Development of Hybrid ACOR-BP Algorithm for Estimation of Shear Wave Velocity from Porosity Logs

Reservoir characterization is an important step for exploration and development of oil and gas fields. Shear and compressional wave velocity can provide the accurate data for petro-acoustic studies from a hydrocarbon reservoir. Compressional wave velocity (Vp) is very easily obtained from sonic logs that are available in most of oil and gas wells, but some wells, especially old wells, may not have shear wave velocity (Vs) data. In this research, shear wave velocity was predicted from porosity well log data (sonic, neutron and density) using Hybrid ACOR-BP technique. For this purpose a total 3190 data points from Asmari reservoir which have Vs and porosity log data were utilized. These data are divided into two parts, (1) one part included 2090 data points which are used for constructing model and (2) the other part included 1100 data points used for testing and validation model. Hybrid ACOR-BP has better performance in simulating reservoir characterization than NN-BP and ACO individual algorithms. Results of simulation in the test and validation stage calculated root mean squared error, and correlation coefficient 0.06 and 0.97, respectively. Results of this research show Hybrid ACOR-BP algorithm is a reliable intelligent method for estimation of Vs. This study is based on the data collected from the two wells of the Cheshmeh Khush oilfield; moreover, the results of the study can be generalized to other developmental wells.