Logging Recognition Methods Of Igneous Reservoir Oil, Gas And Water

With the igneous reservoir discovered continuously, the igneous rock has already become the important part that can't be neglected in the petroleum exploration. Petroleum engineers in the world have taken the identification for oil-gas-water layers as an important question to study with much attention and effort. So, the identification of the igneous reservoir is very important for the petroleum exploration and development. But, the valid methods for igneous rock interpretation from common well logs have not been found up to now. The main purpose of this thesis is to identify oil and gas layers in the igneous reservoirs by using the neural network based on the common well logs. This thesis firstly carries on an investigation in the general situation of the igneous reservoir. The rock type of the igneous reservoir, the pore type of reservoir, the combination type of reservoir space and the structure characteristic of the reservoir are introduced. Those have been the foundation for the discussion to the logging-response characters. The logging-response characters of igneous reservoir in researched region are summed up based on actual data. The calculation methods of reservoir parameters based on common well data are introduced in thesis. But they all produce the bigger error margin. So, BP neural network technique is made use of identifying the oil-gas-water layers in igneous reservoirs. The basic structure and calculation of BP neural network are discussed in the detail in this thesis. The process to identify oil and water layers by BP neural network is given. Genetic algorithm (GA) is led into BP network in allusion to the flaws of BP network. Finally, the actual data are disposed by using the procedure of BP neural network, and the result is very good. Therefore, the neural network technique is very useful to identify the oil-gas-water layers in igneous reservoirs based on the common well logging data.