Multi-point Statistics In Facies Prediction Application

Mufti-point statistics is a very important research rising in geological field in recent years.In geological facies forecast, based on two points of the variation functions statistics predict is not high, so it is difficult to meet the needs of practical application, in this environment, some scholars have suggested multi-point statistics, it can be integrated all the prior information instead of using just two point or several points prior information, this greatly improving the geological facies prediction accuracy.In recent years, with the booming of social science technology and world economy,the geological industry as an important areas of the relationship between people’s daily life has become one of the focus of attention, with today’s demand for oil is increasing, how to have better information to understand the oil extraction process must be studied, and further study of the relationship between the line of porosity, permeability and saturation is the problems to be solved. In response to these issues, this paper analyze the traditional two points statistics variation functions deficiencies, put forward mufti-point statistics on this base,and to simulate the key topic of the twelve countries xinjiang oil field "sand66well", accuracy rate of more than80%,reflect the value of multi-point statistics in the rock like a prediction. This paper aims for master data, analog effects and realistic effects contrast method to predict geological facies, this paper describes the first principle of the traditional two-point statistics, variation functions and multi-point statistics, then simulate the66wells on the sand, finally, to show the results of the study.