Estimation of Well Log Data by Conjunctive Use of Seismic and Petrophysical Data (Hendijan Oil Field, Iran)

Well logs are of the important tools for reservoir characterization. Parameters such as porosity, volume of shale, water saturation, permeability, lithology and production zones could be computed from processing and interpreting of well logs. Such information is obtained from petrophysical logs. Considering coverage and continuity of seismic data, log estimation in each location of reservoir by using seismic data, not only before the drilling but also after it will be applied and invaluable. So that using the predicted wells log from seismic data, reservoir evaluation and assessment over the seismic data coverage is possible. In this research, a formulation is established between well log data including sonic and density logs and their corresponding seismic attributes using neural networks in Hendijan oil field, Persian Gulf. For this purpose, optimal input seismic attributes were obtained using a forward stepwise regression. The results show a good agreement between measured and neural network predicted data at test wells. So, the estimated well logs could be used for reservoir parameters evaluation and reducing risks of future exploration and field’s development. Also such predicted well logs enable us to gain more vital parameter about a reservoir.