Study On Safety Evaluation Methods Of Oil And Gas Drilling Engineering Based On Hierarchy Structure Model

Drilling engineering is a direct approach to develop underground fluid resources,and it is also a systemic engineering with long period,complex technology and various processes.In the process of drilling,many complex situations may affect the safety progress of drilling,and then affect the exploration and development of underground fluid resources.Therefore,the evaluation of drilling safety is of great significance to ensure drilling safety and esource development.On the basis of intensive analysis on drilling safety evaluation methods,this paper proposes a hierarchy structure model for quantitative evaluation of drilling safety.Based on the actual data from drilling site,the model integrates the analytic hierarchy process,membership function and nonlinear regression model,which can comprehensively consider various factors affecting drilling safety and make quantitative evaluation on drilling safety.Firstly,structure and characteristics of the hierarchy structure model are introduced.Then,a comprehensive evaluation system of drilling safety is built,through investigating and analyzing of drilling engineering data.Finally,the method and process for establishing the hierarchy structure model is presented,the application of the model and is analyzed.The results show that three-layer index system can be used to characterize the influence of field factors on drilling safety.The first layer index contains direct observation factors from drilling site.The second layer index is consist of four potential variable factors,including natural,technical,hominal and economic factors The third layer is corresponding to the drilling safety level.The hierarchy structure model contains two models,the first layer model is related direct observation variables and potential variables,and the second layer model is related to potential variables and drilling safety level;Membership function of the first layer model can be represented by different empirical function model combinations.Nonlinear regression model can be used to represent the second layer model.With the optimization analysis,the exchange condition mathematical expectation model(ACE)is due to the artificial neural network model(ANN)and support vector machine model(SVM).Case analysis shows that the result of prediction of hierarchy structure model is consistent with the actual security situation,and the prediction effect is accepted.