining theconforming value of flush times. The pumping unit can adjust pumping speed without stopingwork, which does not affect the crude oil production and can extend the service life of suckerrod, and reduce the production cost at the same time.Low-yielding well production management system database is designed by SQL2000. The original oil well production data, the derivation of data during the data processing and thefinal results of the analysis are stored in the database. With the aid of VC++6.0platform, theconnection with the database, the data mining algorithm, and the low-yielding oil wellproduction management system interface design are realized.Users can understand thedifferent well conditions of oil wells by suitable production plan through a simple operation,which provides a reliable basis for improving the production efficiency.