Use of genetic algorithm in reservoir characterisation from seismic data: A case study

Abstract

In the present paper, a seismic inversion based on genetic algorithm (GA) is performed to characterise the reservoir using seismic data only from the Blackfoot field, Alberta, Canada. The algorithm is first tested on synthetically generated data to optimise the GA parameters. The error analysis between the inverted and the expected results suggested that the performance of algorithm is exceptionally satisfactory. Thereafter, the inversion is performed for real seismic data from the Blackfoot field. The seismic data is first inverted for acoustic impedance section and then it is transformed into the velocity and density sections using the relation derived from the well-log data. The interpretation of the inverted/derived results depicts a low-amplitude anomaly zone between 1055 and 1065 ms time interval, which is characterised as a reservoir. The results demonstrate the efficacy and applicability of the GA in reservoir characterisation from the seismic data alone. This study is very helpful for the offshore projects where the information about well logs are missing.