A flowchart for porosity and acoustic impedance mapping using seismic inversion with semi hybrid optimization combining simulated annealing and pattern search techniques

Porosity and acoustic impedance are important in the study of subsurface properties of rocks and soil. Porosity is influenced by the type of minerals, and fluids, and their distribution within the subsurface material. Acoustic impedance is a key parameter in seismic inversion because it governs the reflection and transmission of seismic waves at interfaces between different rock layers. Mapping porosity and acoustic impedance using seismic inversion poses several challenges such as low resolution, longer convergence times compared to other optimization techniques, and handling large datasets. To address these challenges, our current study has employed a semi-hybrid optimization approach by incorporating a pattern search (PS) method into the globally recognized simulated annealing (SA) technique. In our devised methodology, seismic data is meticulously inverted, trace by trace, initially utilizing the simulated annealing process and subsequently integrating the pattern search which further reduces computational Complexity. The output from SA serves as the foundation for the PS optimization, preventing it from getting trapped in local minima or maxima. To evaluate the algorithm, we initiated a systematic analysis using synthetic data. The hybrid optimization method performed well, yielding highly accurate inversion results with a remarkable high resolution and correlation between original and inverted impedance. We then applied this approach to actual seismic reflection data from the Blackfoot field in Alberta, Canada. Notably, the inversion identified a sand channel between 1055 and 1070 ms two-way travel time, characterized by low impedance and high porosity, suggesting the potential presence of hydrocarbon reservoirs. The level of performance demonstrated in this context may not be anticipated when utilizing SA or PS optimization alone. Hence, the newly devised semi-hybrid optimization approach emerges as a highly recommended solution, offering the potential to address the constraints of individual optimization methods and deliver thorough subsurface insights.