Study On Drilling Formation Engineering Characteristic Based On Rock Physical Property

It is well-known that drilling is a chief way for oil&gas prospecting and exploitation. During the drilling process, bit meets and breaks various rocks, so the rock is the main working object for drilling. There are many different kinds of rock in the earth's crust, and they are anisotropic and heterogeneous. Therefore, it is an important research content to test rock performance parameters and study interrelation between the parameters. As a basic research work, the study and application of rock performance have significant sense for improving drilling technology.In this paper, the research objects are 668 pieces of cores which are selected from Shengli oilfield. Firstly, the measurement experiments of rock physical properties and drilling formation engineering characteristics are carried on and some datas are obtained. After the datas are analysed, the double-parameter predictive models of drilling formation engineering characteristics are formed based on rock porosity, permeability, density and acousticvelocity. The average predication accuracy of drilling formation engineering characteristic parameters is about 80% when using the predictive models. Lastly, the multi-parameter predictive models of drilling formation engineering characteristics are established with the application of BP neural network. The neural network predictive models can improve the predication accuracy of drilling formation engineering characteristic parameters in some degree.The research result of this paper indicates that there are definite correlations between the rock physical property and drilling formation engineering characteristic, and it's an effective approach to predict drilling formation engineering characteristic parameters with using rock physical property parameters.