Characterization Of Igneous Reservoir Pore Structure And Classified Evaluation Of Reservoir Parameter

AS the society developing, it has a growing tendency of the demand for energy, especially for oil and gas resources. The reservoir which has been considered as important in past, has can't meet the need, while also contribute to the improvement of the techology of exploration wokersNow has set the sight on anoher type.of reservoirs. More and more contries and regions have exploration and exploitation of igeous rock reservoirs, these examples of the theory and pratical have show that igneous rocks can be good reservoirs.Combination of variying storage spaces and structure of pore of reservoir rock formed reservoir with different properties and capacity of production. Even if the reservoir has the same lithology and same genetic, it also consistes of betterness, adiaphorism and weak reservoirs.For the resivoirs which are composed by the good ,ecumeical, bad parts,the differences of capacity of store ,permeablity , validity of reservoirs and pore structure are object exsitence. Some complex reservoirs even existence cosiderable chasms.conventional approachs will find themself hard to accuratly evaluate the reservoir.Therefore, in order to give full play to different reservoir capacity, improve the level of development of gas and oil, classification and evaluation of the reservoir is necessary.Moreover, the complex structure of pore, and strong heterogeneity, so the classification of igeous rock reservoirs seems more necessary.Capillary pressure curve is the relationship between the wet phase (or non-wet phase) saturation of fluid and the capillary force curve. It is not only the function of pore throat radius and pore volume, is also a function of pore connectivity of throat of pore. It is indispensible information to study the structure.This artical base on the deep region of ying cheng zu .firstly, we compare the capillary pressure curve with each other, then classify according to different shapes and pressure of displacement, maximum and average of radius .Evey curve represent a class reservoir.Thus we connsider the capillary pressure curve into 7 categories, corresponding to seven kinds of reservoir by the sequence from good to bad.Supplemented by reservoir quality index (RQI) for comparison.According to the curve that reflects information of the structure of pore and the completeness of the well datas,we optimize the7 curves including spontaneous potential logging, natural gamma logging, deep, shallow lateral resistivity, density logging, and neutron porosity degree as the samples.There were fomerly many defects in capillary pressure curves obtained from laboratory,such as too many times for measuring ,cost too expensive to accept ,the use of mercury is toxic,and so on.Furthermore,after all, the information that the cores represent is limited.This articals consider T2 NMR spectrum which can also refelcts the structure of pore as the original datas. Using classified conventional log response of core as samples and using self-organizing neural network to classify the conventional curve are meaningful to fitting petrophysical parameter for classification and calculation of porosity, permeability and saturation, especially for igneous rocks reservoir whose all kinds of characteristics are complex.IN the progress of make sure the value C,we mainly make use of the interdependency between the two sets datas of C/T2~A m and PC ~ S Hg and don't invole some specific parameters.so the values of C is selected relatively objective and simple. So the the method that determines conversion factor is especially suitable for volcanic reservoir whose structure of pore are complex, and parameters of core are not very typical.After we get C, we obtain a whole continuous pseudo- wells capillary pressure curve.We drawn the shape of capillary pressure curve as a sample of self-organizing netrual network,in order to classify the pseudo-capillary pressure curves .The way of converting a two-demensional array of datas into a one-dimension is adapt to the requirements for samples, but also retrain the original shape of the curve information.