Study On Logging Evaluation Method Of Glutenite Reservoir In Niudong Area,Northern Margin Of Qaidam Basin

The glutenite reservoir characteristics of Oligocene lower Ganchaigou Formation in Niudong area,northern margin of Qaidam Basin are complex.In order to realize the reservoir evaluation by logging data,firstly,the reservoir characteristics and the main controlling factors of reservoir physical properties are determined.Under the premise that conventional logging and simple linear algorithm can not meet the actual needs,conventional logging combined with support vector machine,nuclear magnetic logging combined with generalized regression neural network are designed to achieve the purpose of reservoir classification,and conventional and array acoustic logging combined with crossplot technology,conventional logging combined with multi-scale wavelet analysis technology to identify fluid properties.Finally,the above methods are applied to the actual data,and good results are achieved.The above logging evaluation methods and main conclusions are as follows:1.The reservoir characteristics are analyzed by using the statistical description method based on core thin section,X-ray diffraction,pore permeability analysis,capillary pressure and other experimental data,and the comparison results of lithology,grain size,cuttings,physical properties,structure and other reservoir characteristics of the upper and lower Ganchaigou Formation are obtained.The results show that:the lithology and structural characteristics of the reservoir have no obvious inheritance in the stratum,and the strata of the upper and lower members of lower Ganchaigou formation separated by faults have obvious similarity,which indicates that the reservoir characteristics are mainly controlled by the structure.The hanging wall is a mixed accumulation unequal grain reservoir with complex pore structure and complex pore-permeability,while the footwall is a low porosity low permeability reservoir with fine particles and contains secondary dissolved pores and fractures.2.Secondly,main controlling factors of reservoir physical properties are determined by cross plot technology.The main factors affecting reservoir physical properties are particle size,clay content,pore structure and carbonate content.3.In this paper,two methods are used to achieve reservoir classification evaluation.The first method is support vector machine technology to realize the classification in non coring well sections,the training sample is the conventional logging data of known capillary pressure curve classification.The sample return rate calculated by the support vector machine is 84.6%,and the prediction success rate is 100.00%;the second method is using generalized regression neural network(GRNN)to establish the reservoir classification model,and the input of neural network is the nuclear magnetic pseudo capillary pressure curve obtained by power function method.The accuracy of GRNN is 98.6%and 87.5%respectively.The two methods are applied to the actual data,and the classification results are consistent with those of mercury injection.4.For the main gas producing layer(footwall E₃¹)in Niudong area,two fluid identification methods are used in this paper.The first method is cross plot method:density resistivity cross plot can separate the dry layer with poor physical properties,the gas bearing index resistivity cross plot can distinguish the gas and water dry layers,and the petrophysical parameter chart obtained by array acoustic logging can identify the high-yield gas producing layer.The second method is wavelet analysis technology to analyze the product value of resistivity and porosity,and the characteristic database of fluid properties on energy and scale are established.The actual well data processing results show that the wavelet multi-scale energy weighted cumulative value can effectively identify the gas reservoir.The well logging reservoir evaluation technology formed in this paper can effectively classify the reservoir and identify the fluid properties of the complex glutenite of lower Ganchaigou formation in Niudong area.