SOLVING THE PROBLEM OF DIFFUSION OF PETROLEUM AND WATER IN UNDERGROUND RESERVOIRS BY USING NEURAL NETWORKS

Porosity and Permeability are two important characteristicsof a hydrocarbon reservior. Yet estimation of these parameters is difficulttasks encountered in geophysical and reservior engineering. Formationporosity is often measured in the labratory from cores or well logs. Estimationof these parameter from well logs needs a prior knowledge of theformation and pore fluids.Formation permeability is often measured in thelabratory from core analysis or evaluated from test data. Another methodfor this target is use of empirical equation that any of these methods havemany problems. Artificial neural network is a new method recently is usedin oil industry for prediction of petrophysical parameters. This methodhas little cost and it’s important advantage is that no prior knowledgeabout rock materials and pore fluids is required. The main achievent ofthis research is to study of the Diffiusion problem. Then predict and approximatethem by artificial neural network. Then after definition of netstructure, these nets are generalized to another well in this field.

SOLVING THE PROBLEM OF DIFFUSION OF PETROLEUM AND WATER IN UNDERGROUND RESERVOIRS BY USING NEURAL NETWORKS Keywords:

Porosity , Permeability , Diffiusion problem , approximate , Artificial neural network