The knowledge of the mechanical properties of the rocks is essential in the calculation of the stability of the oil wells and in the elaboration of drilling projects, because they help in the selection of the best equipments for each geologic scene. In these activities, the most important mechanical property is the rock compressibility, which can be calculated from the sonic geophysical log, but this one, not always available in the well data set. In order to minimize this limitation, it was developed, in this work, a methodology to simulate it through an algorithm that utilizes fuzzy logic concepts, using as input data gamma ray and resistivity logs. The basic principle of this methodology is to propose that any geophysical log, including the sonic log, can be considered function of other measured geophysical logs in the same depths. On the other hand, to test the confidence of this approach, it was compared with two others commonly used in the simulation of logs: the linear multiple regression and the neural network back-propagation, showing, our methodology, however, better results. Finally, to validate the method, it was tested using wells from Namorado Oil Field in Campos Basin, which contains gamma ray, sonic and resistivity logs.
sonic; gamma ray and resistivity logs; fuzzy logic; compressibility; Namorado Oil Field; Campos Basin
