عنوان مقاله [English]
During the last decade, there has been an increasing interest in the use of attributes derived from 3-D seismic data to define reservoir physical properties such as porosity and fluid content. Therefore, significant advances in the study and application of expert systems in the petroleum industry is needed so that we are able to use such attributes in reservoir characterizations. Establishment of an intelligent formulation between two sets of data (inputs/outputs) has been the main topic of such studies. One such topic, of great interest, was to characterize how 3D seismic data can be related to lithology, rock types, fluid content, porosity, shear wave velocity and other reservoir properties. Petrophysical parameters, such as water saturation and porosity, are very important data for reservoir characterization. So far, several researchers have worked on predicting them from seismic data using statistical methods and intelligent systems (Russell et al., 2002; Russell et al., 2003; Chopra andMarfurt, 2006).
Two sources of information are commonly available for structural modeling and reservoir characterization. These data are well log data (depth data) from wells and geophysical measurements from seismic surveys, which are often difficult to integrate.While the well data provide the most accurate measurements of depths, there are rarely enough wells to permit an accurate appraisal from well data alone. On the other handthe seismic data is generally less precise but more abundantThe Main purposeof this study was to enhance the Â characterization of subsurface reservoirs by improving the prediction of porosity through a combination of reservoir geophysics (seismic attributes) and well logs data.First, for statistical determination of reservoir parameters seismic attributes were combined by using the classical techniques of multivariate statistics and more recent methods of neural network analysis were developed.However, there were important questions to answer: Which attributes had to be combined to estimate the porosity? How the best attributes were selected to achieve the goal? Were all the attributesÂ used in different combination methods? Was there any software that contains all attributes relevant to the Â petrophysical parameters? To answer these questions, it should be noted that, generally speaking,Â conventional attributes which exist in any software were used for these ideas but each software was developed for specific tasks Â with specific attributes. Therefore, integration of different attributes from different softwares will improve process of estimation of petrophysical parameters. We used two very developed and famous softwares and their attributes for estimation of porosity. During the usage of these software programs, we found that, iso-frequency component, instantaneous bandwidth and time gain had more relation with porosity. The mentioned attributes do not exist in Hampson Russel software as main software for reservoir characterization. Then these attributes beside many other attributes extracted from the Petrel software were used in a different process of combination of attributes to estimate the porosity at well locations. For this study, well logging and seismic data were used in order to estimate the porosity in an Iranian oil field. Â At the first step, an inversion was carried out on seismic data and well logs. Subsequently, seismic attributes were extracted from the mentioned data by mathematical algorithms. Next, the extracted seismic attributes were combined using a step by step regression algorithm. In next stage, we determined a relationship between a set of seismic attributes and a reservoir parameter such as porosity in well locations by using a neural network, and then this relationship was used to calculate reservoir parameters from sets of appropriate seismic attributes throughout a seismic volume. In this study, firstly existence attributes in Hampson-Russell software with well data were used for porosity estimation. At this stage, the porosity was estimated with good accuracy. Further, to improve the estimation of petrophysical parameters, other seismic attributes from the Petrel software related to the petrophysical parameter were extracted. Then, these attributes with associated attributes available in Hampson-Russell software were used in the estimation of porosity. At this stage, the results were better than before. During this study, the best attributes that were related to reservoir characteristics from different software were used and the best combination of attributes for porosity estimation was investigated with using multilinear regression and different neural network methods