نوع مقاله : مقاله پژوهشی
نویسندگان
1 دانشجوی کارشناسی ارشد، گروه زمین‌شناسی نفت و حوضه‌های رسوبی، دانشکده علوم زمین، دانشگاه شهید چمران اهواز، اهواز، ایران
2 گروه زمین‌شناسی نفت و حوضه‌های رسوبی، دانشکده علوم زمین، دانشگاه شهید چمران اهواز، اهواز، ایران
چکیده
کلیدواژه‌ها
موضوعات
عنوان مقاله [English]
نویسندگان [English]
We used petrophysical data, by means of the self-organizing neural network clustering technique, to recognize petrophysical rock types in the bangestan Formation. Using petrophysical logs, such as water saturation level, sonic and density, gamma ray, along with calculated reservoir data (effective porosity and permability) from 10 wells (5 cored and 5 uncored wells) led to recognition of 5 electrofacies (EF1-EF5). These electrofacies were used to develop our interpretations to 5 uncored wells. Flow zone index, Porosity and permeability and water saturation were used to estimate reservoir quality of each electrofacies. Rock type 1 and 2 show low percent of water saturations and high values of porosity, permeability, which characterizes them as good reservoir quality or good reservoirs rocks. Rock type 3 show proper percent of water saturations and values of porosity, permeability, which characterizes them as medium reservoir quality. Rock type 4 and 5 with low porosity or permeability, high water saturation are considered as poor reservoir quality. In the second part of this research, Permeability was then predicted from geophysical well log using artificial neural network (ANN). An advanced approach for testing and training of the ANN was developed which provided reliable predictions and consistent. This method is done in two steps; permeability estimation once for the whole interval and once for each of the determined electrofacies.The results showed that the average permeability error estimated in the electrofacies of the reservoir is lower than the permeability error obtained in the entire interval of the reservoir.
کلیدواژه‌ها [English]
)