Study On The Evaluation And Application Of Pore Structure In Tight Sandstone Reservoir

With the advancement of oil and gas exploration,tight sandstone reservoir has become the focus of exploration in the field of petroleum industry.The characteristics of low porosity and permeability,strong reservoir heterogeneity in tight sandstone reservoir lead to the challenges in quantitative evaluation of pore structure,calculation of reservoir parameters,and comprehensive classification of reservoir.The research on the method and application of pore structure evaluation of tight sandstone reservoir is of great significance for the division of reservoir “sweet spots” and the prediction of productivity.In this paper,pore structure evaluation,reservoir parameters evaluation,and comprehensive classification of tight sandstone reservoir are studied based on the Triassic Chang 8 Formation in Ordos Basin in China.Firstly,based on nitrogen gas adsorption(N2GA)experiments,mercury injection capillary pressure(MICP)experiments,X-ray diffraction(XRD)experiments,particle size analysis experiments,thin section analysis and scanning electron microscopy(SEM)experiments,the pore structure and fractal characteristics of tight sandstone are investigated,the effects of mineral components and diagenesis types on the physical properties and pore structure of rocks are analyzed.The monofractal and multifractal characteristics of the pore size distribution curves obtained from N2GA(PSDN2)and MICP(PSDCP)are analyzed,it is concluded that,compared with PSDN2 curves,PSDCP curves are more complex and have a wider range;the fractal characteristics of both are influenced by the information of coagulation adsorption pores and small pores respectively;when evaluating the pore structure of tight sandstone using different pore size distribution curves,the fractal characteristic parameters of different pores should be calculated.The physical properties and pore structure characteristics of rocks under different mineral components and diagenesis types are compared and analyzed,the results show that,the increase of clay mineral content will intensify the complexity degree of pore space of tight sandstone;constructive diagenesis and destructive diagenesis lead to the improvement and deterioration of the physical properties and pore structure characteristics of rocks respectively.Nuclear magnetic resonance(NMR)experiments have the advantages of fast measurement speed and nondestructive analysis compared with other core experiment experimental methods,and NMR logging can continuously obtain the pore structure information.In this paper,we calculate the fractal dimension of NMR echo data(Decho)and predict the capillary pressure(CP)curves for evaluating the pore structure of tight sandstone reservoir.Fast Fourier transform is performed with the NMR echo data and then the Decho is calculated.The Decho and the fractal dimension of the NMR T2 distribution are compared,it is concluded that: the calculation of Decho is more stable and Decho can better evaluate the pore structure of tight sandstone reservoir.In addition,two methods for predicting the CP curves using NMR data are proposed: the prediction method based on classified piecewise multi-parameter power function transformation(CPMPFT)model and the prediction method based on echo data characteristic parameters regression(EDCPR)model.The reliability analysis of the model verifies the effectiveness of the CP curve prediction methods.The differences between the two CP curve prediction methods are compared and analyzed,and the corresponding applicability conditions are pointed out: the prediction method based on CPMPFT model is not limited by the number of core samples,and can predict the capillary pressure in the range of 0 ～ 100% mercury saturation,but this method is affected by inversion uncertainty and the shape of the T2 distribution;the prediction method based on EDCPR model can predict the CP curves directly using the NMR echo data,but this method can only obtain the CP curves within the range of experimental pressure value and needs to ensure that the number of core samples is sufficient.For the evaluation of the parameters of tight sandstone reservoir,the porosity is calculated using conventional logging and NMR logging;to address the errors in the experimental results of oil-water saturation of core samples,the compaction coefficient and the vertical projection correction method are used to correct the measured oil-water saturation,meanwhile,different saturation calculation formulas are used to calculate the oil saturation for different sedimentary characteristics;a permeability prediction method based on multi-parameter-mind evolutionary algorithm-artificial neural network model is proposed,different pore structure parameters are selected as model inputs to predict the reservoir permeability;the reservoir brittleness index is calculated using array acoustic logging and conventional logging,respectively.For the comprehensive classification of tight sandstone reservoir,two comprehensive reservoir classification methods are proposed based on the lithology characteristic evaluation parameters,petrophysical quality characteristic evaluation parameters,engineering quality characteristic evaluation parameters,and pore structure characteristic evaluation parameters: the classification method based on “comprehensive reservoir quality index”and the unsupervised classification method based on principal component analysis-simulated annealing genetic algorithm-fuzzy cluster means.Well data processing results show that the reservoir classification results match well with the lithologic profile: in the type I reservoir,the reservoir thickness is thick,the evaluation parameters are large,and the core descriptions all show oil-bearing characteristics;in the type II reservoir,the reservoir thickness is medium,the evaluation parameters are medium,and the core descriptions partially show oil-bearing characteristics;the type III reservoir is mainly composed of thin interbeds,with low evaluation parameter values and no oil-bearing features in core description.The aforementioned results demonstrate the effectiveness of the classification methods.