Porosity prediction from model-based seismic inversion by using probabilistic neural network (PNN) in Mehar Block, Pakistan

Seismic inversion is an approach to provide a useful connection between observed seismic data and interpretative elastic physical properties of potential reservoirs. For the calculation of reservoir characterization, Post stack seismic inversion plays an important role to estimate the reservoir properties i.e. porosity and acoustic impedance etc. In this paper, we use the post-stack time migrated seismic data (PSTM) and log data to delineate the reservoir properties and also the fundamental properties including acoustic impedance and porosity of target zone. To complete this work, seismic inversion and geostatistical method has been used. The methodology of inverting seismic data into acoustic impedance is pertinent to key information of this study. In addition, suitable wavelet representative of the given conditions is important for encouraging result. In the next stage, geostatistical inversion is applied by using applications of probabilistic neural network (PNN) to estimate the porosity of low impedance sand body of Lower Ranikot using well Mehar-02. By using PNN the impedance volume is transferred into porosity volume while the finally 18% porosity is predicated with the help of the petro physical properties of Mehar-02, and display on the seismic section as high porosity zone with low impedance.